Ancient German Identity in the Shadow of the Roman Empire: The impact of Roman trade and contact along the middle Danube frontier, 10 BC–AD 166 9781407303611, 9781407333892

Table of contents :
Front Cover
Title Page
Copyright
Table of Contents
List of Tables
List of Charts
List of Figures
Preface
PART I: INTRODUCTION AND BACKGROUND INFORMATION
Chapter 1: Archaeological Goals and Methods
Chapter 2: Historical Account of Southwest Slovakia: 100 BC–AD 166
Chapter 3: Archaeological Account of Southwest Slovakia: 100 BC–AD 166
Chapter 4: Frontiers, Romanization, and German Identity
PART II: THE ARCHAEOLOGICAL PROJECT
Chapter 5: The Survey Phase of the Project at Urbárske Sedliská
Chapter 6: The Excavation Phase of the Project at Urbárske Sedliská
Chapter 7: Analysis of Huts 1–2 and Features G2–G3
Chapter 8: The German Pottery from Features G1–G4
Chapter 9: The Roman and La Tène Pottery from Urbárske Sedliská
Chapter 10: Miscellaneous Cultural Material and Faunal Analysis
PART III: SYNTHESIS
Chapter 11: German Identity in the Shadow of the Roman Empire
Appendix A: Data Collected from Shovel Test Pits
Appendix B: Catalog of Finds from the Survey Phase and Units: Non-Roman Period Artifacts
Figures
List of Journal Abbreviations
Bibliography

Citation preview

BAR S1881 2008 VRBA ANCIENT GERMAN IDENTITY IN THE SHADOW OF THE ROMAN EMPIRE

B A R Vrba 1881 cover.indd 1

Ancient German Identity in the Shadow of the Roman Empire The impact of Roman trade and contact along the middle Danube frontier, 10 BC – AD 166

Eric Michael Vrba

BAR International Series 1881 2008

13/11/2008 13:49:09

Ancient German Identity in the Shadow of the Roman Empire The impact of Roman trade and contact along the middle Danube frontier, 10 BC – AD 166

Eric Michael Vrba

BAR International Series 1881 2008

ISBN 9781407303611 paperback ISBN 9781407333892 e-format DOI https://doi.org/10.30861/9781407303611 A catalogue record for this book is available from the British Library

BAR

PUBLISHING

Table of Contents List of Tables List of Charts List of Figures Preface with Pronunciation Guide for Slovak and List of Abbreviations

vii ix xi xv

Part I: Introduction and Background Information Chapter 1. Introduction: Archaeological Goals and Methods 1.1. Significance of Southwest Slovakia for the Study of the Roman Frontier 1.2. Archaeological Goal of the Project 1.3. Romans, Celts, Germans, and Archaeological Cultures 1.4. Why Slovakia? 1.5. Background Information for Urbárske Sedliská and Mást 1.6. Archaeological Methods 1.7. Conclusion Endnotes Images

3 4 6 7 8 9 10 11 12

Chapter 2. Historical Account of Southwest Slovakia: 100 BC–AD 166 2.1. The Geographic and Chronological Framework for Slovakia from the Final La Tène to the Great Migration Period 2.2. Review of the Ancient Sources for Southwest Slovakia 2.2.1. Authors for the Final La Tène Period: 100–10 BC 2.2.2. Authors for the Early Roman Period: AD 10–166 2.3. Historical Account of the Region 2.3.1. Late La Tène: 100–10 BC 2.3.2. Early Roman Period: 10 BC–AD 166 2.4. Conclusion Endnotes Tables and Charts Images

20 20 21 22 24 24 24 29 29 35 36

Chapter 3. Archaeological Account of Southwest Slovakia: 100 BC–AD 166 3.1. Final La Tène: 100–10 BC 3.2. The B1 Phase (AD 10–70/80) 3.3. Description of Quadi Burial Practice 3.4. Description of Quadi Settlements 3.5. The B2 Phase (AD 70/80–166) 3.6. The Roman Presence Along the Middle Danube Frontier: 10 BC–AD 166 3.7. Overview of the Roman-Style Structures in the Barbaricum 3.8. The Roman-Style Structure at Stupava 3.9. Conclusion Endnotes Images

42 44 47 48 48 50 51 52 58 58 61

Chapter 4. Frontiers, Romanization and German Identity 4.1. The Roman Frontier 4.1.1. The Meaning of the Term Limes 4.1.2. The Concept of Frontiers 4.1.3. Client Kings 4.1.4. Roman Authors, Views on Imperialism, and the Instructions of Augustus 4.1.5. Germans and Soldiers on the Middle Danube Frontier 4.1.6. Comments on Dio’s “Neutral Zone” 4.2. Romanization 4.2.1. The Origins of Romanization 4.2.2. The Concept of Romanization Today 4.3. Ancient German Culture and the Topic of Identity 4.3.1. A Brief Overview of Ancient German Culture

79 79 80 82 82 83 85 86 87 88 89 89

iii

4.3.2. Cultural Identity and the Role of Objects 4.3.3. German Identity and the Middle Danube Frontier 4.4. Conclusion Endnotes

92 94 96 97

Part II: The Archaeological Project Chapter 5. The Survey Phase of the Project at Urbárske Sedliská 5.1. Geologic Information for Stupava and the Surrounding Region 5.2. The Magnetometer Survey 5.3. The Archaeological Survey 5.3.1. Survey Results in Regards to Soil, Stratigraphy, and Overall Artifact Distribution 5.3.2. Types and Quantities of Artifacts Found in the STPs 5.4. Discussion of the Artifacts 5.4.1. Pottery 5.4.2. Metal 5.4.3. Lithics 5.4.4. Brick and Slag 5.4.5. Bones 5.4.6. Other 5.5. Unit 1 5.6. Area 6 5.7. Conclusion Endnotes Charts Images

108 109 110 110 111 111 111 111 112 112 113 113 114 116 120

Chapter 6. The Excavation Phase of the Project at Urbárske Sedliská 6.1. Unit 2 6.2. Units 3, 3A–3E 6.3. Units 4, 4A–4C 6.4. Unit5 6.5. Unit 6 6.6. Unit 7 6.7. Unit 8 6.8. Summary of Findings Endnotes Charts Images

132 132 133 133 133 134 134 135 136 137 147

Chapter 7. Analysis of Huts 1–2 and Features G2–G3 7.1. The Grubenhaus 7.2. Huts 1 and 2 (Features G1 and G4) 7.2.1. Hut 1 (Feature G1) 7.2.2. Feature G2 7.2.3. Feature G3 7.2.4. Features G2 and G3 and their relation to Hut 1 7.2.5. Hut 2 (Feature G4) 7.3. Period of Occupation for Huts 1 and 2 7.3.1 The Date of Hut 1 7.3.2 The Date of Hut 2 7.4. Conclusion Endnotes Charts Images

157 159 159 161 161 162 163 165 165 166 166 167 169 171

Chapter 8. The German Pottery 8.1. Introduction to German pottery of the Early Roman Period 8.2. Typology for German pottery at Urbárske Sedliská 8.2.1. Fine-Ware Vessels: Form I 8.2.2. Fine-Ware Vessels: Form II

188 189 191 191

iv

105 105 107

8.2.3. Fine-Ware Vessels: Form III 8.2.4. Fine-Ware Vessels: Bases and Unidentifiable Rims 8.2.5. Coarse-Ware Vessels: Form I 8.2.6. Coarse-Ware Vessels: Form II 8.2.7. Coarse-Ware Vessels: Form III 8.2.8. Coarse-Ware Vessels: Bases and Unidentifiable Rims 8.3. Catalog of the types of ornamentation 8.3.1. Rolled Wheel Patterns 8.3.2. Dotting by Small Wheel 8.3.3. Filled Triangles 8.3.4. Incised Lines with Dots 8.3.5. Dimples 8.3.6. Fine Punctures 8.3.7. Barbotine 8.3.8. Grooved Lines/Furrows 8.3.9. Flute/Impressed Lines 8.3.10. Roughening 8.3.11. Wedges 8.3.12. Nail/Finger Impressions 8.3.13. Comb-Lines 8.3.14. Incised Lines 8.3.15. Finger Fluting 8.3.16. Rim Incising 8.3.17. Combination Patterns 8.4. Analysis of Forms and Decorations 8.4.1. Analysis of the Form Data 8.4.2. Analysis of the Decoration Data 8.5. Conclusion Endnotes Tables and Charts Images Chapter 9. The Roman and La Tène Pottery from Urbárske Sedliská 9.1. Types of Roman Pottery Found in Southwest Slovakia 9.1.1. Brick-Colored Ware 9.1.2. Pannonian-Striped Ware 9.1.3. Marbled Ware 9.1.4. Mortaria 9.1.5. Smooth-Gray Ware 9.1.6. Pannonian-Stamped Ware 9.1.7. Coarse-Gray Ware 9.1.8. Beakers with Impressed Walls and Thin-Walled Mugs 9.1.9. Containers with Horizontal Rims 9.1.10. Barbotine Ware 9.1.11. Raetian Ware 9.1.12. Glazed Ware 9.1.13. Terra Sigillata 9.2. Catalog of Roman Pottery from the Site 9.2.1. Brick-Colored Ware 9.2.2. Pannonian-Striped Ware 9.2.3. Mortaria 9.2.4. Smooth-Gray Ware 9.2.5. Coarse-Gray Ware 9.2.6. Beakers with Impressed Walls and Thin-walled Mugs 9.3. Discussion of the Roman Pottery from the Site in Comparison to That of Other Sites 9.4. The La Tène Pottery from Hut 1 9.4.1. Late La Tène Pottery in Southwest Slovakia 9.4.2. La Tène Pottery at Urbárske Sedliská 9.5. Conclusion Endnotes Table and Charts v

192 193 195 196 198 200 202 202 202 203 203 203 203 204 204 204 205 205 205 206 208 208 208 208 209 209 210 211 211 212 222 266 267 268 268 268 269 270 270 270 271 271 271 272 272 273 273 274 275 275 275 275 275 276 276 277 277 278 279

Images

281

Chapter 10. Miscellaneous Cultural Material and the Faunal Analysis 10.1. Roman Imports Other Than Pottery Recovered from German Settlements and Graves 10.1.1. Metal Imports 10.1.2. Lamps 10.1.3. Glass Products 10.1.4. Beads 10.1.5. Items from Bone 10.2. Artifacts Recovered from Huts 1 and 2 and Features G2 and G3 10.2.1. Bronze Artifacts 10.2.2. Iron Artifacts 10.2.3. Bone Hair-pin 10.2.4. Bead 10.2.5. Stone 10.2.6. Slag 10.3. Spindle Whorls Possibly of the Roman Period 10.4. Animal Husbandry among the Germans 10.5. Analysis of the Faunal Assemblage from Urbárske Sedliská 10.5.1. Analysis of Bones from Stratum 2 of Units 2–4 10.5.2. Analysis of Bones from Huts 1 and 2 10.5.3. Comparison of the Faunal Assemblage from Urbárske Sedliská with That of Huts from Other Sites 10.6. Summary of Findings Endnotes Tables and Charts Images

292 292 293 294 294 294 294 294 295 296 296 296 297 297 297 298 298 299 300 300 301 302 306

Part III: Synthesis Chapter 11. German Identity in the Shadow of the Roman Empire 11.1. Southwest Slovakia: Archaeological Overview of the Quadi Presence 11.2. Summary of the Project at Stupava-Urbárske Sedliská 11.3. How Does Urbárske Sedliská Enable a Better Understanding of Ancient German Identity? Endnotes Charts

317 319 323 325 326

Appendix A: Summary of Artifacts Found in the Shovel Test Pits

327

Appendix B: Catalog of Non-Roman Artifacts Found in Survey and Excavation B.1. Pottery B.1.1. Early Neolithic: LBK B.1.2. Middle to Late Neolithic: Lengyel B.1.3. Late Neolithic to Early Eneolithic: Bajč-Retz B.1.4. Middle Eneolithic to Final Late Bronze Age B.1.5. Great Migration Period to Present Day B.2. Metal B.3. Stones B.4. Coins Endnotes Figures

333 333 334 335 335 336 336 337 337 338 339

List of Journal Abbreviations Bibliography

345 346

vi

List of Tables 2.1. 3.1. 3.2. 5.1. 5.2. 5.3. 5.4. 5.5. 6.1. 6.2. 6.3. 6.4. 7.1. 7.2. 7.3. 7.4. 7.5. 7.6. 7.7. 7.8. 7.9. 7.10. 7.11. 8.1. 8.2. 8.3. 8.4. 8.5. 8.6. 8.7. 8.8. 8.9. 8.10. 9.1. 9.2. 9.3. 9.4. 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. 10.16.

Archaeological Periods with Subdivisions for Slovakia and Only the Roman Period for Bohemia/Moravia Stamp Types Found on Roof-tiles and Their Find-spots at Stupava Terra sigillata Found at Stupava STPs Dug Deeper Than 60 cm Artifact Totals from the Surveyed Area Pottery Totals from the Surveyed Area STPs with Bone Artifact Totals from Unit 1, Strata 2 and 3 Units and Their Size in Meters Pottery Totals from Unit 6, Strata 2 and 3 Pottery Totals from Unit 7, Strata 1–3 Artifact Totals from Stratum 1 Distances Between Postholes of Hut 1 (in meters) Diameter and Depth of Hut 1 Postholes (in centimeters) Pottery from Grubenhaus 1 (Feature G1) Pottery Types and Amounts from Feature G2 Artifacts from Feature G2 Pottery Types and Totals from Feature G3 Artifact Types and Amounts from Feature G3 Distance between Postholes for Hut 2 (in meters) Diameter and Depth of Hut 2 Postholes (in centimeters) Pottery Amounts and Weights (in grams) from Hut 2 (Feature G4) Radiocarbon Dates from the Two Charcoal Samples Form I Form II Form III List of Abbreviations Used in the Catalog The 17 Decoration Types Total Frequency of Occurrence for the Various Forms Summary of the Form Totals in Fine and Coarse Wares by Amount and Percentage from Huts 1 and 2 Total Number of Each Form Found in the Various Huts Frequency of Occurrence for the Various Decorative Types by Feature Summary of the Occurrence of the Various Decoration Types at the Various Sites Mentioned in the Text Distribution of Roman Pottery at Urbárske Sedliská Pottery Present in the Various German Huts Frequency of Occurrence of Roman Pottery Types Found in German Huts Vessel Forms Present at the Various German Huts Iron Artifacts Found by Feature Bone/Teeth from Units 2–4 and 8 Animals Present and Their MNI and NSIP from Units 2–4, Stratum 2 Information Gleaned from the Bones, Units 2–4, Stratum 2 Bones Showing Teeth/Gnawing Marks, Units 2–4, Stratum 2 Bones Showing Cut or Chopping Marks, Units 2–4, Stratum 2 Bones Showing Evidence of Burning, Units 2–4, Stratum 2 Bone and Teeth Totals from Huts 1 and 2 Animals Present and Their MNI and NSIP from Huts 1 and 2 Total Amount of Bones per Animal and Totals of Domestic and Wild Animals Present in the Two Huts Information Gleaned from the Bones, Huts 1 and 2 Bones Showing Teeth/Gnawing Marks, Huts 1 and 2 Bones Showing Cut or Chopping Marks, Huts 1 and 2 Bones Showing Evidence of Burning, Huts 1 and 2 Ages of Animals Found at Urbárske Sedliská Comparison of MNI Totals of Urbárske Sedliská Huts with Those from Moravia vii

35 57 57 107 109 111 112 112 132 134 134 135 159 160 161 161 161 162 162 164 164 165 165 190 190 190 212 202 213 209 210 210 213 279 276 276 276 295 298 298 302 302 302 302 299 302 303 303 303 303 304 299 304

11.1. 11.2. 11.3. 11.4. B.1.

Summary of Pottery Totals from the Five Huts in the Stupava Area Summary of Pottery Totals Found in Huts Dating to AD 150–166 Summary of Pottery Totals Found in Huts Dating to AD 166–180 Summary of Pottery Totals Found in Huts Dating to AD 180–250 Metal Items from Units and STPs

viii

322 322 322 322 336

List of Charts

5.1. 5.2. 5.3. 5.4. 5.5. 5.6. 5.7. 5.8. 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. 6.15. 6.16. 6.17. 6.18. 6.19. 7.1. 7.2. 7.3. 7.4. 8.1. 8.2. 8.3. 8.4. 8.5. 8.6. 8.7. 8.8. 8.9. 8.10. 8.11.

Percentages of Artifacts Found in the Entire Survey Area Percentages of Artifacts Found in Area 1/2 Percentages of Artifacts Found in Area 3/4 Percentages of Artifacts Found in Area 5 Percentages of Pottery by Period/Culture in Entire Survey Area Percentages of Pottery by Period/Culture in A1/2 Percentages of Pottery by Period/Culture in A3/4 Percentages of Pottery by Period/culture in Area 5 Pottery by Period/Culture from Unit, Stratum 1, Showing Amount and Percentage Artifacts from Unit 2, Stratum 1, Showing Amount and Percentage Pottery by Period/Culture from Unit 2, Stratum 2, Showing Amount and Percentage Artifacts from Unit 2, Stratum 2, Showing Amount and Percentage Pottery by Period/Culture from Unit 3, Stratum 1, Showing Amount and Percentage Artifacts from Unit 3, Stratum 1, Showing Amount and Percentage Pottery by Period/Culture from Unit 3, Stratum 2, Showing Amount and Percentage Artifacts from Unit 3, Stratum 2, Showing Amount and Percentage Pottery by Period/Culture from Unit 4, Strata 1 and 2, Showing Amount and Percentage Artifacts from Unit 4, Strata 1 and 2, Showing Amount and Percentage Pottery by Period/Culture from Unit 5, Stratum 1, Showing Amount and Percentage Pottery by Period/Culture from Unit 5, Stratum 2, Showing Amount and Percentage Artifacts from Unit 5, Stratum 2, Showing Amount and Percentage Pottery by Period/Culture from Unit 8, Stratum 2, Showing Amount and Percentage Pottery by Period/Culture from Unit 8, Stratum 3, Showing Amount and Percentage Artifacts from Unit 8, Stratum 2, Showing Amount and Percentage Artifacts from Unit 8, Stratum 3, Showing Amount and Percentage Pottery by Period/Culture from all Units, Showing Total Amount and Percentage Artifacts from all Units, Showing Total Amounts and Percentage Pottery from Grubenhaus 1, Showing Amount and Percentage Artifacts from Grubenhaus 1, Showing Amount and Percentage Pottery Amounts and Percentages from Hut 2 (Feature G4) Artifacts from Hut 2, Showing Amount and Percentage Frequency of Forms I–III by Percentage Frequency of Forms Shown by Percentage (Only those forms shown which had representative sherds from the features) Frequency of Forms Found in Fine Ware by Percentage Frequency of Forms Found in Coarse Ware by Percentage Frequency of Forms from Hut 1 by Percentage Frequency of Forms from Hut 1 by Percentage (Without Bases and Unidentified Rims) Frequency of Forms from Hut 2 by Percentage Frequency of Forms from Hut 2 by Percentage (Without Bases and Unidentified Rims) Total by Percentage of Decoration Types Found in the Four Features Total by Percentage of Decoration Types Found in the Four Features Including Undecorated Total by Percentage of Decoration Types Present in Hut 1 ix

116 116 117 117 118 118 119 119 137 137 138 138 139 139 140 140 141 141 142 142 143 143 144 144 145 145 146 169 169 170 170 214 214 215 215 216 216 217 217 218 218 219

8.12. 8.13. 8.14. 8.15. 9.1. 9.2. 9.3. 10.1. 11.1. 11.2.

Total by Percentage of Decoration Types Including Undecorated Found in Hut 1 Total by Percentage of Decoration Types Found in Hut 2 Total by Percentage of Decoration Types Including Undecorated in Hut 2 Shows the Percentage Frequency of the Decoration Types from all Eight Huts Combined Roman Pottery Found at Urbárske Sedliská by Percentage Distribution by Percentage of the Roman Pottery at Urbárske Sedliská Percentage Frequency of Roman Pottery Vessel Forms Found at Urbárske Sedliská Animals Present at Urbárske Sedliská by Number of Bones Found, Without the Unidentified Category Artifact Totals and Percentages from Hut 1, Urbárske Sedliská Artifact Totals and Percentages from Hut 2, Urbárske Sedliská

x

219 220 220 221 279 280 280 305 326 326

List of Figures 1.1. 1.2. 1.3. 1.4. 1.5. 1.6. 1.7. 1.8. 1.9. 1.10. 1.11. 1.12. 1.13. 1.14. 2.1 2.2 2.3. 2.4. 2.5. 2.6. 2.7. 2.8. 2.9. 3.1. 3.2. 3.3 3.4. 3.5. 3.6. 3.7. 3.8. 3.9 3.10. 3.11. 3.12. 3.13. 3.14. 3.15. 3.16. 3.17. 3.18. 3.19. 3.20. 3.21. 3.22. 3.23. 3.24. 3.25. 3.26. 5.1. 5.2. 5.3. 5.4. 5.5.

Modern-day Slovakia Roman provinces along the Danube River and location of Roman-style structures beyond the frontier Aerial view of Cífer-Pác View of the bath-building at Dúbravka during excavations Archaeological plan of the excavations at Milanovce Map of Stupava View of area surveyed at Mást Areas surveyed at Stupava in 1982 Grubenhaus excavated at Urbárske Sedliská by Dr. Turčan A sample of German pottery from the Grubenhaus A sample of Roman provincial pottery from the Grubenhaus German pottery and crucibles from the Grubenhaus Metal artifacts from the Grubenhaus Map of Urbárske Sedliská showing location of trenches dug by Dr. Farkaš Chronology I: Hallstatt and La Tène Periods Chronology II: Roman and Great Migration Periods Map of Slovakia showing the different regions Map of the Middle Danube region around 80 BC Map of the Middle Danube region around 40 BC Map of the Middle Danube region around 7 BC The Empire of Maroboduus Map of the Middle Danube showing the extent of Quadi territory Map of the Middle Danube showing main paths of the Amber Route Map of Slovakia showing La Tène period sites Map of Moravia showing the La Tène settlement area with German incursions Archaeological groups of the Middle Danube c. AD 50 Map of sites mentioned in Chapters 2, 7–9 Site-plan of Dúbravka Site-plan for the settlement at Branč Site-plan for Bratislava-Trnávka “Zadné” Site-plan for Blučina Site-plan for Křepice Site-plan for Komořany Site-plan for Hrušky Map of Pannonia by AD 200 Map showing location of Stupava Plan of the site at Stupava according to Ondrouch’s excavations Plan of the site at Stupava after the 1970s excavations Wall-painting fragments from the site Photo of the windows in-situ at the site Reconstruction of the heating system and suspended floor in Building D Reconstruction of canal KV1 Reconstruction of canal KV2 Plan of CIII from Stupava The site at Stupava with nearby German huts marked Lion-head ornament from a helmet Examples of Legionary stamps from Stupava Late Roman German pottery from Stupava Various finds from the site at Stupava Map of southwest Slovakia showing location of Záhorie Lowlands Map of the Pannonian and Vienna Basins Map showing location of survey area and its subdivision Map showing area covered by the magnetometer survey View of the cesium magnetometer

xi

12 12 13 13 14 14 15 15 16 16 17 18 19 19 36 37 38 38 39 39 40 40 41 61 62 62 63 65 66 67 67 68 69 69 70 71 72 73 74 74 75 75 75 76 76 77 77 78 78 120 121 121 122 122

5.6. 5.7. 5.8. 5.9. 5.10. 5.11. 5.12. 5.13. 5.14. 5.15. 5.16. 5.17. 5.18. 5.19. 5.20. 5.21. 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. 6.15. 6.16. 6.17. 6.18. 6.19. 6.20. 6.21. 7.1. 7.2. 7.3. 7.4. 7.5. 7.6. 7.7. 7.8. 7.9. 7.10. 7.11. 7.12. 7.13. 7.14. 7.15. 7.16. 7.17. 7.18. 7.19. 7.20. 7.21. 7.22. 7.23. 7.24.

Dr. Tirpák and his crew during the survey Map of surveyed area produced by Dr. Tirpák using Geosoft software Map of Residual Magnetic Anomalies Same map as in Fig. 4.20 but with the various anomalies circled Map showing layout of transects Crew working along transects 1 and 2 Shovel test pit at transect 1, meter mark 5 Example of form used during the survey Types of sampling using shovel test pits Two examples of STPs for the two types of stratigraphy encountered on site Photograph of Unit 1 after excavation Drawing of the strata for Unit 1 Magnetometer Map of the magnetic readings of the soil Image of the German sherds from Unit 1 Area 6 is on the other side of the ruined house LBK knobs 1 and 2 from Area 6 Map of the survey area and position of units Close-up map showing position of Units 3–5, 7 and 8 Drawing of Unit 2 showing the sectors Photograph of the plough furrows Photograph of dark stains Photograph of dark stains with Grubenhaus indicated Photograph of north baulk with posthole evident Profile drawing of north baulk of Unit 2 Map showing all the units making up Unit 3 Profile of Unit 3, south baulk Image showing the dark stain of the hut, with the pipe cut indicated Posthole in STP 3-E-15 Map showing sectors of Units 4, 4A, 4B and 4C Entire hut revealed as a stain before its excavation Reconstructed profile of south baulk from Unit 4 Baulk drawing for Unit 5 Photo showing the excavated pit in Unit 6 Profile drawing of the south baulk of Unit 6 Reconstruction of probable size of the pit Profile of Unit 7 Profile drawing of south baulk for Unit 8 Reconstruction of what a Grubenhaus may have looked like Reconstruction of German long-houses from Feddersen Wierde The five types of Grubenhäuser as proposed by Titus Kolník The four subcategories of Type III huts Ground plans of two Grubenhäuser with a pit present near the entrance niche Hut 19/61 from Branč with a storage pit at the center Ground plans of Huts 1 and 2 with pipe cut and features G2 and G3 shown Ground plan of Hut 1, showing the depth of postholes and other features Layout of line measurements and labels for the postholes Profiles of the seven postholes found in Hut 1 Outline of stains indicating the modern trench for a pipe Photograph of Unit 3 with hut fill removed and stratum 4 revealed Feature G2 is clearly seen here as the lighter colored stain The west end of the modern pipe-cut excavated to reveal the plastic pipe Daub 437 was recovered in stratum 2 of Unit 3 Close-up photograph of one side of daub no. 433 The east end of Feature G2 showing the microstratigraphy Field mouse den at the center of the photograph Image A shows the baulk between Units 2 and 3 with pit portion indicated. Image B shows the pit excavated The ditch portion of Feature G3 with fill still in place Feature G3 with pit and ditch indicated Profile of West baulk of Unit 3 extension View of Hut 1 before the fill (stratum 3) is excavated Ground plan of Hut 2, showing the depth of postholes and other features xii

123 123 124 124 125 125 126 127 128 128 129 129 130 130 131 131 147 148 148 149 149 150 150 151 151 151 152 152 153 153 154 154 154 155 155 156 156 171 171 172 172 173 173 174 174 175 175 176 176 177 177 178 178 179 179 180 180 181 181 182 182

7.25. 7.26. 7.27. 7.28. 7.29. 7.30. 7.31. 8.1. 8.2. 8.3. 8.4. 8.5. 8.6. 8.7. 8.8. 8.9. 8.10. 8.11. 8.12. 8.13. 8.14. 8.15. 8.16. 8.17. 8.18. 8.19. 8.20. 8.21. 8.22. 8.23. 8.24. 8.25. 8.26. 8.27. 8.28. 8.29. 8.30. 8.31. 8.32. 8.33. 8.34. 8.35. 8.36. 8.37. 8.38. 8.39. 8.40. 8.41. 8.42. 8.43. 8.44. 9.1. 9.2. 9.3. 9.4. 9.5. 9.6. 9.7. 9.8. 9.9. 9.10. 9.11.

Hut 2 (Feature G4) in Unit 4 before the fill is removed Profile of the Hut 2 fill along the long axis from the southeast to northwest Profiles of the seven postholes found in Hut 2 Plan of hut 2 with the fifteen identified stains indicated with a number Profiles of small stains 1–5 and 15 from Hut 2 (Feature G4) Image A shows a chunk of daub with a flat outer surface Overall map of the site with the two huts in situ (not to scale) Map of sites mentioned in Chapters 7–9 Varsik’s Form I Typology Examples of Form I vessels Varsik’s Form II Typology Examples of Form II vessels Varsik’s Form III typology Varsik’s Form IV typology Examples of Form III vessels Fine-ware Form I Fine-ware Form II Fine-ware Form III Fine-ware Form III Fine-ware Bases Fine-ware Unidentified Rims Coarse-ware Form I Coarse-ware Form I Coarse-ware Form I Coarse-ware Form II Coarse-ware Form II Coarse-ware Form II Coarse-ware Form II Coarse-ware Form III Coarse-ware Form IIIA and IIIB Coarse-ware Form IIIB Coarse-ware From IIIB Coarse-ware Form IIIB Coarse-ware Form IIIC and Bases Coarse-ware Bases Coarse-ware Unidentified Rims Coarse-ware Unidentified Rims Ornament Typology I–VI Ornament Typology VII–IX Ornament Typology X–XII Ornament Typology XIIc–e Ornament Typology XIIf–h Ornament Typology XIIi–n Ornament Typology XIIIa–f Ornament Typology XIIIg–k Ornament Typology XIIIl–m Ornament Typology XIIIo–s and XIVa and b Ornament Typology XIVc Ornament Typology XIVd Ornament Typology XIVe and f Ornament Typology XV–XVII Map showing the location of sites mentioned in Chapters 7–9 Examples of Brick-Colored Wares Examples of various Roman wares Examples of various Roman wares Examples of various Roman wares Examples of various Roman wares Brick-Colored Wares from Urbárske Sedliská Pannonian-Striped Ware from Urbárske Sedliská La Tène pottery: Types 1–5 La Tène pottery: Types 6–10 La Tène base from Urbárske Sedliská xiii

183 183 184 185 186 187 187 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 281 282 283 284 285 286 287 288 289 290 291

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. 10.16. 10.17. 10.18. B.1. B.2. B.3. B.4. B.5. B.6. B.7. B.8. B.9. B.10. B.11. B.12. B.13.

Grave-goods from a German elite burial at Zohor Silver lanx from Krakovany-Stráže Various Roman fibulae found in Slovakia Bronze armband from Grave 26, Abrahám Reconstruction of a noric-pannonian style belt Two examples of belt-buckles Partial inventory of items from Grave 35, Kostolná pri Dunaji Belt end-piece with spoon-like end Iron knife blade with tong from Grave 44, Abrahám Iron scissors from Grave 40, Abrahám Inventory of items from Huts 1 and 2, Urbárske Sedliská Gaming stone/sling-bullet Gaming stone/sling-bullet Pestle Broken pestle (15), a mortar (16) and a whetstone (13) Reverse of the broken pestle, mortar and whetstone A possible polishing stone for applying graphite The four spindle whorls recovered from Units 3 and 4 Various LBK vessels (Not to scale) LBK base, body, and handle LBK body sherds with incised line decoration at bottom LBK handle (conserved in lab with some plaster addition) Lengyel vessels (early phase) from Nitriansky Hrádok-Zámeček Bajč-Retz sherd Bajč-Retz sherd Bajč-Retz sherd Bird Caller Bird Caller with the dowel removed and opened to show filament Metal pin Sling-stone Whetstone

xiv

306 307 308 308 309 309 310 310 311 311 311 312 312 312 313 313 314 314 339 340 340 341 341 342 342 342 343 343 344 344 344

Archaeological Museum in Bratislava, had faith in my sincerity about wanting to do archaeology in Slovakia when first introduced and then referred me to Vladimír Turčan. At the Archaeological Museum, a large group of people provided a wide range of help, from identifying pottery to conservation and storage. I thank Zdeněk Farkaš (also for allowing me to use unpublished work and material from Urbárske Sedliská), Juraj Bartík, Igor Bazovský, and also the administrative secretary for allowing me to do so much photo-copying. From the conservation labs at the museum, the greatest thanks go to Marta, who did all the conservation work on the pottery and to Mr. Slovak, who dealt with the metal artifacts. Radoslav Čambal provided help in understanding the pottery as well. Peter Šimčík deserves recognition for his amazing drawings of the artifacts, which in some ways look better through his work than they ever could in reality. Alena Šefčáková did the faunal analysis. The greatest thanks go to my crew: Peter Nagy, Daniela Nagyová, Vratislav Zervan, Michal Fratrič and from the survey portion, Andre Mongiello. Their energy and drive to get the job done were without equal, but above all, their friendship and openness to my vision of what I wanted done enabled the project to be a success.

Preface This book is a modified version of my doctoral dissertation, “Beyond the Roman Frontier: A Case-Study in Slovakia of the Impact of Roman Trade and Culture on Ancient German Settlements,” which was completed in 2007 at Boston University. The archaeological project described in this work and the dissertation that grew from it would not have been possible without the help of a large group of people, both here in the United States and in Slovakia. I shall do my best to thank everyone, but if I have skipped a name, please know that it was not intentional. In the United States my greatest thanks go to Valentin and Eva Vrba. Their unwavering faith in my ability to complete this work and the financial support they unquestioningly gave formed the foundation upon which everything else was built. I also want to thank my dissertation readers James R. Wiseman, Curtis Runnels, and Peter S. Wells for their help in making the dissertation better and persevering through the editing process. Christopher Roosevelt and Ksenija Borojevic comprised the remainder of the dissertation committee and provided helpful suggestions for both the dissertation and other matters. Special recognition goes to James R. Wiseman, who was a superb mentor during my stay at the department and whose detailed knowledge of the GrecoRoman world inspired me. A note of thanks goes to Paul Goldberg, who also read portions of the dissertation. In the Department of Archaeology, Mary Beaudry, Ricardo Elia, Norman Hammond, and Paul Zimansky (now at SUNY) provided helpful advice about various topics related to the dissertation process and their fields of interest. The Humanities Foundation, the Department of Archaeology, and the Center for Archaeological Studies at Boston University provided funds that helped partially finance the project and tests. Anna-Margarite McCann made available funds for me to participate in a pottery seminar at the American School at Rome in 2000, which provided important help in understanding pottery, but also from where I traveled to Slovakia and laid the foundations for the dissertation project. Archer Martin hosted the seminar at Rome and shared his erudite knowledge of Roman pottery.

There were also other archaeologists who provided a lot of advice and observations about various aspects of the project. From the Komenius University, Edward Krekovič provided a lot of publications and help with Roman pottery. Vladimír Varsik provided copies of his publications and help with German pottery. From the Archaeological Institute at Nitra, I also thank Ján Tirpák for his magnetometer survey and Ján Rajtár, Kristian Elschek, and Titus Kolník for various articles, publications, and advice. My family in Slovakia not only provided a lot of logistical help to make the project succeed, but their openness in hosting me and willingness to go that extra length to help out truly made me feel lucky. Above all, special thanks go to Karol and Katarina Pritz, Miloš and Etela Bučko, and Ján and Zuzana Rafajdus for opening their homes to me and making me feel as if I had always been a part of the family and not a distant relation from far away. Karol Pritz provided wonderful conversations that helped take my mind off the stress during the project and Katarina Pritz deserves recognition for being a wonderful host while I lived in Bratislava. Ján Rafajdus ran around Bratislava with me and constructed the shifters based on my scrawls. I am also grateful to Zuzana and Ján Rafajdus for introducing me to Štefan Holčík which started everything back in 1999. Thank you also to Miloš and Etela Bučko for their help in obtaining various supplies and storing them, driving me all over the place when needed, as well as providing a great retreat away from Bratislava and Stupava, when I would stay with them at Nové Mesto Nad Váhom. Thanks also to Vojtech and Helena Zeman and everyone else in Vajnory for the great meals and company.

In Slovakia, I want to thank especially Vladimír Turčan, who allowed me to participate in excavations at Stupava, made available the sites of Urbárske Sedliská and Mást, but even more importantly has been a wonderful mentor over the years. Through his help I was able to obtain necessary background literature; his insightful advice enabled me to complete the archaeological project within the time allotted; and also because of him I met a large group of archaeologists, not only from Slovakia, but also Moravia and Austria who gave me a better appreciation for the archaeology of the Middle Danube region. Ivan Staník provided some great conversations about Roman and Medieval architecture, the phases at Stupava, and many other topics. Štefan Holčík, former director of the

Last of all, but not at all the least, I want to give a very special thank you to Ilean, my wife, who put up with a xv

lot, especially my ill-humor and melancholy during various stages of the writing process. Her unfailing good humor and upbeat spirit are something to marvel at and her bottomless ability to put up with me is astounding. She was a rock for me during the writing process, and without her loving support and cajoling to stay on topic, the dissertation may still lay unfinished.

The letter h is pronounced with slight voicing and ch is pronounced from the back of the throat with a guttural sound. The letter j is pronounced as the y in yes. The letters č, š, and ž are palato-alveolar consonants and are pronounced as follows: č š ž

ch in chop sh in shop zh as in the letter s in pleasure

Pronunciation Guide for Slovak The letters ĺ and ŕ have the status of long vowels in words such as hĺbka and fŕkať, or vŕba, where an English speaker would expect to have the vowel e before the letter l and r. The digraph dz is pronounced as in the word adz; dž is dzh.

The Slovak language is a western Slavic language related to Czech and Polish. The language uses the Latin alphabet with additional diacritical marks and digraphs, which are letter combinations representing separate sounds. To help the reader pronounce the various names and places mentioned in the book, a list of letters with their pronunciation is given below. The list is derived from O. Swan and S. Galova-Lorinc’s Beginning Slovak (Slavica Publishers, Inc., 1990), pp. xxx–xxxvii.

List of Abbreviations For the ancient sources mentioned in the text, in both author’s names and their works, I use the abbreviation system devised for the Oxford Classical Dictionary, 3rd edition (1996, xxix–liv). For journal abbreviations see bibliography.

The Alphabet a á ä b c č d ď dz dž e é f g h ch i í j k l ľ ĺ m n ň o ó ô p q r ř ŕ s š t ť u ú v (w x) y ý z ž

The following abbreviations are used in reference to institutions that deal with archaeology in Slovakia.

Vowels a á ä e é í/ý i/y o ó ô u ú

u in cup a in father a in have e in get e in fiancé ea in seat, but without a glide ea in seat, but with the length of i in sit o in off oo in floor diphthong consisting of uo u in put oo in stool

Consonants The letters b, d, f, g, m, n, s, v, and z have the same phonetic value as in English. The letter c is pronounced ts. The letter r is similar to the trilled r in Spanish (but trilled less energetically). The letters w and x are used in words of foreign origin such as waltz or taxi. The consonants ď, ň, and ť are palatalized dental consonants and ľ is a palatalized lateral (formed by an occlusion with the tongue in such a way that breath escapes along the sides). These letters are produced by pronouncing the sounds d, n, t, and l while holding the tongue against the palate, creating an effect similar to the ñ in Spanish, so that there is a ya sound after the consonant. ď = dya ť = tya ň = nya ľ = lya

xvi

AM-SNM

Archeologické múzeumSlovenské národné museum

AÚ-SAV

Archeologického ústavu Slovenskej akadémie vied

SNM

Slovenské Národné Múzeum

PART I INTRODUCTION AND BACKGROUND INFORMATION

region in antiquity also became a transitional zone between the urban societies to the west and the seminomadic peoples to the east. During what Slovak archaeologists call the Roman Period in Slovakia, from the last quarter of the first century BC to the end the fourth century AD, there was heightened cultural interaction along the Middle Danube, encompassing southwest Slovakia and the northern portions of Austria and Hungary. Slovakia in antiquity was part of the region called the “Barbaricum” by the Romans, 4 which encompassed the Germanic areas controlled by the Marcomanni and Quadi as well as territory occupied by the Sarmatian Iazyges (Fig. 1.2). Roman traders and soldiers had close dealings with the Quadi and, in fact, many Germans from various parts of the barbaricum served in the imperial army and later in Roman government positions.5

Chapter 1 Archaeological Goals and Methods 1.1. Significance of Southwest Slovakia for the Study of the Roman Frontier Slovakia is a convergent zone of three interrelated spheres of study in Roman history and archaeology. These three spheres are the frontier, Romanization, and cultural identity.1 In terms of Roman history and archaeology in general, the frontier and Romanization have been studied by various scholars for many years (see Haverfield 1905 and 1912; Dyson 1985; Millett 1990; Blagg and Millett 1990; Wood and Queiroga (eds.) 1992; Whittaker 1994 and 2004; Elton 1996; Mattingly 1997; Freeman 1997; MacMullen 2000; Burns 2003), while cultural identity, having been usually in the hands of anthropologists or archaeologists studying other parts of the world, has finally found a voice among classical archaeologists studying Romans and non-Romans of antiquity (see Todd 1975 and 1998; Wells 1980, 1999, 2001; Wolfram 1990; Laurence and Berry 1998; Miles 1999; Whittaker 2004). In Slovakia (or the former Czechoslovakia) these three spheres of study have been explored in various ways (see especially Eisner 1933; Ondrouch 1938; Pelicán 1960; Dobiáš 1964; Kolník 1962, 1964, 1966, 1977, 1984, 1991, 1998a; Friesinger et al. 1994; Pieta et al. 1995),2 but most of the publications are in Slovak, Czech, or German, thereby limiting how much of this research has reached an English speaking audience. The three spheres overlap in Slovakia with the study of the Roman frontier along the Danube, of the “Romanization” of the people living along this frontier, and of how the presence of the Roman Empire as a neighbor influenced the cultural identity of groups living in Slovakia. These groups include the Germanic Quadi tribe as well as a surviving Celtic culture in the north and a mixture of Sarmatian and Dacian groups in the east.3 The Quadi were at times clients of the Romans and had kings appointed by the Roman Emperor, while at other times the Germans asserted their independence and fought against the Romans. Because of the close proximity of the Quadi to the Roman Empire, it is also possible to study how the Germanic cultures in Slovakia affected the Romans and provincial residents along the frontier of the province of Pannonia.

The frontier along the Middle Danube was not a barrier to people and both Romans and Germans moved freely across it into each other’s territory in times of peace. There had been attempts in the past to identify the frontier as a barrier and a way to keep the “barbarians” outside (Luttwak 1976, 80, 87–88, 193; Whittaker 1994, 3 and 6–7; Wells 1999, 125); archaeological evidence and literary documentation show, however, Roman activity beyond the frontier and German travel into the provinces, thereby implying a fluid frontier. Recent works by C. R. Whittaker, H. Elton, and P. Wells have taken this evidence into account, creating a more realistic view of what the frontier was like (Whittaker 1994 and 2004, 8– 9; Elton 1996, 4, 5, 111; Wells 1999, 125–126; Wells 2005, 50). In their views, the Middle Danube region saw the interaction of many different parts of Roman and German society that are visible as superimposed zones of interaction (Elton 1996, 4, 5, 8–9, 111; Wells 1999; Whittaker 2004). Originally, Whittaker argued in his 1994 publication that the economic value of the land was the primary cause for Roman expansion and creation of the frontiers, and that other issues were secondary considerations (Whittaker 1994, 85–86, 95, 97, 131–132), but since then he has taken a more complex view of the frontier as a zone of interaction among many elements of cultures present there (Whittaker 2004, 5–6). The idea of the frontier as a zone of interaction among various groups, creating a complex mix of various peoples with varied backgrounds is something that Lightfoot and Martinez have argued should be the primary concept of the frontier (Lightfoot and Martinez 1995). One problem in dealing with a fluid frontier and Roman activity in what can be considered foreign territory is trying to define those activities with precision, beyond the obvious case of warfare. There are archaeological sites in Slovakia that tantalize us with Roman material, but fall short of showing definite Roman presence. Sites such as Cífer-Pác, Dúbravka, or Milanovce have buildings that appear architecturally Roman amidst German settlements, but have no clear sign of actual Roman occupation, instead implying German habitation (Figs. 1.2–1.5). Archaeologists use the term “Roman-style” to describe the buildings because these structures incorporate Roman

The region of southwest Slovakia has been a frontier between cultures for many centuries. The country is situated along the Danube River, which connected the region to the wider ancient world as trade and people moved along it (Fig. 1.1). Slovakia also straddles part of the Amber Route, the path used from prehistoric times to late antiquity for the trade of amber and other raw materials from northern Europe, along the Baltic Sea, for metal and other manufactured items from the Mediterranean region (Todd 2004, 95–96) (Fig. 2.9). The

3

architectural design and material, but do not have strong evidence of Roman habitation.6 This situation has caused much speculation as to the purpose of such structures and what the Romans, if they were indeed there, were doing amidst the Germans. Explanations vary from identifying these buildings as living quarters for German elite to Roman military outposts, but so far the evidence is inconclusive. These sites possibly represent some type of Roman influence that may not involve the physical presence of Romans.

focused on the Early Roman Period, AD 10–166, but provide also some background information for the Late La Tène because some Celtic material was found in one of the huts. On a wider level, this case study also provides information on how Roman and Germanic cultures interacted and about the influence that was exerted by each of them consciously or subconsciously on the other. When one considers the cultural interaction between Germans and Romans and how the two groups influenced each other, it is clear that Urbárske Sedliská was occupied by Germans using Roman pottery and metal products that appealed to their own needs, such as for competition among households, gift exchange, or for functional value in cooking and other household or productive activities. It is well known that Roman material was used by neighboring cultures at other frontiers, such as along the Rhine and north of Hadrian’s Wall in Britain. I am not concerned to see if the use of Roman material happens but instead to try to explain why this happens. Numerous archaeological reports provide evidence of non-Romans who lived near Romans and used Roman manufactured items including Germans living in the territory of Slovakia near the frontier. This study, however, will demonstrate that Germans maintained their cultural identity and used foreign goods within the mechanics of their social structure, thereby showing that the free use of the term “Romanization” by scholars in the past to explain the presence of Roman goods is in fact wrong (see below and Chapter 4).

One of the Roman-style buildings, located in the modern town of Stupava, has produced evidence to show that it may have had actual Roman inhabitants and maybe even soldiers. The structure is located between the Morava River and the western foothills of the Little Carpathian Mountains, 16 km north of the Danube River, which was the Roman frontier (Fig. 1.2). Archaeological work along the Morava Plain (also called the Záhorie Lowland) has identified German sites (burials and dwellings) around Stupava as well as further to the north and south (Elschek 1995, 39 and 42; Droberjar 1997, 11; Kolník 1998, 144; Turčan 2005). The Roman-style building at Stupava, located in close proximity to other purely German sites, provides an opportunity to study German and Roman interaction from the first to third centuries AD, the lifetime of the Roman-style building (Fig. 3.15). The evidence at the site includes iron shoe-nails characteristic of Roman legionary footwear mixed with other Roman material such as pottery and stamped roof tiles, all of which may indicate a Roman presence at certain levels of the site (Staník and Turčan 2000, 23) (Figs. 3.23–26). Stupava not only provides an opportunity to examine Roman activity beyond the frontier, but also to study the cultural influence that Romans and Germans had on each other. Surveys in and around Stupava have located more German sites (Turčan 1985, 93). Recent excavations have also located German structures just outside the boundary walls of the Roman-style building (Turčan 2005, 293).7

The use of Roman and other foreign material in German context is a common pattern in southwest Slovakia. What we have at Urbárske Sedliská is the mixing of economic zones, incorporating both German local wares and production, while involving the importation and trade of Roman goods. Urbárske Sedliská, being German, provides a chance to study how the different zones of interaction were focused through the filter of Stupava and affected local German development. In conjunction with comparing Germans at Urbárske Sedliská with the “Romans” at the site at Stupava, it becomes important to do comparative studies with German sites beyond the frontier zone to identify differences and similarities, and to do the same with Roman sites at the frontier of the province.

1.2. Goal of the Archaeological Project The aim of the project that forms the core of this book is a greater understanding of how identity functions, as reflected in a culture’s material remains, and what affect outside agents have on identity, if any. To achieve this understanding, I conducted excavations at a site called Urbárske Sedliská, just outside Stupava. Archaeological work at Urbárske Sedliská had already yielded one hut in 1982 and survey had identified promising areas for further huts (see section 1.4 below). Urbárske Sedliská lies around 800 m from the Roman-style building at Stupava and 16 km north of the Danube River (the Roman frontier). Through survey and excavation two more huts were uncovered in 2002 by me, and their material was then studied in comparison to previously excavated huts in the area and in comparison to material found at the Roman-style building. The project thus offers a case study of how the Roman presence, both at the Roman-style building and at the frontier, affected the development of German identity at Urbárske Sedliská. I

In addition to the zones of interaction in southwestern Slovakia, I shall also examine the concept of Romanization, which has changed in recent years (Wells 2005, 51). The traditional concept has been that of acculturation; specifically, that the more advanced Romans brought benefits of civilization to primitive peoples, who then adopted Roman culture and became like Romans or “romanized” (Wells 1999, 126; Hingley 1996, 39; Woolf 1998, 5). This concept is too simple and one-sided. Cultural interaction was more complex, with more than just one group adopting the ideas of another. Most scholars now agree that local peoples were not the only ones who adopted new ideas and material culture, but the Romans themselves adopted concepts from locals as well (Wells 1999, 127). When one attempts to 4

understand Romanization today, there is a need not only to identity what Romans brought to indigenous people but also to look more closely at why locals adopted the things they did. It now seems clear that the indigenous population adopted new ideas because they were advantageous to their own needs and not because they were inherently superior, while rejecting other ideas and even resisting some outright. The local elite, for example, in places such as Gaul or Pannonia may have adopted Roman customs to help them maintain their own status in the eyes of the Romans and also to increase it further in the eyes of the ordinary locals. In Gaul, for example, the elite may have dropped the use of Gallic clothing and adopted the Roman toga because it would help them in dealing with Romans, while also showing to other Gauls that they had access to new prestige goods, further enhancing their identities (Woolf 1998, 12–13; 18–19; 124–125). At the same time, the Romans wanted efficient governance and used local elites when possible; these elites and especially their children were inculcated with the trappings of what it was to be Roman (Woolf 1998, 18, 33).

The primary focus of this project is cultural identity. The ideas and models created by Whittaker, Wells, and Elton primarily deal with material culture and cultural interaction for the frontier, but a theoretical approach to define how cultural identity manifests through material remains is needed. How material evidence bears witness to cultural identity is a difficult and complex question. Many archaeologists have attempted to provide an answer or at least a possible method for determining an answer (see Wells 1980, 1999, 2001; Shennan 1994; GravesBrown et al. 1996; Jones 1997; Laurence and Berry 1998). As the frontier has come to be seen as a mixture of various people with varied backgrounds involved in a complex interaction, so too identity has become viewed as a complex interaction of various elements both within the culture and within the person or group, along with influence from outside the person or group (Ucko 1994, xviii; Shennan 1994, 16; Jones 1997, 85–86; Wells 2001, 24). One part of identity is the manifestation of a boundary that helps differentiate the person or group from other groups. This approach was first identified in the study of ethnicity by Fredrik Barth in his book Ethnic Groups and Boundaries (1998, first published in 1969). Barth’s views are expressly concerned with how a group identifies itself in accordance with the social boundaries it sets up with other groups (Barth 1998, 15). As part of this project, I shall explore how Barth’s views of ethnic boundaries can be applied to the use of material culture by the Germans to create their identity in opposition to the Roman presence along the frontier.

The idea of Romanization is further complicated when applied to the frontier zone. The frontier was not just a mixing of Romans with “barbarians,” but instead a mixing of many different peoples from other regions of the empire. The frontier was garrisoned not only by Roman legionaries, but also by auxiliary forces, some of which might have come from other provinces. At the same time, these auxiliary forces may have been followed by locals from their home area, or previous garrison post, to conduct trade or other business. For example, the small fort at Gerulata, located on the Middle Danube frontier, about 22 km east of the Roman provincial capital of Carnuntum, was garrisoned with German cavalry from Cananefati, near Ryna in Holland, and had people from the eastern part of the empire, such as traders and slaves, who had followed the Cananefati from their previous post in the eastern part of the empire (Kolník et al. 1993, 222). Some of the soldiers worshipped the god Jupiter Dolichenus, a form of Jupiter that had been syncretized with the god Baal from Doliche, in ancient Commagene (Kolník et al. 1993, 224). There was a mixing of ideas and beliefs from all over the empire, and much of it was non-Roman. At the same time, with the frontier being fluid, Germans from beyond the frontier would have been coming to Gerulata and seeing this cultural mix. It is necessary to consider how the Germans reacted to it all and what ideas they took home with them. The Germans were also in contact with non-Roman cultures to the north and northeast, such as the Przeworsk culture in Poland, the Púchov culture in central Slovakia, and the SarmatioDacian cultures of eastern Slovakia and northeastern Hungary. These cultures also would have possibly influenced the Germans of southwest Slovakia. The term “Romanization,” therefore, falls short of labeling the interaction at the frontier, since there was more than just Roman culture at the frontier and in the region as a whole.

When archaeologists examine the material culture of the Germans, even though they may see a wealth of Roman material, they recognize that there are still German items there, and what is more, we know from ancient sources that German identity persisted. Therefore, it is obvious that Germans had other criteria for maintaining “difference.” It is particularly interesting to see how Roman material may have been used to define what is German. If one can dress like a Roman, eat like a Roman, use Roman dishware, but still be German, then there are other criteria in action, such as language, world-outlook, or even just still living in a “German” house. This same issue applies to Britain in the fourth and fifth centuries AD, when the Romans abandoned the province. It appears that Romanization was tenuous in Britain, since by the seventh and eighth centuries it seemed almost as though Romans had never been there (of course taking into account the invasion of other outside groups as well) (see Cleary 1990 and Hines 1996). Again, it can be argued that the Britons maintained a cultural boundary or separateness from Romans, while living, eating, and dressing like Romans. This approach to cultural identity will be explored in Chapter 4 and adapted to the needs of archaeology. In Part I of this book, I outline the ancient German culture along the Middle Danube River using ancient literary evidence and archaeological material. I shall also provide an overview of the latest views on the structure of the Roman frontier, the idea of Romanization, a term which has many problems, especially for the frontier, and 5

finally what is meant by cultural identity. Part II is the account of the archaeological project conducted at Urbárske Sedliská, along with detailed descriptions of specific artifact groups, such as pottery, seen in southwest Slovakia. The final section, Part III, is a synthesis of parts I and II, bringing together the known archaeological data of the region and the theoretical discussions with the new data recovered from the excavations.

whatever cultural antecedent was present (Hingley 1997, 85–86). In this book, when the term Roman is employed it is in fact being used more to describe the material remains from within the Roman Empire as a whole to create a contrast with the German material in southwest Slovakia. The discussion in Chapter 9 of Roman pottery, for example, shows that many of the Pannonian forms have Celtic antecedents that were melded with Roman designs or forms, creating new provincial types, which in the grand scheme of things could be identified as Roman pottery, since they were produced within the boundaries of the Roman Empire by people who may have held Roman citizenship, but were not of pure “Roman” origin. Woolf provides a definition of Roman culture as “the range of objects, beliefs, and practices that were characteristic of people who considered themselves to be, and were widely acknowledged as, Roman” (Woolf 1997, 11). These objects, beliefs, and practices, however, were always changing so that he adds, “becoming Roman was not a matter of acquiring a ready-made cultural package, then, so much as joining the insiders’ debate about what that package did or ought to consist of at that particular time” (Woolf 1997, 11).

1.3. Romans, Celts, Germans, and Archaeological Cultures When one hears the names Roman, Celt, and German or recalls how they are represented in movies, immediate and specific images come to mind (Wells 1999, 19–20). Whether with the Romans being portrayed as despotic, anti-Christian emperors and consummate gladiators, the Celts as druidic tree-huggers sacrificing humans, or the ancient Germans as barbarous men with long beards wearing animal skins, these groups have a specific identity in popular culture. At the same time, in the scholarly community, a similar situation has existed for many decades in terms of how the Romans, Celts, and Germans were viewed and portrayed in various works. The Romans were the torch-bearers for civilization, while the Celts were the ready recipients of this civilization and the Germans were resistors to civilization and the ultimate destroyers of the Western Roman Empire (Wolfram 1988, 5; Jones and Graves-Brown 1996, 10; see Chapter 4.3.2). With the development of identity studies first in anthropology and then in archaeology during the 1970s and 1980s (Jones and Graves-Brown 1996, 5–6; Wells 2001, 13), it became clear that the traditional views of these three groups needed to be changed. There was also a realization that the association of archaeological cultures with specific peoples—thereby creating the impression of monolithic identities that were static—was not valid and in fact counterproductive (Jones and Graves-Brown 1996, 4, 5; see below and Chapter 4.3.2).

The term Celt is also fraught with problems. The Greeks and Romans in their writings talked about Celts as if they were a unified people, who occupied a specific region and had a static identity (Green 1995b, 4; Wells 1999, 32–33). Even after closer contact and the identification of tribes, which were given names in Greek and Roman literature, the over-arching name Celt (or Gaul) was still applied to them. The problem with Celt or Celtic is that there is no indication that the people so named viewed themselves as a unified people to whom the designation Celt could be applied (Green 1995b, 4; Renfrew 1996, 131–132). They may have adopted the terms Celt and Gaul after coming into contact with Greeks and Romans, but before that time, it is difficult to find evidence. Archaeologically, there is a shared material culture during the later Iron Age (primarily after 450 BC and up to the Roman conquests) that has broad similarities from Britain and Gaul in Western Europe to countries in Eastern Europe and even in Turkey as well (Green 1995b, 3–4.7). While material similarities exist, some scholars argue that they do not make everyone Celtic for the European Iron Age (Green 1995b, 3; Fitzpatrick 1996, 238). There have been attempts to link the archaeological evidence with language and ethnic evidence based on modern groups to create a unified Celtic culture, but there has been little success and no wide acceptance of the results (Trigger 1989, 155; Green 1995b, 6; Fitzpatrick 1996, 243; Jones and Graves-Brown 1996, 8; Renfrew 1996, 128–129, 132). Instead of helping to identify an ancient culture, some of these attempts have been more in step with modern political aspirations in Europe to create unifying links between countries (Fitzpatrick 1996, 250– 251; Jones and Graves-Brown 1996, 5, 15; Wells 2001, 20–21). As a result the term Celt has become a politically and nationalistically loaded term. Some scholars see little use for the term Celtic in labeling the archaeological

It has become clear that even though one may use the term “Roman” when describing something within the confines of the Roman Empire, the term itself does not really express the intricacy of the situation (see Chapter 3.2). The Roman Empire covered vast territory and incorporated various peoples at different stages of complexity, with a wide range of material remains. When the term Roman is used, many times it is really a simplification that hides the fact that, in many instances, the people being discussed, while within the Roman Empire, were not really Romans. As Hingley rightly states, there was no “monolithic ‘Roman’ social image; Roman material culture did not necessarily project some form of abstract ‘Roman’ identity” (Hingley 1997, 85; see also Woolf 1998, 7, 11). When someone talks about Roman material, the objects usually have a complicated developmental history, which includes earlier non-Roman antecedents that were melded with Roman types, so that they created a new style, not necessarily “Roman,” but also not purely Celtic, German, Syrian, Egyptian, or 6

groups seen in Europe (Green 1995b, 3; Renfrew 1996, 128, 129, 132–133), but this rejection has also not found broad appeal. In this book, the terms Celt and Celtic Expansion are used to identify a specific material culture and hut types, as defined in Chapters 3.1, 7.1, and 9.4.1, that are found in Slovakia and which differ in comparison to the later German material and huts that appear in the same region.

southwest Slovakia based on the archaeological and textual evidence. For ease of understanding, in this book, I use the term German to refer to the material culture found in southwest Slovakia for the Roman Period, instead of Elbe Group. The term then denotes a specific material culture, hut types, etc., that are different from the earlier material culture identified in Slovakia as Celtic, or from the material appearing in the province of Pannonia within the Roman Empire. I would suggest for the English speaking reader to read German with a hard ‘g’ sound, such as in the word garrulous, to denote the ancient group.

Along with the challenge to the term Celt, there has been a similar call against the term German. There are varied reasons for this challenge and these are discussed briefly in Chapter 4.3. One of the problems that should be mentioned here has been the equation of the Germani of old with the Deutsche of today, for various reasons, but especially nationalistic, by different groups in the 19th and early 20th centuries (Sklenář 1983, 147; Wolfram 1988, 5). The confusion of the two names persists in English, in which Deutsch is translated as German, masking the fact that the two are not the same; this difference is acknowledged in other languages in Europe, such as Allemand in French and Nemec in Slovak when referring to modern Germans. The ancient Germans, much like the later Goths, are a shared people in the history of various modern European nations. By having two different names, it allows access to a shared past, and does not give ammunition to a modern group for territorial ambitions (Sklenář 1983, 146, 151). I mention this since the Nazis are a perfect example of how a group equated the modern Germans with the ancient ones, and so wherever archaeological evidence showed the ancient Germans, it was used by the Nazis to justify their territorial ambitions (Sklenář 1983, 151; Trigger 1989, 163-164, 166).

With the reinterpretation of what is meant by Roman, Celt, and German, I should mention the use of the phrase “archaeological cultures.” In the book there are references to groups, such as the Przeworsk or Púchov cultures. These are archaeologically identified “cultures,” which means an assemblage of artifacts used together in a particular time and place. Archaeological cultures do not immediately correlate with actual groups, whether alive today or mentioned in ancient sources. During and since the nineteenth century, there have been attempts to correlate archaeologically identified cultures with historically attested ones (Sklenář 1983, 91; Trigger 1989, 149, 162; Ucko 1994, xv; Jones and Graves-Brown 1996, 4, 5; Jones 1997, 16, 33; Wells 2001, 20). This attempt can be seen, for example, in the discussion given above for the use of the term Celt (see Jones and GraveBrown 1996). Trying to find links between archaeological cultures and attested groups has fallen out of favor in some archaeological circles, but is still done on a wide scale (such as the linking of the German remains in Slovakia with the Quadi and my mention of trying to link the Púchov culture with the Cotini). Chapter 4.3 provides a discussion of the past use of the term culture, which was seen as a monolithic, static entity, and the reassessment, whereby culture is now seen as dynamic and constantly changing.

The material culture that is attributed to the late Iron Age/Roman Period groups that inhabited southwest Slovakia has broad similarities to the material found along the Elbe River in Germany, thereby providing a link which has been interpreted as evidence for either the movement of peoples or the movement of goods. Most scholars, using the ancient Roman sources, see the appearance of material culture from the Elbe region in the Czech and Slovak Republics as evidence for the movement of the Marcomanni and Quadi into the region (see Chapter 3). The identification of a movement of people instead of merely material culture is corroborated for some archaeologists through the abandonment of earlier sites in Slovakia identified as Celtic or CeltoDacian and the appearance of new settlements in new locations with “German” cultural material, while some sites show continuity with a switch to new material culture and a disappearance of the older (see Chapter 3). The movement of tribes in antiquity is a controversial topic today and the certainty shown in the historical/archaeological writings about this subject in the late 19th and early to mid-20th centuries has been rejected (see Wolfram 1988, 5-11). Much as with “Celts,” there was no pan-German identity and the groups identified as German by the Romans actually identified themselves by their tribal name (see Chapter 4.3). In Chapter 4.3, I attempt to reconstruct ancient German culture in

1.4. Why Slovakia? There are two reasons why Slovakia was chosen for the archaeological project. The first reason has already been stated in that Slovakia provides a superior opportunity to study interaction between two cultures since it straddled the frontier of the Roman Empire. The second reason is that my knowledge of the Slovak language provided me access to the scholarly literature from the country, thereby enabling Slovak archaeological work to be introduced to an English speaking audience through this work. I was able to volunteer on excavations of the Roman-style building at Stupava with the help of Dr. Štephan Holčík, former director of the Archaeological Museum in Bratislava, and spent two seasons (1999 and 2000) working with other students under the direction of Dr. Vladimír Turčan,8 who is in charge of all archaeological studies in the Stupava area and directly supervises the excavations at the Roman-style structure. After talking with Dr. Turčan about my interests and 7

focus of study, he offered me for further archaeological examination two potential sites, Urbárske Sedliská and Mást, known to have German material and located near the Roman-style structure (Fig. 1.6). Dr. Turčan was willing to allow this study in an effort to expand the knowledge of what was happening in Stupava during the Roman Period. Archaeological research in Stupava has focused on the Roman-style structure because of the shortage in funding and manpower, thereby resulting in a limited knowledge of what was happening among the Germans, immediately around the site.9

of pottery was found along with some metal objects (Figs. 1.10–13). The pottery was primarily German, which were of types common in Southwest Slovakia in the first half of the second century AD (Turčan 1985, 93 and 107). Among the provincial13 pottery were jugs and plates (Fig. 1.11). One of the jugs was made of brick red fabric, but not enough survived to identify the shape to fit it precisely into the chronology of Pannonia; it could be dated generally in the second to the fourth centuries AD (Turčan 1985, 110). The dating of this pottery places the site approximately in the same time period as the occupation of the Roman-style structure and therefore provides a chance to see German life in the shadow of the more complex site located nearby.

In late July of 2002, the archaeological project was begun at the two sites of Urbárske Sedliská and Mást. The project was titled the Urbárske/Mást Archaeological Project and was sponsored by the Archaeology Museum, Slovak National Museum (herein AM-SNM) in Bratislava, Slovakia, and the Department of Archaeology at Boston University, with funding from the Humanities Foundation at Boston University and private donations. The survey and excavation were conducted from July into mid-October and artifact conservation and study have continued to the present.10 The museum also provided research and laboratory facilities along with conservation and storage space for the excavated artifacts. In accordance with Slovak law, Dr. Turčan, being in charge of all archaeological work in Stupava, was the primary investigator and I was partnered with him.11 Dr. Turčan was interested in seeing archaeological techniques that are in use outside of Slovakia and accorded me freedom to conduct the excavations as I saw fit, with Dr. Turčan in a supervisory role to provide logistical support and to ensure the laws were being followed.

The most surprising discovery at Urbárske Sedliská was a series of small crucibles,14 only one of which was whole (total of seven; Fig. 1.12: numbers 9–13, 16, and 17; no. 11 is the one intact). The crucibles are rounded with pinching of the walls to create depressions, and are of a type known to have been used in the working of precious metals for jewelry during the La Tène period (Turčan 1985, 110). They have been found both in oppida and also in smaller settlements. This is the first instance of their discovery in a German site in Slovakia, although examples have been found at other German sites outside Slovakia (Turčan 1985, 110). Samples from these crucibles were submitted to a local examination, being sent to a lab; minute traces of metal were found on the surface of the sherds. Five samples of the metals (from five of the seven crucibles) were subjected to a metalographic analysis and microbeam probe to identify the metal (Turčan 1985, 110). One crucible (no. 11) was found to have had pure silver, while another had pure gold (no. 17), and three had copper mixed with tin and zinc (nos. 12, 13, and 16) (Turčan 1985, 112) (Fig. 1.12.). Along with the crucibles were some bronze and iron tools that were also used in the processing of the metals and jewelry making (Turčan 1985, 112) (Fig. 1.13). In the fill of the structure some slag was found as well (Turčan 1985, 113).

1.5. Background Information for Urbárske Sedliská and Mást From the project’s inception, Urbárske Sedliská was the primary focus since it had already proven to contain at least one German structure with the potential of more being found. Urbárske Sedliská is located approximately 800 m to the northwest of the hilltop at Stupava where the Roman-style building sits (Fig. 1.6). Work was first done at the site of Urbárske Sedliská in 1982, when farmers were placing drainage pipes in the fields in an attempt to dry out the land to improve crop growth (Turčan 1985, 94). Dr. Vladimir Turčan and Dr. Zdenek Farkaš (also of AM-SNM) were observers of the digging and noted if any ancient material was uncovered (Turčan 1985, 94; Farkaš 1984, 2). Two areas yielded ancient German material along with other cultures: Area A on the west side of a drainage ditch, approximately 10 m away from the ditch itself; Area B was 150 m east of the drainage ditch (Fig. 1.8).12

This Grubenhaus came from only a portion of the site and pointed to the possibility that expanded excavation of the site may result in discoveries as important. Dr. Turčan believed that the Germans working precious metals at Urbárske Sedliská obtained them and the inspiration to work them from nearby Stupava (Turčan 1985, 114). Turčan’s research had indicated that some Germans in Stupava were dealing with precious metals, but a question remained regarding for whom was the work being done: was it for German elites or for the possible Romans in Stupava? If this is the first evidence of jewelry-making at a German site in Slovakia, does it mean that the Germans in Stupava were developing differently than those at other sites? Were the Germans in Stupava materially better off? Further investigation would provide a better understanding of what had been uncovered so far, while also providing information that would help clarify our understanding of German settlement in Stupava and how it may have differed from other German sites.

In Area A, workers uncovered German material, and when the place was further investigated in 1982 a Grubenhaus was discovered. Dr. Turčan excavated the German hut, which was of trapezoidal shape, the longitudinal axis of which was oriented northeast to southwest (Turčan 1985, 93) (Fig. 1.9). A large amount 8

In Area B, a Bronze Age urn-burial was uncovered by Farkaš.15 Area B was a low hill with gentle slopes leading to the manmade drainage ditch, 150 m to the west (Fig. 1.14). It was noted by Dr. Farkaš that along the slopes and at the western edge of the hilltop, to the south of the BA burial, there were various types of pottery sherds, including German and Roman pieces (Farkaš 1984, 4). He postulated that if a site did exist, it would be on top of the low hill (Farkaš 1984, 3 and 4).16 Dr. Farkaš had started lower down the hill, extending trench I/82 south for 7.2 meters, then started a second trench, II/82, perpendicular at the south end of trench one, heading east, creating an “L” shape (Farkaš 1984, 2)(Fig. 1.14). By heading east, trench II/82 went six meters up the hill. In trench I/82, at a depth of 80 cm a large piece of a German pot was uncovered, decorated with half-circles made by a comb (Farkaš 1984, 3)(see figure 8.37: Type XIIIf for examples of that decoration motif). The sherd was found in dark earth with a large amount of other sherds and animal bones.17 The German sherd found in I/82 could be dated to the second half of the second century AD, along with some provincial sherds found with the German piece (Farkaš 1984, 4). In trench II/82, more sherds were found between 2.2 and 3.4 meters along the trench, at an approximate depth of 100 cm. The sherds were a mix of periods, from the Neolithic to the Roman, along with a flint blade (belonging to the Lengyel culture) and a piece of obsidian (Farkaš 1984, 4). Because the material was located on a slope, with all periods jumbled together, Dr. Farkaš postulated that slope wash had brought the artifacts to their present location and that cultural features would be on top of the hill (Farkaš 1984, 3 and 4).

1.6. Archaeological Methods The timetable for the project at Urbárske Sedliská called for a three-week survey period followed by eight weeks of excavation and ten weeks of lab work. The survey and fieldwork were divided into three phases, starting in July with phase one and ending by mid-October 2002 with phase three. The first phase was a magnetometer survey. The probable existence of material remains such as pottery, fired wattle-and-daub, along with hearths and potentially more slag provided a good chance of detection by a magnetometer. It was postulated that these cultural features and artifacts would show up as anomalies on the resulting map and would help in the decision process for choosing where to excavate in phase three of the project. Chapter 5.2 deals with this first phase and also provides some background information on magnetometer surveys in general. For the second phase of the project, a surface survey was planned, whereby a crew would be spread ten meters apart, walking transects while looking for any visible cultural material. The material seen would be marked as to location and type, creating a distribution map of surface finds based on artifact density. Flexibility was built into the plan by having equipment to do shovel test pits if necessary. The two survey maps from phases one and two would then be combined and based on magnetometer anomalies and artifact densities, the best areas for excavation would be chosen. Once in the field, changes had to be made to the project design. The magnetometer survey was done on July 25th, but the results were not provided until late September, too late to help in identifying where to place the excavation units, the task of which had to be completed by late August. Phase three then had to rely on the successful identification of artifact densities through surface survey. The surface survey had originally been designed as a pedestrian survey, but because of a dense overburden of plants and root mat, this method had to be abandoned in favor of shovel test pits. The shovel test pits proved successful with two pits having very high concentrations of German material. Chapter 5.3 provides a description of the survey methodology used for phase two of the fieldwork.

My 2002 project was organized to explore Area B in the belief that the German pottery may be indicative of some type of structure or at least cultural feature (Fig. 1.12). Area A with the Grubenhaus was not chosen because no other ancient material was found during the excavation; what is more, the land there, near the drainage channel, is wet and choked with vegetation, which would have slowed archaeological work. Mást is located approximately 2 km to the southwest of the Roman-style structure and was discovered in 1995, when a pipe was being laid across a series of six private garden plots in the area of Mást, a small sub-division of Stupava located on the western side of the town, closer to the highway (Fig. 1.10).18 Another surface survey and a magnetometer survey were planned and the goal was to create a better understanding of the site to see if future excavation should be conducted. There was hope at the time that Mást would prove to be a German settlement, thereby providing further potential evidence to elucidate Stupava’s impact on the region. From the outset there were problems with the site, such as re-locating it and obtaining permits, which eventually resulted in the abandonment of work there for the archaeological project.19

The third phase was the actual excavation of the promising areas within the surveyed area. The excavation portion of the project began with units placed near the two shovel test pits that had yielded the most promising data. The excavation of the units followed natural stratigraphy. If a unit provided sufficient evidence of cultural activity that unit was expanded in an effort to reveal as much as possible in regards to the potential discovery. Because German sites usually do not have great time depth (see Chapter 3.4), vertical control was not as critical as obtaining an understanding of the horizontal picture. Excavation used shovel and pick for the plough-zone and then trowels for potential cultural 9

strata. All identified cultural features were troweled and one hundred percent artifact recovery was attempted. All earth removed from cultural features was screened through a ¼-inch mesh for the recovery of small finds. Samples were collected for radiocarbon dating, along with samples through the use of flotation for paleoethnobotanical study, and soil was collected from the cultural features for geoarchaeological/ micromorphological analysis. Chapter Seven provides a detailed account of the third phase of the project.

Hungarian territory east and southeast of the Danube. The common definition of the barbaricum is “a foreign land.” The word is not a new invention and was in fact used by the Romans themselves to refer to the lands beyond the Danube as a whole. Suetonius used it in reference to the Germans (Suet. Calg. 47), and barbaricum also appears in late antique works such as the SHA, the Epitome by Eutropius, and the Summary of Roman History by Festus (c. fourth century AD). The term has received wide support and is frequently used in the scholarly community of Central Europe, since it is an efficient alternative to naming all the tribes or four modern countries in scholarly discourse. The term can be applied to all lands outside the Roman Empire, but also in regards to particular areas (Elton 1996: 36–37). The term, of course, is derived from the word barbarian, coming from the Greek βάρβαρος and βάρβαροι, meaning anyone not Greek and later Roman, especially those who had no accomplishments that were comparable to classical civilization.

1.7. Conclusion Southwest Slovakia was a vibrant frontier region with constant mixing of cultures and goods, but where identity was maintained as seen in the strong German presence in the form of the Quadi tribe. The power of the Quadi to maintain their identity is evident in the passages of various ancient authors and in the archaeological record. The study of Urbárske Sedliská helps clarify German development in southwest Slovakia under the shadow of the Roman presence both along the frontier and at Stupava, while also helping to show what type of interaction took place between Romans and Germans beyond the empire. In the following pages, I shall use the evidence from other sites in southwest Slovakia and that of Urbárske Sedliská to demonstrate that the ancient Germans used their own material culture along with that of the Roman Empire to reinforce and maintain their particular sense of identity.

5. These occurrences will be examined in greater detail in Chapter 2 and Appendix B, which deals with the ancient sources. 6. Slovak archaeologists refer to these structures individually as either a Rímska stavba (“Roman structure”) or a Rímska stanica (“Roman station”), since they are so out of place amidst German habitation, which was primarily composed of semi-submerged huts made of waddle-and-daub with thatch roofs, termed Grubenhäuser in German (see Chapter 7). In an effort to avoid confusion and to take into account that there is controversy about these structures in the literature, I will refer to them as “Roman-style” structures or buildings, just as C. R. Whittaker referred to them in his book Frontiers of the Roman Empire, where he calls them Romanized buildings, Roman-style buildings, and Roman-looking buildings (Whittaker 1994, 115–117). Hugh Elton also uses the term Roman-style to refer to these sites in his book Frontiers of the Roman Empire (Elton 1996, 106). The term then refers to the architectural style of the building, which uses Roman design and material, in clear contrast to the traditional German style of house construction.

Endnotes: 1. All three of these terms have caused great debate as to their meaning and how they apply to various peoples and places. They are presented here as part of the introduction and will be explored in greater detail in the following chapters. 2. There are a few books and articles dealing with Slovakia published in English (see Neustupný 1961, Kraskovská 1978 and 1981, Chropovský et al. 1981, Turčan 1991, Krekovič 1994, 1997a, 1997b, 2000b, 2001a and 2001b, and Kuzma et al. 1996). There are two articles written by the British scholar L. F. Pitts (1987 and 1989) dealing with the Roman Period sites in Slovakia, the Czech Republic and eastern Austria. Various books and articles published in Slovakia may have summaries usually in German or Russian, but sometimes in English.

8. Vladimír Turčan is also a member of the Archaeology Museum in Bratislava, which is a branch of the Slovak National Museum system.

3. A more detailed description of the Quadi and other Roman Period cultures will be given in Chapters 2 and 3.

9. Turčan in 1985 had already made statements to this effect (Turčan 1985, 93).

4. The term barbaricum in this work refers to the land occupied by the Germanic Suebi, Marcomanni, and Quadi, as well as lands of the Sarmatian Iazyges. Today these lands encompass parts of Germany, the Czech Republic, Slovakia, the northeast portion of Austria, and

10. See Vrba 2004, 129.

7. Personal communication from V. Turčan.

11. In Slovakia, foreign archaeologists are not allowed to work without a Slovak archaeologist as partner. Turčan

10

obtained the official permits and permission from various landowners for the project. 12. Three other sites were identified, having material from the Lengyel period and Bronze Age. One site was approximately 500 m to the northeast of Area B, while the other to sites were directly north of Area B, approximately 1 and 1.5 km distant, in an area marked as Moligrunty pod hradskou on the maps. 13. The term “provincial” or later on the phrase “provincial pottery” is used in this book to refer to ceramics made in the Roman province of Pannonia, using shapes and clay of the area. The phrase “Pannonian Ware” can also be used to refer to some types of pottery forms found in the region. See Chapter 9 for more information on Pannonian Wares. 14. Crucibles are containers made to hold metals for melting and fusing or for calcining ores, metals, etc. 15. This material has not been officially published and the account here is taken from the field records on file at AM-SNM. Z. Farkaš has given me permission to describe the German material, but not the Bronze Age burial, which he expects to publish in the near future. 16. See Vrba 2004, 131. 17. Because of this description and from my experience with the two Grubenhäuser I excavated, it would seem that Z. Farkaš had excavated a corner of a one. A Grubenhaus is usually filled with a dark earth (usually charcoal flecked) that stands out from the surrounding soil/sediments (such as sand), and this dark soil is full of pottery, bone, and other material. 18. V. Turčan and J. Bartík (also of AM-SNM) watched the construction and walked the area being excavated to see if any material was being turned up during digging. The survey yielded 29 items made up of 23 sherds of pottery: 14 German, 6 provincial, 1 terra sigillata, and 2 unknown. Five metal fragments were collected: 1 bronze appliqué, 1 arrow head, 1 bronze slag, 2 pieces of thin material, either bronze or copper, and a possible piece of a Celtic torque. There was also one piece of baked earth from a wattle-and-daub structure. The area surveyed was roughly 200 x 50 meters and most of the material came from the southern portion, while the northern portion became wetter as it neared the Mláka stream and a line of trees (Fig. 1.11). The material found may have only been an agglomeration of material resulting from farming activities, and the site seemed less promising than Urbárske Sedliská based on the description and material recovered. The material from Mast was shown to me by Turčan during the Stupava field season of 2000. He has given me permission to discuss the site and its associated material, even though nothing has yet been published.

11

Figures

Figure 1.1. Modern-day Slovakia (based on a map from Kirschbaum 1995, xii).

Figure 1.2. Roman provinces along the Middle Danube River across from Slovakia, showing the position of the Quadi and Iazyges as well as the location of Roman-style structures beyond the frontier (based on a map from Genser 1990, 50:Abb. 22). Trenčín has a Roman inscription identifying the site as Laugaricio. The town after which the site of Milanovce was named is now known as Veľký Kýr.

12

Figure 1.3. Aerial view of Cífer-Pác. The Roman-style structure is the large one at the bottom of the excavated area (modified image from Kolník 2000, 41).

Figure 1.4. View of the bath-building at Dúbravka during excavations (Elschek 2000, 27)

13

Figure 1.5. Archaeological plan of the excavations at Milanovce (Kolník 1959, 33: Abb. 4).

Figure 1.6. Map of Stupava (shaded in gray) showing location of the Roman-style structure (letter A), the site at Urbárske Sedliská (letter B) and the site at Mást (letter C).

14

Figure 1.7. View of area surveyed by Dr. Turčan and Dr. Bartík in 1995 when pipes were being laid.

Figure 1.8. Areas surveyed by Dr. Turčan and Dr. Farkaš in 1982. 15

Figure 1.9. Grubenhaus excavated by Dr. Turčan in Area A (modified image from Turčan 1985, 95: Obr. 2).

Figure 1.10. A sample of German pottery found in the Grubenhaus (modified image from Turčan 1985, 97: Obr. 3).

16

Figure 1.11. A sample of the Roman provincial pottery found in the Grubenhaus (Turčan 1985, 106: Obr. 10).

17

Figure 1.12. German pottery and crucibles from the Grubenhaus (modified image from Turčan 1985, 103: Obr. 8).

18

Figure 1.13. Metal artifacts from the Grubenhaus (Turčan 1985, 107: Obr. 11)

Figure 1.14. Map of Urbárske Sedliská showing the location and shape of excavation trenches dug by Dr. Farkaš in 1982. 19

Because there was pottery found in the German features that were identified as Late or Final La Tène (see Chapter 9), information is provided for these periods in this chapter, section 2.3.1, Chapter 3, section 3.1, and in Chapter 9. The La Tène is divided into five broad periods of Pre, Early, Middle, Late and Final, with further subperiods: A (450–400 BC), B (400–250 BC), C (250–150 BC) and D (150–10 BC). The sub-periods are usually referred to as LTA–D, with each further subdivided and identified by subscript numbers, such as LTA1 or LTC2 (see Table 2.1 and Fig. 2.1).3

Chapter 2 Historical Account of Southwest Slovakia: 100 BC–AD 166 Chapter 2 is divided into four main sections: 1) the first part describes the geographic and chronological framework for Slovakia with emphasis on the Final La Tène and Roman Periods; 2) an overview of the ancient sources that mention the region of southwest Slovakia during the periods discussed in this chapter; 3) an historical account of southwest Slovakia from 100 BC to AD 166, using the ancient sources; and 4) summary of the chapter. There is no recent historical account in English of the Middle Danube region using the ancient sources with a focus on the Czech Republic and Slovakia.1 Section 2.3 of this chapter, therefore, provides detailed coverage of the historical account.

For the Roman Period chronology, there are two broad divisions of Early (AD 10–166) and Late (AD 180–375) and a break between the two identified with the Marcommanic Wars, covering the B2/C1 period, AD 166– 180 (Table 2.1 and Fig. 2.2). Both the Early and Late Roman Periods are further subdivided, with the Early Period identified as B1a–c (AD 10–70/80) and B2a–c (AD 70/80–166) (Table 2.1). The Late Period is divided into C1a–b (AD 180–250/60), C2 (AD 250/60–310) and C3 (AD 310–375) (Table 2.1). There is a difference between the chronologies used in Slovakia and Moravia (see Table 2.1; Droberjar 2002, 103)

2.1. The Geographic and Chronological Framework for Slovakia from Final La Tène to the Great Migration Period

The period in which Germanic tribes moved into the Roman Empire and Pannonia as a province disappeared, is known as the Great Migration Period (AD 375–568) (Fig. 2.2). This period is divided into D1 (AD 375–425), D2 (AD 425–475) and D3 (AD 475–568). It is during the late D2 or early D3 that the first Slavs came into the territory of Slovakia in all three regions (Steinhübel 2000, 37; Ruttkay 2000, 69).

When discussing archaeological cultures for Slovakia, there are three distinct zones that follow basic geographic divisions: west or southwest Slovakia, central Slovakia, and eastern Slovakia (Fig. 2.5). Depending on the period, various cultural groups existed in each region or occupied multiple regions. It should be noted that Central Slovakia is divided between north and south because these regions differed archaeologically, with the southern portion sharing a close development with southwestern Slovakia, while the northern section shares more parallels with eastern Slovakia or southern Poland.2 From the Paleolithic to the arrival of the Slavs, western/southwestern Slovakia was primarily shaped by groups seen also in the Czech Republic, Austria and northwestern Hungary; central Slovakia showed influences from Poland, north-central and northeastern Hungary, and Romania; eastern Slovakia was primarily influenced by, or was part of, cultural groups seen in northeastern Hungary, southeastern Poland, the Ukraine, and Romania. Many times, Central Slovakia served as the transitional zone between western and eastern Slovakia.

The chronological system outlined in this section is a modified one first introduced by H. Eggers in 1951 and 1955 (Kolník 1971, 511; Droberjar 2002, 68 and 103). His system had a Period A for the end of the La Tène (50–1 BC), Periods B1 (AD 0–50) and B2 (AD 50–150) for the Early Roman Period and C1 (AD 150–200), C2 (AD 200–300), and C3 (AD 300–350) for the Late Roman Period (Droberjar 2002, 103). In Slovakia, Egger’s period A has become LTD3, so that there is no official Period A, but instead a direct start with B1a; for all the barbaricum an intermediate period B2/C1 was added for the Marcomannic War, and the Late Roman Period was moved further in the chronology to start at AD 180. Other chronologies have been proposed and used both before Eggers and after him, but Egger’s has gained wide acceptance. The chronology, with modifications, is used for the barbaricum not only in Slovakia, but also Poland and the Czech Republic (Dorberjar 2002, 103).

Since this book is focused on the Quadi and German interaction along the middle Danube frontier, the following discussion of chronology is oriented around southwest Slovakia, with reference to the other two regions. The periods and sub-periods along the timeline for Slovakia are shared by all three regions. Appendix A provides two charts with chronological information for all three regions, from the Late Hallstatt to the Great Migration Period, showing the various cultural groups of each region and where they fit in time.

2.2. Review of the Ancient Sources for Southwest Slovakia The following two sections provide a brief discussion of the available written sources, listing the most important authors and what they cover, while also mentioning their shortcomings.4 The authors are divided by the two 20

periods covered in Chapter Two: 1) The Final La Tène, Roman expansion to the Danube, Celtic withdrawal, and the arrival of the Germans in southwest Slovakia, 100–10 BC and 2) Roman and German interaction along the Middle Danube frontier, marked by a mixture of diplomacy and war, AD 10–166. Some of the authors overlap periods and may be mentioned more than once below.

criticized and he has been considered an ‘inept, if enthusiastic, amateur’ who tried to imitate Thucydides or Tacitus, but failed (Bucher 2000, 411-412). This view, however, has now been reappraised and one can say that Appian’s history was not really a history, but instead a way to explain Rome’s rise to dominance and the benefits of monarchy to the world (Bucher 2000, 429, 438-439, 442). Appian has been shown to have written objectively, in that while being an advocate for monarchy, he does so on a broad scale, thereby not being concerned with creating heroes or monsters, enabling him to give accounts of various figures, such as Octavian, in a not so favorable way (Bucher 2000, 440, 442). Appian’s work is important for the Middle Danube region for his account of the Illyrian Wars, where Pannonia and the Danube are discussed, and Roman actions in the region during the period of the Civil Wars.

Even though a wide range of written sources have survived from ancient times, no sources deal solely with the Middle Danube. There are no accounts written from the German perspective for the period being discussed in this chapter. There are, however, numerous works by Greek and Roman authors that mention southwest Slovakia, but they must be used with caution. Many authors had biases, hidden agendas, lack of accurate information of the region, or were sloppy in scholarship, relying on unreliable accounts. Peter Wells makes a good point that the ancient texts are mostly descriptive, providing information on who did what with little interpretation (Wells 1999, 66). Other important sources of written information include inscriptions from monuments and burial steles along and beyond the Danube frontier or even from other parts of the empire, as well as coin inscriptions and legionary stamps on roof tiles and bricks (see Chapter 3 for mention of these materials).

Pliny the Elder mentions in his Naturalis Historia the movement of the Boii, who had occupied much of the Czech Republic and had expanded into Slovakia. He also provides an account of a Roman soldier traveling beyond the Danube along the Amber Route to obtain a large supply of amber. The Natural History, written in the mid1st century AD, is a colossal work and contains much information, but care needs to be given because Pliny used sources indiscriminately, usually with no critical eye toward accuracy. He also wrote Bella Germaniae in 20 books, covering in detail Roman campaigns against the Germans. The work does not survive, but was used extensively by Tacitus in his Annales and Germania (Benario 1999, 3-4).

2.2.1. Authors for the Final La Tène Period: 100-10 BC

Strabo’s Geography provides a description of the region of the Middle Danube. The work was written between 5 and 9 BC with revisions in AD 18 (Clarke 1997, 102103).6 Strabo prides himself on his extensive travels, but does not list the northern frontier of the empire as one of the places visited. He does mention that in the period he writes there was new information available to provide better accounts of the Britons, Germans, Caucuses, etc. (Strabo 2.5.12). Strabo is important because he provides a detailed description of both Celtic events in the middle Danube region before the arrival of the Germans and also a list of groups living in the region after the Germans arrive.

For the first period, covering the Middle Danube up to 10 BC, there are fewer sources than for the other period. During this time, the Middle Danube region was still outside the realm of Roman interests and so little was known about the area. Greeks had also limited information on this region. The earliest author to mention the middle Danube region is Herodutus, but the primary authors available for 50-10 BC are Appian, Pliny the Elder, Strabo, Cassius Dio and Augustus’ Res Divi Augusti, with some information provided by Julius Caeser, Velleius Paterculus, and Suetonius.5 Appian was an Alexandrian Greek born at the end of the 1st century AD and wrote the Roman History sometime between AD 150 and 166, arraigned ethnographically, detailing each people that Rome conquered, starting with Italians and ending his work with an account of Trajan’s campaigns against the Dacians, Jews, Pontic peoples and Arabians (Bucher 2000, 416). Of the 24 books, 1-9 are fragmentary, 10 and 18-24 are lost, thereby only leaving us complete books 11-17, which are Appian’s account of Rome’s wars against the Iberians, Hannibal, Carthaginians, Macedonians and Illyrians, Syrians and Parthians, Mithradates VI, and the Civil Wars. Appian’s account of the Civil Wars is the only surviving connected narrative, covering the period from 133-35 BC and used many works that have not survived; thereby making Appian one of the most important authors for the period (Bucher 2000, 412). Appian’s abilities have been heavily

Of the four primary sources mentioned, Dio Cassius is the most removed in time from the first period under discussion here. Dio lived in the late third century and wrote the Roman History between AD 207 and 219, with some additions in 229 (Millar 1964, 30). The work starts with the arrival of Aeneas in Italy and ends in AD 229 (Cary 1914, x; Millar 1964, 38). The work does not survive intact; the first portion from the founding of Rome to 69 BC is missing, the period 69 BC to AD 46 has survived with some gaps, but the remainder only exists in lengthy excerpts and epitomes by the later Byzantine writers Zonaras and Xiphilinus (Millar 1964, 1-2). As stated by Dio himself, he read through many sources first before writing, and because of this it is thought he then relied on his notes since there are many 21

errors and at times his accounts are too brief or shallow in detail (Rich 1996, 299-300). There are also some omissions and some periods have events combined so that a clear picture is hard to obtain. Many of these omissions and conflations are thought to have been done by Dio for effect of style, which he felt was more important, so that the narrative would flow and not be burdened by details (Millar 1964, 43-44). Dio is important for this period because of his account of the Augustan campaigns in Illyria, Dalmatia and Pannonia.7

verified many details provided in the Germania (Benario 1975, 157; Martin 1981, 57; Mellor 1993, 44-45). For the study of the frontier and ancient Germans, Tacitus’ Germania is by far the most important ancient work (Todd 1975, 28; 2004, 6). The Germania was written in AD 98, contemporary with his Agricola, but published second (Martin 1981, 49). It has been argued that Tacitus attempted not only to provide an ethnographic essay about the Germans, following traditional structure and topics of the genre, but also to provide a contrast between the virtues, morals and politics of Romans and Germans (Martin 1981, 49-50; Benario 1999, 3). This contrast was attempted to warn the Roman readers about the moral decline of the empire and a need to go back to the moral codes of the ancestors (Martin 1981, 49). Tactius, however, also lists the shortcomings and weaknesses of the Germans, along with strengths, providing detailed lists of various tribes, belief structures and lifestyles. Because of these details, his work has also been seen as a way for Tacitus to warn Roman readers of the continuing German threat on the frontiers (Martin 1981, 50; Benario 1999, 3). While he may have been trying to use the Germans for various motives, there is no evidence that Tacitus actually visited the German provinces or the frontier (Benario 1999, 3).10 Tacitus relied on earlier works, such as Pliny the Elder’s history of the German wars, along with accounts by traders and soldiers (Benario 1975, 31; 1999, 3-4).

The last work listed as an important source for this period and the early portion of the next one is Augustus’ biography, Res Gestae Divi Augusti, which survived from an inscription on a temple of Rome and Augustus at Ancyra in Asia Minor (Jones 1970, 168).8 The document is a summary of his acts as emperor, listing many of his accomplishments. The text itself was not written in AD 14, but first compiled in 2 BC and added to or amended afterward (Shipley 1924, 338). Being somewhat of an autobiography by a master propagandist, the details provided in the Res Gestae are important, but the motives for what is told need to be kept in mind (Bosworth 1999, 1). While the document may provide many of Augustus’ acts, some were not listed and others were presented a certain way in order to cast the best light possible on the emperor (Jones 1970, 168-169). The work is important because it provides information on Augustus’ accomplishments against the Germans and also his actions in Pannonia.

Once again, Dio Cassius’ Roman History provides important details about the activities of various emperors along the frontier, especially Domitian, Trajan and Hadrian; he is important also for his account of the Marcomannic Wars, AD 166-180. The drawback is that these sections of Dio only survive as epitomes, as mentioned above.

2.2.2. Authors for the Early Roman Period: AD 10166 For the second period, AD 10 – 166, there are more works available. The most important ones are by Tacitus, Dio Cassius, Vellieus Paterculus, and the Scriptores Historiae Augustae (herein SHA), with supporting information from Amminaus Marcellinus, Appian, Aurelius Victor, Eutropius, Florus, Martial, Silius Italicus, Statius, and Suetonius.9 Another significant work dating to this period is the Geography by Ptolemy Claudius.

Velleius Paterculus wrote what has come to be called the Historia Romana, which was actually a universal history written in a brief style that Velleius himself called transcursus (Starr 1981, 164 and 166).11 The work was finished sometime in the latter half of AD 30 (Sumner 1970, 28). Velleius has not had a favorable standing among historians, usually referred to as a “bluff old soldier, ham-fisted amateur, and sleazy toady” (Sumner 1970, 257 and 281).12 There has been a reappraisal, however, of both Velleius and his history, showing that his military experience and service under Gaius Caesar and Tiberius provided him with hands-on and eyewitness experience to many events he relates in the last portion of his work (Sumner 1970, 270 and 279). Also, with the realization that his work is a universal history, the accusation that Velleius was an apologist and propagandist for Tiberius is weakened and the tone of the work can be recognized for what it was––a “miniature universal history” as R. Starr called it (Starr 1981, 174).13 Velleius is important because he provides an account of Tiberius’ wars against the Germans and activities in Pannonia as general under Augustus. The author was also a participant, having served as praefectus equitum under

Of these authors, Cornelius Tacitus is the most important. His Historiarum and Annales provide important details of various events along the Middle Danube, such as the rise and fall of Maroboduus and Vannius, while his ethnographic work, Germania, provides unparalleled information about German culture and organization beyond the Roman Empire. Tacitus, however, did not write objective accounts of historical events, but instead attempted to criticize Roman morals and political values, trying to be a ‘moral historian’ (Martin 1981, 10; Mellor 1993, 2, 15, 30 and 48). He disliked monarchy and his works show an interest in the concept of the principate and the slide toward autocratic rule (Benario 1975, 147 and 148; Gowing 1990, 316). In using Tacitus, one must be aware of his prejudices and agenda, but at the same time, he has been found to be fairly accurate, not changing facts to fit his concepts, and archaeology has 22

Tiberius for the campaign on the Rhine and legatus Augusti in Pannonia (Sumner 1970, 268 and 272). The SHA is a late Roman document and problematic work, but still provides some useful information. The work is supposedly a compilation of imperial biographies composed by six writers sometime between the reigns of Diocletian and Constantine. This belief was generally accepted as true until the very late 19th century, when H. Dessau put forward the idea that the SHA was in fact written by one person some time in the late 4th century AD (Syme 1968, 1). Dessau’s theory at first found little support (Syme 1968, 1 and 217), but over time it received more and more adherents, and with P. White’s (1967) and R. Syme’s (1968) further research, it can be said that there is now wide support for the belief that there was only one author (Birley 1987, 229). The exact date of the SHA is still debated, with a general consensus of the late fourth century,14 but there are dates proposed of earlier in the fourth as well as later into the fifth and even sixth centuries (Honoré 1987, 156).

how to make maps, the remainder of the work is Ptolemy’s attempt to create a world map using the system he devised and also to provide tables of locations and coordinates so that anyone can reconstruct the map at a later date (Riley 1995, 231; Berggren and Jones 2000, 4). Ptolemy is important because of his description of the Middle Danube region, including the territory of the Qaudi and beyond. He lists a number towns and important places that were beyond the frontier and among the Germans. The problem is trying to actually locate where they were in reality, because Ptolemy’s system is not accurate, especially as you move further from Italy and the maps that have survived with the manuscripts are Medieval in date, thereby having their own host of problems (Riley 1995, 232; Marsina, ed., 1998, 115; Berggren and Jones 2000, 5-6). For the Middle Danube, only two sites, Celemantia and Leukaristos, have been identified, and the other places have not (Valachovič 1998, 24).16 A document that helps us reconstruct how the Pannonian province and frontier were organized is the Antonine Itinerary. The itinerary dates to the late third century and provides information on the Roman road network, listing the towns, settlements, forts and stations, also providing distances (Purcell 1996a, 775; Valachovič 1998, 24).

The biographies are filled with falsities, lies, and misconceptions, with a focus on the bizarre and titillating (Syme 1968, 204-207). The author decided to follow the Suetonian model for his lives, thereby focusing on the fantastic (White 1967, 118). The interesting aspect of the SHA is that the falsities multiply through the work, leaving scholars such as Syme to argue that the author warmed to his task (Syme 1968, 204-205). It is thought that there are more facts in the earlier lives and we know that Herodian and Marius Maximus (whose work does not survive) were used as sources for some of the Vitae (White 1967, 129; Birley 1987, 230),15 but starting with the Vita Macrini there is a steady increase in the exaggerations (Syme 1968, 97).

Another source worth mentioning is the Notitia Dignitatum, which lists various sites along the Middle Danube frontier in the fourth century. The work is a list of the offices of the empire, from the praetorian prefects down to the provincial governors, given in order of precedence or geographic sequence for both parts of the empire, along with details on the Roman order of battle (Tomlin 1996, 1049). While the information within the Notitia does not extend beyond AD 395 for the eastern part of the empire and 400 for the western half, the work itself is thought to have been compiled sometime around 408, possibly for Stilicho (Mann 1991, 215, 216, and 218).

The SHA is still an important text because it provides a continuous account of events from the early second through the third century AD, a period which is not well documented as a whole. The work also provides information on emperors such as Lucius Verus and Marcus Aurelius, whose principates are described by Dio, but only the epitomes survive, while Herodian starts with the death of Marcus Aurelius. For the Middle Danube, the life of Marcus Aurelius and Commodus provide interesting statements about the Marcomannic Wars. There are also sporadic references to the Middle Danube frontier in the other lives.

The Notitia Dignitatum provides information on the position of various Germanic tribes in the empire, along with a detailed description of Pannonia, listing forts and the position of various units (Marsina, ed., 1998, 250251). Even though a late work, the Notitia helps to reconstruct the organization of the frontier for earlier periods in conjunction with the above mentioned Antonine Itinerary.

The Geography by Ptolemy Claudius, who wrote sometime in the middle second century AD, is not like the Geography of Strabo. While Strabo’s work gave detailed descriptions of various regions and those people living in them, with digressions on various topics, Ptolemy’s work had a different goal. The work is primarily a treatise on cartography, providing a detailed topography of Europe, Africa and Asia, a discussion on the role of astronomy and other types of data-gathering for geographical investigations and detailed instructions on how to construct maps; these three aspects of the work ensured its survival through the Middle Ages (Berggren and Jones 2000, 3). After putting forward his theories on

A different type of source is the Peutinger Table, which is a Medieval copy of a late Roman map that is based on a modified 4th century map of a 2nd century one (Purcell 1996b, 1151). The map provides important information on the location of sites within the province of Pannonia, which can be used in combination with the Antonine Itinerary and the Notitia Dignitatum.

23

When Burebista was still alive and the Dacians at the height of their power, Caesar had planned for a campaign against them (App. Ill., 13). Before he could carry this out, Caesar was assassinated, Burebista died, and Dacian power in the Middle Danube region disappeared. Octavian, however, also created a plan for a Dacian campaign, which was started in 35 BC with attacks against the Iapydes, who were located east of Aquileia (App. Ill., 18). This tribe had conducted raids against the Romans and had also stopped paying taxes, which were used as a pretext for the Roman attack (App. Ill., 18; Dio Cass. 49.34.2). After the defeat of the Iapydes, Octavian moved against the Pannonians, taking the town of Siscia on the Save River in 35 BC (App. Ill. 22). Dio describes the Pannonian campaign as a way to provide practice for the soldiers and to show the power of Rome (Dio Cass. 49.36.1). The following years, 34–33, Octavian campaigned against the Dalmatians (App. Ill., 23; Dio Cass. 49.38.1). All of these campaigns have been explained as an attempt by Octavian to strengthen his position against Anthony, creating a land bridge between Italy and the Balkans, and/or for Octavian to gain prestige as a commander, using the Dacians as an excuse for conducting these various campaigns (Jones 1970, 33; Mócsy 1974, 22–23; Burns 2003, 198–199, 203).

2.3. Historical Account of the Middle Danube Region: 100 BC–AD 166 2.3.1. Final La Tène: 100–10 BC In the Final La Tène (100–10 BC), Celtic expansion reached its largest extent in Europe, with attempts of settlement in Italy, southern Gaul, as well as further east in Central Europe. Celtic society was going through many changes in this period, with increasing urbanization, complexity, and instability sparking warfare and rivalries between groups (see section 3.1; Wells 1999, 32–33, 71–72). Along the Middle Danube, the Celtic Boii had expanded from the Czech Republic east and southeast, gaining control of southwest Slovakia and Pannonia (Strabo 5.1.6; Dobiáš 1964, 29; Mócsy 1974, 14–15) (Fig. 2.4). Another Celtic group in Slovakia were the Cotini, occupying central Slovakia, and to the east of them were the Anartii (Mócsy 1974, 19) (Fig. 2.4).17 At this time, in the late 50s into the 40s BC, the Boii were ruled by Critisarius, who also had control over the Taurisci and it is thought that the Osi, Cotini and Anartii were allied with or were part of the Boii (Mócsy 1974, 19) (Fig. 2.4). By LTD3 (50–10 BC) the Boii tried to expand even further eastward into eastern Hungary and Romania, while at the same time the Dacians had been expanding westward from Romania, across the Tisza River into northern Hungary and eastern Slovakia (Pieta and Zachar 1993, 147). A conflict arose between the two, mentioned by Strabo, ending in 45 BC along the Tisza River (Strabo 7.3.11, 7.5.2). The Getan chief Burebista led the Dacians, Scordisci, and Getae to victory over the Boii and Taurisci. In the archaeological record there Dacian material appears in southwest Slovakia by the second half of the first century BC, with the Boii being pushed back into the Czech Republic (Pieta and Zachar 1993, 200–201). Pliny the Elder wrote that for a brief time the westward boundary of Dacian territory was the Morava River, but that later they had been driven out by the Sarmatian Iazyges (Plin. HN. 4.12.80; Pieta 1981, 104).

Roman attention in the region shifted to the Lower Danube after 33 BC, with wars against the Thracians and Getae (Dacians) (Mócsy 1974, 23). There was relative stability in the Middle Danube region from 28 to 16 BC, when in that year there was trouble in Macedonia, and at the same time the Pannonians, with Norican allies, invaded Istria (Dio Cass. 54.20.3). While Tiberius was sent to deal with the problems in Macedonia in 14 BC, a new campaign was initiated against the Pannonians with M. Vinicius and/or Agrippa in command (Vell. Pat. 2.96; Flor. 2.23; Dio Cass. 54.28.1).19 By 13 BC, Tiberius had suppressed the Scordisci in Macedonia and then was sent against the Pannonians, taking over the command of Agrippa, who had died in either 13 or 12 BC. Three years of fighting followed and in 9 BC the Pannonians were finally defeated, with the province of Illyricum created around 11 BC (Dio Cass. 54.31.2; 55.2.4; Vell. Pat. 2.96.2; Mócsy 1974, 34). At this point the region of Pannonia was part of the greater province of Illyricum and the Roman frontier was brought to the southern border of Slovakia, along the Danube (Mon. Anch. 5.30).

The Boii were weakened by their battles with the Dacians and soon came under pressure from Germanic tribes pushing south into the Czech Republic from Eastern Germany. The combination of the military defeat and Germanic pressure caused the Boii to leave the Middle Danube region and push westward, allowing Pliny the Elder to refer to the Boii homeland as “deserta Boiorum” (Plin. HN. 3.14.146). Some Boii crossed the Danube into Noricum, attacking Noreia and then joining the Helvetii in their migration southwest (Caes. B Gall. 1.5).18 The Dacian occupation of the Pannonian plain was short-lived and by the late 40s BC, after the death of Burebista (in 44 BC), the Pannonian tribes had pushed the Dacians east (Mócsy 1974, 19). The Osi, identified as Pannonian (Tac. Germ. 43), had crossed the Danube at the bend and settled along the Ipeľ River (southeastern Slovakia and Hungarian territory directly north of the Danube at the bend) (Fig. 2.5), but Dacian material still appears archeologically in southern Slovakia.

2.3.2. Early Roman Period: 10 BC–AD 166 As already mentioned, by the last quarter of the first century BC, the Boii had suffered setbacks because of the Dacians and were coming under pressure from the north by Germanic groups, both of which led to their exit from the Czech Republic. This German pressure has been identified with the Marcomanni and allied German tribes, who were pushing east and southeast, but attributing the pressure to the Marcomanni does not work archaeologically. German finds start to appear in small amounts within Bohemia already at the beginning of the first century BC (Dobiáš 1964, 75, 354; Droberjar 2002, 24

Elbe (Flor. 2.30; Vell. Pat. 2.97.2).21 After the death of Drusus in 9 BC, Tiberius took over and consolidated Roman victories from 8–7 BC (Cary 1965, 500). It was probably these campaigns under Drusus and Tiberius that caused the Marcomanni and Quadi to leave their German homes and move east to escape the reach of the Roman military (Dobiáš 1964, 78, 355).22 We know that the Marcomanni were definitely out of Germany around 7 BC, because Dio reports that Domitius Ahenobarbus, governor of Illyricum, allowed the Hermunduri to settle in the vacated Marcomannic lands (Dio Cass. 55.10a.2).23

72). It would appear that there was a steady infiltration of Boii land by German settlers, which probably grew steadily so that by the 80s BC, the Boii started to leave the Czech Republic (Dobiáš 1964, 29, 349–348).20 From the archaeological material in the Czech Republic we may conclude that Germans were definitely the primary power by the start of the first century AD, with their cultural material dominating the archaeological assemblages (Droberjar 2002, 168). The Celts, however, had not disappeared and Pliny’s comment of a “deserta Boiorum” should not be taken at face value. Celts were still in the area, but their strength and organization did not match what had been present in the second or first part of the first centuries BC (Kolník 1991, 432).

There is also mention that in 10 BC M. Vinicius was in command of forces that crossed the Danube to deal with a Dacian threat.24 He not only defeated the Dacians’ but their allies, which included the Cotini and Osi, indicating that Vinicius’ campaign brought him close to or even into eastern Slovak territory (Dobiáš 1964, 34, 353; Marsina 1998, 146).

Velleius reports that the Marcomanni were led by Maroboduus and they migrated east into Boii lands to escape Rome’s power in the west (Vell. Pat. 2.108.2). Tacitus also provides similar information and that while the Marcomanni occupied Bohemia, the Quadi under their king, Tudrus, settled near the Danube (Tac. Germ. 42). Tacitus also mentions that various groups were subjugated by the Germans in the region (i.e., Czech Republic and Slovakia), namely the Marsigni, Cotini, Osi and Buri, further adding that the Cotini and Osi paid tribute to the Quadi; of the last two, Tacitus reports that the Cotini were Celtic and Osi, Pannonian (Tac. Germ. 43). By the last quarter of the first century BC, the Cotini were probably located in the northern part of central Slovakia and the Osi in northeastern Hungary between the Ipeľ River and the Danube bend (Dobiáš 1964, 33– 34, 352–353) (Fig. 2.6).

Once in Bohemia, Maroboduus solidified his power over the Marcomanni and increased the military strength of the tribe (Vell. Pat. 2.108.2). Velleius reports that Maroboduus set himself as king and wanted to create an empire of his own, subduing or by treaty allying neighboring tribes (Vell. Pat. 2.108.2); he created a kingdom that ruled territory from the Lower Elbe in the west to the Danube opposite Noricum and Pannonia, up to the Morava River in the east (Dobiáš 1964, 95–96; Kossack 1981, 298) (Fig. 2.7). The army he formed was well-trained and disciplined (Vell. Pat. 2.108.2). Little is known about Maroboduus, but that he was of noble birth and spent his youth in Rome. Dobiáš postulates that he may have been of a similar position as other sons of foreign allies and client kings under Augustus, thereby receiving the same education as the grandsons of the emperor and other foreign boys, such as Herod’s (Dobiáš 1964, 89, 356; see also Braund 1984, 9–17).25 Growing up in Rome, Maroboduus would have been exposed to the workings of the Roman Empire, learning of state administration and military organization (Dobiáš 1964, 89, 356).

Both the Marcomanni and Quadi were originally located in Germany, but precisely where is debated. Earlier research had the Marcomanni on the Upper and Middle Main River and the Quadi slightly further north, somewhere in the area of Harz or Vogelsberg (Dobiáš 1964, 75, 354). In recent years, it is thought the Marcomanni may have originated somewhere in Mecklenburg or Brandenburg (Droberjar 2002, 168). The Quadi have been identified as originally being part of a larger tribe (or tribal confederation) called the Suebi, or Suevi, (see Tac. Germ. 38), mentioned by Caesar as allied with the Marcomanni and other Germans under Ariovistus in 58 BC (Caes. B Gall. 1.51; Dobiáš 1964, 75–76, 354; Droberjar 2002, 148). After the defeat of Ariovistus and the retreat of German forces back over the Rhine, there is nothing more known about the Marcomanni for almost 50 years, while the Suebi continued to harass and trouble the Romans (Dobiáš 1964, 76, 355). In 55 BC, Caesar mounted his first German expedition to deal with the Suebian threat, by showing the power of Rome in crossing the Rhine and penetrating German territory (Caes. B Gall. 4.1–19).

During this period, the Romans still held German territory up to the Elbe River, preparing it to be made into provinces. Maroboduus’ kingdom abutted the Roman Empire both on the Elbe and on the Danube, while extending to the Baltic in the north and Vistula in the east (Droberjar 2002, 172) (Fig. 2.7). He established diplomatic relations with Rome, sending embassies, sometimes as supplicant other times as an equal (Vell. Pat. 2.109.2). It is thought that by having extended his power to the Elbe, the local Langobardi and Hermunduri tribes became subjects of Maroboduus, which would have caused Roman consternation since Hermunduri territory crossed the Elbe into Roman territory (Dobiáš 1964, 97, 358). In AD 4 and 5, Tiberius had continued operations in western Germany, reducing any opposition. As Tiberius approached the territory of the Langobardi, they fled over the Elbe into Marcomanni territory, along with the Semnones (Dobiáš 1964, 97, 358). Because he harbored these groups, who were escaping Roman power,

The Suebi, however, continued to be a menace and the Rhine was an unstable frontier from 55 to 16 BC, in which year there were new raids by the Sugambri. In 12 BC, a new campaign was launched into Germany under the leadership of Claudius Drusus, who overran German tribes, including the Marcomanni and Quadi, reaching the 25

him in Ravenna (Tac. Ann. 2.63; Suet. Tib. 37).27 Catualda then became ruler over the Marcomanni, but only briefly, being overthrown by the Hermunduri under the leadership of Vibilius; Catualda then fled to the Romans, obtaining asylum at Forum Julium, a colony in Gallia Narbonensis (Tac. Ann. 2.63).28

Maroboduus came to be seen as a direct threat to stability in the German territories between the Rhine and Elbe and also along the Danube (or he had been already identified as a threat to the Romans and this incident was used a pretext for what happened next). In AD 6, a campaign was called for by Augustus and organized by Tiberius for the invasion of Marcommani territory and the destruction of Maroboduus’ power. The attack was to have been twopronged, with Tiberius invading Marcomanni territory from Carnuntum, marching up the Morava River, and Sentius Saturninus invading from the west, following the Main River through Chatti lands (Vell. Pat. 2.109.5).

At this point, the Romans considered the followers of both Maroboduus and Catualda too dangerous to allow them to go back beyond the Danube on their own.29 Tacitus reports that they were settled between the Marus (Morava) and Cusus (Váh) Rivers with Vannius from the Quadi named king over them sometime in AD 19 (Tac. Ann. 2.63).30 This territory is in southwest Slovakia, but German cultural material already starts to appear there in the early part of the B1a phase (Kolník 1981, 116). It is most likely that the Quadi, having occupied the territory east of the Moravia, were slowly expanding into Slovakia, when the Romans gave Vannius the territory between the Morava and Váh Rivers (Fig. 2.8). This territory had been under Dacian control up to AD 15, when in that year the Iazyges first appeared in the region, forcing the Dacians to retreat further east (Dobiáš 1964, 150). It appears that with Vannius appointed ruler of the lands in southwest Slovakia, the core of Quadi territory also shifted from Moravia to Slovakia (Dobiáš 1964, 152–153, 365; Kolník 1998b, 62).31 We do not know, however, where Vannius’ capital was located.32 From this time on, archaeologically we see continued German expansion in Slovakia, moving from the southwest to the east, eventually stopping on the western border of the Sarmatian Iazyges, who maintained control of Hungarian territory between the Danube and Tisza Rivers (see section 2.4.2).33 By the end of Vannius’ reign, the Quadi had control of territory extending from the Morava River in the west to the Dura River (most likely the modern Ipeľ) in the east, bordering with the Osi who still lived there and the Iazyges (Fig. 2.8). Within this territory, the Germans occupied the Morava River and the lower reaches of the Váh and Hron Rivers, while their northern border extended to the Lesser Tatra Mountains, where the Cotini resided.

In that same year, AD 6, a new rebellion broke out among the Dalmatians, who were soon joined by the Pannonians, both led by two separate chieftains named Bato. It is thought that their rebellion was caused by Tiberius’ burdensome requisitions in Illyricum in preparation for the invasion of the Marcomanni (Dio Cass. 55.29.1–2; Cary 1965, 503; Mócsy 1974, 37). This rebellion occurred just as Roman forces were reaching their objective against Maroboduus in the Czech Republic. The Romans were forced to retreat, however, with Tiberius marching back to Pannonia and taking command to crush the rebellion, conducting campaigns until AD 8; at which point Illyricum was divided into the provinces of Pannonia and Dalmatia (Dio Cass. 56.14–16; Suet. Tib. 16; Mócsy 1974, 39). Maroboduus had gotten a reprieve and held on to power until AD 17/18. By this time the political power of the Marcomanni had declined and the once subject Semnones and Langobardi rebelled, joining the Cherusci under the leadership of Arminius (Tac. Ann. 2.45). They saw Maroboduus as a traitor to the Germans and a puppet of the Romans (Tac. Ann. 2.45). In the ensuing battle somewhere north of Bohemia, the outcome was indecisive on the battlefield, but because Maroboduus left the battlefield to seek higher ground, this act was seen as a defeat and allied tribes, such as the Hermunduri, abandoned him (Dobiáš 1964, 104, 362; Droberjar 2002, 171). With his authority weakened, Maroboduus was forced to retreat into Bohemia (Tac. Ann. 2.46). At this point, he appealed to the Romans for aid against the Cherusci, but the Romans reminded Maroboduus that he himself had not provided aid when asked for by them against the same tribe (Tac. Ann. 2.46).26 Drusus was sent to settle some kind of peace among the Germans, at least on the surface, but in fact he was sent by Tiberius to ensure the fall of Maroboduus (Tac. Ann. 2.47 and 62; Mócsy 1974, 40).

By the late 40s AD, Vannius had amassed a large fortune, which Tacitus claims was accumulated through praedationibus et vectigalibus (Tac. Ann. 12.29). The use of the term praedationis was a Tacitean device to paint an image of a tyrant and may not have been that severe, but the vectigales has more truth behind it (Vrba 1997, 18). One of the primary paths of the Amber Route was along the Morava River, whereby one would have then traveled into Poland and onward to the Baltic.34 Since the Morava was in the midst of Quadi territory, Vannius would have had direct control of goods traveling along the last leg of the route before entering the Roman Empire (Krekovič 1997a, 234). Vannius most likely accumulated his great wealth by imposing taxes and tariffs on the trade (Krekovič 1997a, 234; Vrba 1997, 18; Valachovič 1998, 28). However, it should be noted that Vannius did exact tribute from subject tribes, such as the Cotini and Osi, and probably did amass booty from campaigns against

Drusus used for Roman purposes a Gotonian named Catualda, who had been banished as a youth by Maroboduus and now wanted revenge (Tac. Ann. 2.62). Catualda put together a strong force, invaded Marcomanni land, and stormed the palace and fortress of Maroboduus in AD 19 after bribing Marcomannic nobleman to abandon their support for the king (Tac. Ann. 2.62). Seeing himself forsaken and without support, Maroboduus fled to Noricum, from where he wrote a letter to Tiberius seeking asylum, which was granted to 26

other Germans, thereby providing Tacitus with some evidence for his use of praedatio.

with Maroboduus and Vannius) in that the two at first had the support of the people but then were hated (Tac. Ann. 12.30). This eventual hatred may be explained not only by possible tyrannical rule of some sort, or Tacitean device, but also that being loyal vassals of Rome, they may have lowered taxes on the Amber Route, cutting revenue; also, they may have relaxed hold of the route itself, allowing more control to be passed to Roman merchants (Vrba 1997, 20). We know that Roman merchants operated beyond the frontiers of the empire, partially from the famous passage by Tacitus, when relating the downfall of Maroboduus, which states, “[In the palace Catualda and his force] discovered the ancient Suebian spoils, together with a number of settlers and traders out of the Roman provinces, drawn from their respective homes and implanted on hostile soil first by the commercial privileges, then by the lure of increased profits, and finally by oblivion of their country” (Tac. Ann. 2.62).37 Pliny the Elder provides evidence that may support the idea that Roman merchants gained a greater control of the Amber Route through Vangio and Sido. Pliny states that the origin of amber, along the Baltic, had been only recently confirmed because of a Roman knight who was commissioned to procure amber during the principate of Nero (Plin. HN. 37.11.45–46). The knight had actually traveled along the Morava River northward, then on to the Oder, lower Vistula, and finally the Baltic coast, bringing back a great quantity of amber for the gladiatorial games (Fig. 2.9). According to Pliny amber had been known to Greeks and Romans for quite some time and by the period of the empire the conveyance of it into Pannonia was controlled by the Germans (Plin. HN. 37.11.43). By stating that the distance from Carnuntum to the place of amber’s origin was only recently confirmed would seem to indicate that Romans had not traversed the Amber Route before, and this may be explained by a tight control maintained by the various German tribes. The ease by which the Roman knight had traveled the route under Nero would seem to indicate a change, explained by the easing of control by Vangio and Sido.38

Vannius was able to hold on to power for thirty years. Tacitus states that the Quadi king began with the support of his countrymen, but abuse of power transformed his reign into one of tyranny (Tac. Ann. 12.29). In the last years of Vannius’ reign, there were internal disputes and neighboring German groups had turned hostile (Tac. Ann. 12.29). Based on what Tacitus provides, that Vibilius, king of the Hermunduri, provided aid to Vangio and Sido, the sons of Vannius’ sister, to rebel (Tac. Ann. 12.29), there must have been a power struggle over who would succeed the old king among the Quadi elite. Vannius having ruled for thirty years was probably an old man by AD 50 and he possibly had no clear successor. Vangio and Sido may have been excluded from power, and so decided to overthrow their uncle. Vibilius of the Hermunduri saw Vangio and Sido as an opportunity to weaken the power of the Quadi and to plunder the amassed wealth of Vannius. The Hermunduri assembled a large army, joined by the forces of the Lugii and other tribes, but Vangio and Sido probably had the help of disaffected Quadi as well. Vannius immediately sent appeals to the emperor Claudius for aid, which were refused. The only consolation Claudius provided was the promise of safe refuge for Vannius if he lost and the posting of a Roman force on the Danube to protect the losers and intimidate the winners to ensure peace along the frontier (Tac. Ann. 12.29). On the surface, it would appear that Claudius was keeping the Romans out of what seemed to be an entirely German affair, but this was not the case. Since Vannius was created by the Romans and was supposedly a client king, Claudius’ refusal of help probably meant the Romans were unhappy with Vannius.35 This unhappiness could have arisen from the taxes imposed on the Amber Route (see below).

The next we hear of the Quadi is toward the middle of AD 69, at which time Vitellius and Vespasian were vying for the principate. When the legions of Pannonia and Moesia threw in their support for Vespasian, the officers were able to obtain also the support of the Quadi kings, Sido and Italicus, who had been loyal to the Romans for a long time (Tac. Hist. 3.5). Apparently Vangio had died; how is unknown, as is the relation of Italicus to Vangio and Sido.39 The two brought troops with them and joined the Pannonian forces under the command of Antonius Primus. They participated in the second battle outside Cremona, taking front-line positions in the battle order, and most likely (if they survived the battle) participated in the later siege and pillaging of the city (Tac. Hist. 3.22; Dobiáš 1964, 155, 368).

Vannius had a force of German infantry and cavalry recruited from the Iazyges, but he realized his forces were too small to defeat in open battle the combined forces aligned against him (Tac. Ann. 12.29). The king had hoped to buy time through defensive actions, using his strongholds, but was betrayed by the impatience of his Iazyge cavalry, who brought about a direct confrontation too soon. In the ensuing battle, Vannius is reported to have fought valiantly, suffering frontal wounds (meaning he got them in combat, not in retreat), but was forced to flee to the Danube. There he boarded the Roman river fleet and was soon joined by his dependants. They were then allowed to settle in Pannonia (Tac. Ann. 12.30).36 Vangio and Sido shared the Quadi kingdom, in that one probably took control of the old Marcomannic territories and the other the Quadi lands, but ruled as a condominium (Dobiáš 1964, 154, 368).

By the time of Domitian, Roman attention had been drawn to the southern course of the Danube and the growing threat of the Dacians. There was a resurgence of Dacian power under the able leadership of their king Decebelus, who led raids across the Danube into Moesia

Once in power, the two were loyal clients of the Roman Empire. Tacitus again follows the same formula (as seen 27

(Dobiáš 1964, 171, 369). The first raid in AD 85/86 led to a defeat of the Moesian governor Oppius Sabinus and destruction of a legion, while a retaliatory raid by Cornelius Fuscus into Dacian territory ended with his defeat in 86 (Suet. Domit. 6; Tac. Hist. Frag. 6). In 88/89, the second Roman expedition under Tettius Iulianus was able to capture the Iron Gates and Decebalus sued for peace, which was ignored. Domitian at this point started preparations for a final campaign against the Dacians. He asked for and fully expected aid from his client kings along the Middle Danube, but instead the Marcomanni and Quadi refused (Dio Cass. 67.7.1). Instead of proceeding with his plans against Decebelus, Domitian decided to punish the Germans for their refusal (Dio Cass. 67.7.1).

anger the Marcomanni and Quadi, who quickly made an alliance with the Iazyges and crossed the Danube into Roman territory (Dio Cass. 67.5.2), ominously foreshadowing what would happen in AD 166. The combined German and Sarmatian force met and destroyed an entire legion, thought to be the XXI Rapax (Eutrop. 7.23; Dobiáš 1964, 174, 371; Southern 1997, 111). The situation became so grave along the Middle Danube that Domitian himself journeyed to Pannonia in the spring of 92 and campaigned until very early 93.41 There is uncertainty about the name of the war and what evidence refers to it, especially various inscriptions dating to this period (Southern 1997, 111). The war is referred to either as the Germanic, Suebian, or Marcomannic War, or German and Sarmatian War, or even just the Sarmatian Wars, which is mentioned by Suetonius and Eutropius (Suet. Domit. 6; Eutrop. 7.23; Dobiáš 1964, 175, 371; Southern 1997, 111). Domitian was able to drive the invaders out of Pannonia and reestablish some stability along the Middle Danube.42

Dobiáš believes that the Marcomanni and Quadi had not wanted to help because a Roman victory in Dacia would have pushed the boundaries of the empire beyond the Danube, bringing the frontier to the eastern edges of Quadi territory, thereby partially encircling them (Dobiáš 1964, 172, 369). Also, with the previous defeat of Domitian’s forces under Sabinus and Fuscus, the prestige of the Romans had been hurt, thus emboldening the Germans to ignore Domitian’s requests (Dobiáš 1964, 172, 369).

This stability was short-lived, because in the second year of Nerva’s principate, hostilities broke out again (Dobiáš 1964, 175, 371). Dobiáš is of the opinion that the Germans invaded Pannonia again, but that they were defeated in the autumn of 97, enabling Nerva to take the title of Germanicus, which was then shared with Trajan (Dobiáš 1964, 175, 371).43 The victory was decisive because there was a relative peace along the Middle Danube during Trajan’s principate.44 It is also of interest that much of the stamped roof-tiles found at sites with Roman-style structures in Slovakia, such as Devín and Stupava, as well as random finds from Bratislava, are marked with stamps from the XIV and XV legions, the presence of which along the Danube are dated to around AD 100 (Dobiáš 1964, 175, 371) (see sections 3.7–8).

Domitian took command in 89 of the expedition against the Marcomanni and Quadi, who in fact did not want war, sending two embassies to sue for peace, of which the second was put to death by the emperor (Dio Cass. 67.7.1). Soon after, Domitian crossed the Danube and invaded the territory of the Germans. The details of this campaign do not survive, but Dio reports that the Romans were defeated by the Marcomanni (Dio Cass. 67.7.2), so it is possible that the major thrust of the campaign had been from Raetia into Marcomannic territory. This defeat forced Domitian to retreat back to the empire and sue for peace with Decebelus.

Around this period, in 98, Tacitus published his Germania in which he described the Marcomanni and Quadi, as well as the territory they held. The Marcomanni were described as strong and the Quadi no less so (Tac. Germ. 42). Tacitus also makes an interesting point, that up to the period right before he is writing, the Marcomanni and Quadi had been ruled by members of the noble houses of Maroboduus and Tudrus (meaning, not only Maroboduus, but also Vannius, Sido, Vangio, and Italicus), but now, they were ruled by foreign kings, whose power rested within the influence of Rome (Tac. Germ. 42). One wonders if Sido and Italicus died at Cremona, having been in the front ranks, which played to Roman advantage, since they could impose whomever they wanted afterwards on the Quadi. But also, the recent victory by Nerva may have also included the imposition of new rulers over the Marcomanni and Quadi, enabling Tacitus to write as he did. Tacitus goes on to say that Romans regularly provided assistance to maintain the status quo either by armed intervention, but more often by subsidies (Tac. Germ. 42). He also states, as already mentioned before, the Quadi and Iazyges also imposed tribute on the Celtic Cotini, who were located in northcentral Slovakia, and the Osi, who were still located between the Danube and Ipeľ Rivers (Tac. Germ. 43).

Around 91 or early 92, there were new problems along the Middle Danube frontier, when a tribe, referred to as the Λύγιοι in Moesia by Dio, requested help from the Romans against the Suebi (Dio Cas. 67.5.2). Based on what is said further, it is clear that this tribe was not in Moesia but north of Moesia and in fact was the Lugii, already mentioned as having allied with the Hermunduri to help Vangio and Sido in AD 50 (Dobiáš 1964, 174, 371; Southern 1997, 110–111). The Lugii were nowhere near Moesia, but in fact were located north of the Marcomanni in southern Poland and identified as belonging to the Vandals and most likely the archaeological culture known as Przeworsk (Dobiáš 1964, 95, 371; Kossack 1981, 299) (Fig. 2.8). Domitian sent off a force of 100 equites, more for show or honor, than actual military help (Dio Cass. 67.5.2).40 Domitian may have been trying to send a message to the Marcomanni and Quadi that the Romans were displeased with them, but the force was kept small since the emperor did not want to become entangled in a war a long way from the frontier. The end result was that the force was too small to have any military effect but large enough to 28

There was a general peace until AD 117, at which time there was a Hadrianic expedition against the Suebi and Sarmatians, as we learn from an inscription commemorating the career of L. Cesennius Sospes found in Anatolia (CIL III 6818 = Dessau, ILS 1017; Dobiáš 1964, 176, 372, and footnote 50 from Chapter Four). The inscription records that Sospes was legate of the XIII legion and was decorated for his actions during Hadrian’s Suebian and Sarmatian expedition (Dobiáš 1964, 176, 372). The SHA also implies a campaign against the Sarmatians at this time in response to raids across the Danube (SHA Hadr. 6.6).

of the Marcomanni, Quadi, Iazyges, and other groups crossed the Danube and invaded the empire, precipitating what came to called the Marcomannic War. 2.4. Conclusion Slovakia can be divided into three geographic regions: southwest, central (further divided between north and south), and east. Each of these regions at various periods either saw the same cultural groups or different ones, but based on the archaeology, southwest Slovakia saw more influences from the west, while east Slovakia saw more influence coming from the east. The Roman Period of Slovakia, AD 10–375, is divided into three main periods: Early Roman (AD 10–166), the Marcomannic War (AD 166–180), and the Late Roman (AD 180–375).

A little later into Hadrian’s rule, there is also a vague reference by Dio about how Batavian forces swam the Danube fully armed, scaring the barbarians on the other side (Dio Cass. 69.9.6). It is known that there was a Batavian cavalry force stationed in Pannonia during the second century at the castellum near modern Füzitö (5 km to the east of Brigetio), and Dobiáš believes, therefore, that this event took place on the Danube opposite the Quadi (Dobiáš 1964, 177, 372). Based on Dio’s account, there were some kind of internal problems among the Quadi and they may have been planning a raid against the Romans, but when they saw the fully armed Batavians swim the Danube, it so terrified the Quadi that they then employed Hadrian as an arbiter to settle their problems (Dio Cass. 69.9.6). It is at this point, Dobiáš believes, Hadrian appointed a new king for the Germans, meaning then the Quadi specifically, sometime around 122/123 (SHA Hadr. 7.7; Dobiáš 1964, 177, 372).

Based on the various ancient Roman and Greek written sources, we may conclude that the Quadi tribe of Germans settled in the region of southeast Moravia and southwest Slovakia, creating a powerful economic and political kingdom that was made a client of the Roman Empire. At various points during the Early Roman Period, the Quadi were either allies or enemies of the Romans, depending on the circumstances, having taken advantage of a weakened or distracted Roman administration to raid the provinces while the Romans, when necessary, either invaded southwest Slovak territory against the Germans, or through political manipulations caused hostile or weak elites of the Quadi to be replaced by those more friendly to Roman policies.

From 123 to 166, there was relative stability, but attacks on the frontier by the Germans and Sarmatians along the Danube still occurred. One such incursion is recorded as having taken place in 136/137, but was dealt with successfully by L. Aelius Caesar, Hadrian’s first designated successor and governor of both Pannonias in 136; and in 138, the governor of Upper Pannonia, Ti. Haterius Nepos, was rewarded with ornamenta triumphalia, which would imply some type of military action against either the Germans or Sarmatians (Dobiáš 1964, 177, 372; Mócsy 1974, 103). We know a new king was given to the Quadi by Antoninus Pius sometime between 140 and 144 because a coin was issued commemorating the event. The coin has on the reverse Rex Quadis Datus, with the emperor shown wearing a toga, not armor, and placing a diadem on a barbarian, who stands at the same level (Dobiáš 1964, 177, 372). The representation of the emperor wearing a toga, a symbol of peace, can be either taken to mean that the creation of the king was done through non-violent means, or it was hoped that there would be peaceful relations with the Quadi through this new king. The idea that the creation of the Quadi king was a peaceful process is supported by Eutropius and Aurelius Victor, who state that Antoninus Pius was so respected and feared by allied kings and barbarian nations, they referred their controversies and disputes to him, abiding by his decisions (Eutr. 8.8; Aur. Vict. Caes. 15). In 166, however, the stability and even survival of the Middle Danube frontier was put to the test when combined forces

Endnotes 1. J. Dobiáš’ Dějiny ČeskoSlovenského Území Před Vystoupením Slovanů (1964) is still the primary secondary source in the Czech Republic and Slovakia for the Middle Danube region during the Roman Period. Dobiáš not only used ancient authors but also the latest epigraphic, numismatic, and archaeological evidence as well, but much has been found since 1964 and there has been comment in Slovakia among archaeologists that a replacement is needed. The book also has an English summary, which is quite extensive, but its greatest weakness is the absence of the footnote numbers present in the Czech text. I used Dobiáš as the starting point for the historical account in Chapter 2, and where he is cited I have tried to provide both the page number from the main text and the English summary, while also providing the appropriate footnote references, because many times these are the critical citations. 2. The reason that the southern portion of central Slovakia would have followed developments seen in western/southwestern Slovakia is because of actual geography. The southern section of Slovakia covers the northern edge of the vast Pannonian plain, while the northern and eastern parts of the country are covered by the Carpathian Mountains.

29

3. The dates being used here are from an updated chronology from the website

created by Mgr. A. Žitňan, an archaeology student at Komenius University under the supervision of Dr. E. Krekovič, an archaeologist and professor at the university.

to have been invented by B. Rhenanus, who rediscovered the work in 1515 and published an edition in 1520 (Sumner 1970, 278-279).

4. The section does not attempt to list every author possible, but instead to provide a brief look at the most important ones. For a complete list, see the extensive catalog of ancient authors that mention the Middle Danube in Part II of Územie Slovenska Pred Príchodom Slovanov (Marsina, ed., 1998, 105-275, 350-360).

13. This accusation is further weakened by placing the work within the context of the times (Sumner 1970, 281283, 293-294).

5. Because Julius Caesar and Suetonius do not give much information about the Middle Danube, they will not be discussed. It should be noted, however, that Julius Caesar does provide the earliest description of the Germans in his De Bello Gallico, while also providing detail about various Celts and their movements, including the Boii.

15. White indicates, as well, similarities between the language of Aurelius Victor’s Caesares in Chapter 21 with a portion of Chapter 9 of the Life of Caracalla in the SHA (White 1967, 128). Syme takes this further with his argument that the author of the SHA used Aurelius Victor as well as Eutropius, and that the Res Gestae of Ammianus Marcellinus was an influencing factor not only in composition but also for the creation of the SHA and its themes (Syme 1968, 70-71, 103-104, 107 and 108). There is still debate about the sources for the SHA, with T. Barnes arguing that there was an unknown biographer of good repute used by the SHA for the facts found in the second century lives (Birley 1987, 230).

12. The major criticism has been that Velleius was an apologist for Tiberius, going to great lengths to cast the emperor in a favorable light (Sumner 1970, 281).

14. Syme believed a date of around 395 (Syme 1968, 220).

6. There has been much debate about the dates for the composition of the Geography and where it was written. Originally, it was thought, as put forward by B. Niese, that Strabo wrote the work around AD 18/19 under Tiberius in Rome (Jones 1917, xxviii; Sihler 1923, 135), but E. Pais made strong arguments for the majority of the work having been written around 7 BC, reworked in AD 18, and written in Amasia, for Pythodoris, Queen of Pontus (Jones 1917, xxvii and xxv). Another line of reasoning has the work unfinished at the time of Strabo’s death, being a work in progress and not published until after his death (Clarke 1991, 103).

16. With the identification of Celemantia as being near Iža, roughly opposite from the Roman fort of Brigetio, then the other sites mentioned by Ptolemy Claudius that would be in Slovakia or Poland are Singone, Arsicua, Parienna, Setovia, Asanca and Carrodunum (Ptom. Geog. 2.11.29 and 30; Valachovič 1998, 24). There is also the site of Leukaristos, which is identified with the Latin Laugaricio and appears on an inscription in Trenčin. The inscription records the presence of a detachment from a Legion spending the winter there, which they called Laugaricio, in AD 179/180.

7. References to Dio will follow the numbering system for the books used by E. Cary in his Loeb edition, not the one devised by U. P. Boissevain. 8. Other copies in less complete form have been found at other temples in Asia Minor (Shipley 1924, 333). The best preserved is the one from Ancyra.

17. The Cotini could be identified with the Púchov culture, which was archaeologically first identified and named by E. Beninger in 1937 (Pieta 1996, 26). The archaeological work by K. Pieta at Liptovská Mara has done much to help in understanding the Púchov culture and its links to the Cotini of the ancient sources (Pieta 1996, 33).

9. The group of authors listed provides only a small amount of information for this period and so will not be discussed. 10. Even though written in 1900, A. Gudeman’s article on the sources of the Germania is still relevant and provides a good discussion of the issue (Gudeman 1900)

18. There is a question of how many Boii groups there were and even if the term “Boii” is accurate. It appears that there may have been at least two branches of Boii, but there may have been more (Burns 2003, 202–203). One inhabited the Hercynian Forest, modern day Bavaria and southern Czech Republic, and another group occupied the Po Valley in northern Italy (Dobiáš 1964, 346–347; Frey 1995, 517). But it is also argued that the Boii as a whole settled in Northern Italy, and that when the Celts were expelled from the Po Valley, the Boii then crossed the Alps and eventually settled in Bohemia

11. The reason Velleius is usually identified as a writer of a ‘Roman History’ is linked to the survival of the second book more or less in tact, which is solely focused on the history of the Romans after the fall of Carthage (Starr 1981, 165). The first book, which only survives in a mutilated form, seems to have covered wider topics and in fact started with the founding of Metapontum by Epeus (Star 1981, 163). The title ‘Historia Romana’ is thought

30

(Mócsy 1974, 14). However, it also attested that after the Boii left Northern Italy, ca. 191 BC, they migrated and settled in Pannonia (Dobiáš 1964, 350). It may be that the name “Boii” may have been like the German “Suebi,” a term encompassing multiple tribal groups in a confederation.

was consul that year, 14 BC, and Velleius reports that Agrippa was placed in command and after his death, Tiberius took over (Vell. Pat. 2.96). Florus, however, writes that Vinnius was sent (Flor. 2.24). The name Vinnius is thought to be actually Vinicius (Mócsy 1974, 36). Dio reports that it was Agrippa who commanded, during the consulship of Marcus Valerius (i.e. Vinicius) and Publius Sulpicius (Dio Cass. 54.28.1). It is possible that both Agrippa and Vinicius commanded jointly.

It is possible that at first the Boii were centered around the Hercynian Forest. A group of Boii split, joining a confederation with the Insubres and Senones, who invaded Northern Italy around 396 BC (Frey 1995, 517). These two groups of Boii then existed separately. In 191 BC, with the expulsion of the Boii from Northern Italy, this group then went to the east, conquering and occupying northern Pannonia. By the 80s BC, the Hercynian Boii were coming under pressure from German groups and started to leave Bohemia, expanding both eastward into Slovakia and further, with others traveling westward. In 58 BC, a large group of Hercynian Boii joined the Helvetii in their push to the west, while the Boii in eastern Slovakia and Pannonia were defeated by the Dacians under Burebista in the early 40s. These Pannonian Boii then retreated back into Pannonia and to the northeast, occupying the lands of the Hercynian Boii in the late 40s BC. These Boii too, however, came under Germanic pressure and left Bohemia so that by 10 BC the territory could be easily occupied by the Marcomanni.

20. See endnote 18 above for the movement of the Boii. For a detailed look at the earliest German remains for Bohemia see V. Salač 1995, 145–176. 21. It is reported that the defeat of the Marcomanni was so great that they were annihilated as a people and that Drusus erected a tropaeum out of the spoils (Flor. 2.30; Dobiáš 1964, 77, 355). 22. There is also another explanation based on climate change in northern Europe that caused pressure on groups in northern Germany to move south, thereby putting pressure on the Marcomanni and Quadi (Kolník 1998b, 60). 23. The settlement of the Hermunduri is also mentioned by Tacitus in his Germania (Germ. 41). 24. The evidence is an inscription partially preserved and now lost from Tusculum, today’s Frascati (Dessau ILS 8 965; Marssina 1998, 146).

Another hypothesis I present here has the Boii retreating from northern Italy into Pannonia after 191 BC, but that then in the 80s BC, the Hercynian Boii under pressure from the Germans, vacated Bohemia and traveled south, joining the other Boii in Pannonia. The Boii then try to push eastward into eastern Slovakia and Hungary up to the Tisza River, but come into conflict with the Dacians. While the war with the Dacians may have ended in 45 BC, Boii groups started to leave the Middle Danube region for the west earlier than that, seeing that there was too much pressure from Germans in the north, Dacians in the east and Romans in the south. These Boii joined the Helvetii traveling west, where they came into conflict with Caesar, who eventually allowed the Boii to settle in the lands of the Aedui (Caes. B Gall. 1.28, 7.9).

25. In Augustus’ Res Gestae there is an interesting point where he mentions which kings and their sons came to him for refuge or were sent as supplicants (Mon. Anch. 6.32). Toward the middle of the paragraph, the Marcomanni and Suebi (Qaudi?) are mentioned and at the point where the name of the king or son sent to Rome would have been, there is a lacuna in the text, both in the Latin and Greek versions. In the Greek version –ρος is preserved, but as to whether that was the end of a name or something else, is uncertain. Tacitus does mention that the Quadi king was named Tudrus, so the –ρος could be referring to Tudrus and that a relation of his was sent to Rome (Tac. Germ. 42).

T. Burns makes the interesting point that the use of the term “Boii” to describe the Pannonian Celts may have been a literary device by Roman Period authors in using names of known groups for rhetorical impact (Burns 2003, 203). We do know as well that the Boii did not disappear from Pannonia, because when the province was organized and civitates peregrinae created, one of them was named civitas Boiorum (Mócsy 1974, 66; Burns 2003, 210). This naming of the civitas would seem to weaken Burn’s idea, unless those Pannonian Celts took the name Boii later or were forcibly given the name by Roman administrators.

26. This aid refers to the period when Arminius had destroyed the three legions of Varus in the Teutoburgerwald in AD 9 and liberated almost all the territory between the Rhine and Elbe from the Romans; after which Arminius had planned to cross the Rhine (Dobiáš 1964, 360). Arminius had sent the head of Varus to Maroboduus as an invitation to join him against the Romans, but Maroboduus turned him down and instead sent the head to Rome (Vell. Pat. 2.119.5; Dobiáš 1964, 360). R. Seager makes a good point that Maroboduus was probably abiding to the treaty he had agreed to with Tiberius in AD 6, possibly seeing little chance of success in Arminius’ revolt (Seager 2005, 36). When the Romans launched their campaign against Arminius, they requested

19. There is some question as to who was in command at the beginning of the Pannonian campaign. M. Vinicius

31

Marcomannic aid, but Maroboduus did not provide any, instead remaining neutral (Dobiáš 1964, 361). Maroboduus may have been hoping that both sides would be weakened, thereby giving him some future advantage either with Germans or Romans. His refusal to aid the Cherusci , remaining neutral, is why Maroboduus was labeled as a traitor to the Germans by Arminius

31. This assertion is supported by the archaeological evidence. Once the Marcomanni lost their position and the Quadi gained prominence around AD 19, there is a distinct shift in Roman material appearing beyond the Danube. While there had been a spike in Bohemia of Roman imports before AD 19, there is a marked decline after that time, while in Moravia, among the Quadi groups there, Roman imports start to increase, but they do not equal in number the sheer volume and variety of goods seen in southwest Slovakia, where Roman goods become readily abundant (see Chapters 9 and 10 for more information on Roman imports in southwest Slovakia; Dobiáš 1964, 367; Kolník 1971, 520; 1980, 9; Bouzek and Ondřejová 1990, 24).

27. Suetonius reports that Maroboduus was lured by decoys and promises to accept the terms of asylum (Suet. Tib. 37). Tacitus provides the details where he mentions that Tiberius promised Maroboduus a secure home in Italy and that hopefully once things settled down he would be allowed to return to the Marcomanni (Tac. Ann. 63). However, Tiberius then told the senate that Maroboduus was as dangerous as Philip to Athens or Pyrrhus and Antiochus III to Rome, and so once the deposed king was in Roman hands in Ravenna he was never allowed to leave. In fact, Maroboduus was then used as a threat against the Marcomanni, that if they misbehaved too much, the Romans would re-impose Maroboduus as king (Tac. Ann. 63; see also Braund 1984, 171–172 about the use of exiled kings as threats, which seems to have also been the case for Catualda, Vannius, and also Bato and the wife and son of Arminius).

32. There is some question about who Vannius was and what did he control before AD 19. J. Dobiáš is of the opinion, and I agree, that Vannius was already king of the Quadi, somewhere on the Morava River and after the defeat of Catualda, he also gained control of the Marcomanni, followed by the territory between the Morava and Vah Rivers (Dobiáš 1964, 149–150, 365). Dobiáš also makes a good argument for the idea that during the rule of Maroboduus, the Quadi were under his control, but when he fell, the Quadi were emancipated with Vannius becoming their king, being a descendent of Tudrus (Dobiáš 1964, 149, 365).

28. It is unknown if Vibilius acted on his own, at the instigation of the Marcomannic nobility, or was Roman interference at play again.

In regards to the location of his seat (i.e., capital), T. Kolník believes it was possibly located in the Trnava region, near the three large and rich cemeteries of Kostolná pri Dunjai, Abrahám, and Sládkovičovo (see section 3.2 for descriptions of these cemeteries; Kolník 1971, 520).

29. When both Maroboduus and Catualda fled to the Roman Empire, they brought with them a large group of followers, most likely not only family members but loyal warriors. Apparently the combined group was so great and dangerous, it was thought best not to let them go free, at which point they might have joined with some other disaffected German leader wanting to fight Rome. The solution was to send them back across the Danube and to place them under the control of a client king, loyal to Rome.

33. In fact, the Quadi expansion into southwest and central Slovakia came at the exact time that the Sarmatian Iazyges, a nomadic tribe, had penetrated into the Hungarian Plain, having originated from the coasts of the Black Sea. The Iazyges arrived sometime between AD 15 and the 20s (Dobiáš 1964, 365). They pushed the Dacians and Anartii out of eastern Slovakia and also from between the Danube and Tisza Rivers, where the Iazyges finally made their home (Plin. HN. 4.12.80). They did not hold on to central Slovakia north of the Danube, leaving territory open that could be easily expanded into by the advancing Germans (Dobiáš 1964, 365; Kolník 1998b, 62–63). It is also possible that when the Iazyges arrived and pushed the Dacians out of Slovakia, the situation beyond the frontier was chaotic for a period, and the Romans took advantage of the moment to cede officially the territory to the Quadi, a known quantity, as opposed to leaving it to the Iazyges, an unknown quantity newly arrived on the scene.

30. There has been debate in the past as to where to place the Cusus, with some scholars wanting it to be the Gusen, a small stream near Linz in Austria (Dobiáš 1964, 149,365; Seager 2005, 79). With the naming of the Morava River without doubt, a comparable river in size needs to be sought for the Cusus (Dobiáš 1964, 149). The Gusen is too small and no other river west of the Morava seems to fit, while to the east of the Morava River, the next river of comparable size within reasonable distance is the Váh River (Dobiáš 1964, 149–150). Dobiáš also makes an interesting point that the source of the Vah River has the name of Kysúca, which has similarities to Cusus (Dobiáš 1964, 149, 365). While the juggling of names by itself may not provide definitive answers, the archaeological evidence is beyond dispute, as seen in section 3.2, that the territory of the Regnum Vannianum needs to be sought between the Morava and Váh Rivers (Kolník 1980, 9).

34. The Amber Route is in fact a phrase that refers to a number of paths that went from the Baltic coast through Central Europe to the Mediterranean. While amber was the original motivator for the route, it was not the only commodity traded; others were animal skins, wheat,

32

honey, wax, horses, slaves, timber and metals from the barbaricum for manufactured goods, such as pottery, bronze cauldrons and buckets, jewelry, and other goods (Wielowiejski 1980, 209; Kolník 1998b, 62; 2000, 21; Burns 2003, 213–214). The Amber Route was also a path for ideas to travel upon and in Late Antiquity, even Christianity (Kolnik 1998, 62). Two paths have been identified through Slovakia, of which one appears to have been the most important of the paths making up the Amber Route with the creation of the Regnum Vannianum (Wielowiejski 1980, 212). The lesser path started at Brigetio, traveled up the Váh River, and at Žilina left the river to travel into the Carpathian Mountains and from there into Poland (Wielowiejski 1980, 208; Kolník 1998b, 62). The other, more important, old path started at Carnuntum, traveling up the Morava River into the Carpathians and then on into Poland (Plin. HN. 37.11.42–45; Wielowiejski 1980, 206; Kolník 1998b, 62; 2000, 21).

35. Claudius also probably did not want to take part in a battle amongst Germans since doing so might have enflamed other Germans and prompt them to join in, thereby escalating the conflict beyond the confines of southwest Slovakia. Also, with the conquest of Britannia only just completed and forces committed there, the Romans probably did not have the resources to assist in a way that would have still preserved security along the entire Rhine and Danube frontiers. 36. There is archaeological evidence to support this settlement, because right around this period, German material appears west of the Neusiedler Lake in today’s Burgenland, Austria (Kolník 1971, 520). 37. Translation by J. Jackson (Loeb series). The Latin reads, “Veteres illic Sueborum praedae et nostris e provinciis lixae ac negotiators reperti quos ius commercii, dein cupido augendi pecuniam, postremo oblivio patriae suis quemque ab sedibus hostilem in agrum transtulerat” (Tac. Ann. 2.62).

J. Wielowiejski reconstructed this main path of the Amber Route that starts from Carnuntum through the archaeological evidence of Roman material concentrations and settlements (Wielowiejski 1980, 208). After leaving Carnuntum, one traveled up the Morava River, through Zohor and/or Stupava, and then onward toward the Carpathians, where one would use the Moravian Gates to cross into Poland. Once in Poland, one traveled toward a large concentration of Przeworsk culture settlements in Głubczyce upland, with the main center at Nowa Cerekiew (in Opole province). The route from there went northeast, coming to Rumin near Konin, turning to the west, through Kleczewo and Paniewo (in Konin province), then northwards along the west side of the Gopło Lake, reaching a settlement at Krusza Zamkowa. The path then went northeast, coming to the crossing of the Vistula near Otłoczyn. From there the path followed the Vistula, on its east bank, to Chełmno and on to the Baltic (see Kulakov 2005 for a detailed archaeological account of the amber lands).

38. There is also a funerary inscription on a Roman stele discovered in Boldog, Slovakia, that provides further evidence of Roman merchants operating on the Amber Route after the ascension of Vangio and Sido. This stele originated somewhere along the frontier in Pannonia, probably Carnuntum, but eventually found its way into Slovakia onto the side of a Romanesque-Gothic church (Kolník 1977b, 481; 2000a, 21). The stele commemorates Quintus Atilius Primus, a provisions procurer who operated on the frontier, sometime in the middle of the first century AD, with the stele itself dating to the second half of the first century (Kolník 1977b, 493; 1998, 62). There are 11 lines in the inscription and some dispute as to how to read it. Q ATILIVS “Q(uintus) Atilius SP · F · VOT · PRI Sp(urii) f(ilius) Vot(uria tribu) PriMVS · INTER R EX mus inter(p)rex LEG XV · IDEM > leg(ionis) XV idem (centurio) NEGOTIATOR · AN negotiator an(norum) LXXX · H · S · E LXXX h(ic) s(itus) e(st) Q ATILIVS COCTA Q(uintus) Atilius Cog(i)taTVS · ATILIA · QLFAV tus Atilia Q(uinti) l(iberta) FauSTA · PRIVATUS · ET sta Privatus et MARTIALIS · HERED Martialis hered(es) · L P · l(ibentes) p(osuerunt)”

Pliny the Elder indicates that the route was 600 Roman miles long (889 km), which corresponds to the reconstructed route by Wielowiejski at 860–890 km (Plin. HN. 37.11.45; Wielowiejksi 1980, 208). There were 40– 42 stage spots, and if one walked 21–22 km per day, it took six weeks of travel time one way (Wielowiejksi 1980, 208). It is interesting that the Roman-style structures found in Austria and Slovakia sit on Amber Routes. Niederleis and Oberleiserberg in Austria and Mušov in Moravia sit on a lesser path that started at Vindobona, while Stupava sits on the main path from Carnuntum, and Milanovce along the path starting at Brigetio (Wielowiejksi 1980, 208). The only site not identified with a trade route is Cífer-Pác, but it does sit strategically between Stupava and Milanovce; also the Váh River may have migrated eastward, away from the site as well, but this supposition would need to be verified.

(Kolnik 1977, 481; 2000a, 20; Pichlerová and Neumann 1979, 53) The third line, “mus inter(p)rex”, has been disputed. The insertion of ‘p’ was done by T. Kolník, but M. Pichlerova and A. Neumann would instead read the line as “mus interrex”, while K. Hošek instead reads the line as “mus inter tre [x] (cenarios)” (Kolník 1977b, 481; 2000a, 20; Pichlerová and Neumann 1979, 53).

33

The translation being, “Quintus Atilius Primus, son of Spurius of the Voturian tribe, interpreter (according to Kolník) or interrex (according to Pichlerová and Neumann) or among chosen soldiers (according to Hošek), centurion and provisions procurer of the 15th Legion, in the 80th year was laid to rest here. Quintus Atilius Cogitatus; Atilia Fausta, freedwoman of Quintius; Privatus, and Martialis as heirs willingly had this made.”

14.686; Dobiáš 175 and footnotes 32–35 from Chapter Four). 43. The evidence for the war and title comes from inscriptions and coins (Dobiáš 1964, 175 and footnotes 37 and 39 from Chapter Four). 44. The SHA does mention that around 108 Hadrian was praetorian legate of Pannonia Inferior, where he kept the Sarmatians in check during the principate of Trajan (SHA Had. 3.9). This statement may indicate actual campaigns, or it may just be that Hadrian kept the Iazyges under control through diplomacy. The Iazyges may have been causing problems because territory that had been captured by Decebelus, identified as the plain of Oltenia, was not returned by Trajan after his Dacian War, and instead it was incorporated into the new province of Dacia (Mócsy 1974, 94–95). It is possible that the Iazyges went to war in an attempt to recapture the territory.

If Kolník is correct, then Quintus was both interpreter and provisions procurer for the 15th Legion. The statement of interpreter would indicate dealings with individuals along the frontier who did not speak Latin, i.e. Germans, Celts, Dacians, or Iazyges and therefore most likely carried Quintus north of the Danube for the acquisition of supplies. If Pichlerová and Neumann are correct, then Quintus had been appointed an interim magistrate possibly of the canabae attached to the legion (Pichlerová and Neumann 1979, 55). They do acknowledge that under Augustus, the title of interrex was changed to praefectus, but that does not preclude the old title from being used in their opinion, especially in the legions or out in the provinces (Pichlerová and Neumann 1979, 55). C. R. Whittaker and M. Todd accept the reading of the inscription as interpreter (Whittaker 1994, 117; Todd 2004, 88); so, too, in AE (1978), 635, where numerous epigraphical examples are cited for interprex=interpres. 39. Italicus could have also been a nickname or alternate name for Vangio. If he had been a hostage raised in Rome or was known for enjoying Roman customs, then he could have earned the name Italicus. 40. As to how the equites got to the Lugii is an interesting question. The force could have marched around the Marcomanni through northern Germany or around the Quadi in the east, but the easiest would have been right up through either Marcomanni or Quadi territory, following the Amber Route (which went through Lugii territory). If the force had indeed marched through Marcomanni and Quadi territory then this act would have helped cause such an angry response from those two German groups. 41. Information on the dating of this campaign is provided through Martial’s poems (Mart. 9.32.3 and 8.8; Dobiáš, 1964, 175 and footnote 31 from Chapter Four). Dobiáš would also like to append to this campaign the line from Rufus Festus’s Brevarium, “Marcomanni et Quadi de locis Valeriae, quae sunt inter Danuvium et Dravum, pulsi sunt” (Festus Brev. 8; Dobiáš 1964, 175 and footnote 31 from Chapter Four). 42. The war had been quite serious apparently because while Domitian had celebrated a double triumph over the Dacians and Rhineland Chatti, he only accepted an ovatio over the Marcomanni, Quadi and Iazyges (Dobiáš 1964, 175, 371). Martial’s poems along with Statius’ Silvae and Silius Italicus’s Punica provide information (Mart. 8.65.8, 9.101.21; Stat. Silv. 4.1.14, 4.7.50; Sil. Pun.

34

Charts and Tables: Table 2.1. Archaeological Periods with Subdivisions for Slovakia and Only the Roman Period for Bohemia/Moravia Slovakia Period Pre La Tène Early La Tène Middle La Tène Late La Tène Final La Tène

Early Roman

Marcomannic War Late Roman

Subperiods A B1 B2 C1 C2 D1 D2 D3 B1a B1b B1c B2a B2b B2c B2/C1 C1a C1b C2 C3

Date 450–400 BC 400–300 300–250 250–200 200–150 150–100 100–50 50–10 BC AD 10–25/30 25/30–50/55 50/55–70/80 70/80–105/10 105/10–130 130–166 166–180 180–210 210–250/60 250/60–310 310–375

35

Bohemia/Moravia Subperiods

Date

A B1a

32/25–10/5 BC 10/5 BC – AD 20/30

B1b

20/30–50/70

B2a

50/70–100/120

B2b B2/C1

100/120–150/160 150/160–180/200

C1

180/200–250/260

C2 C3

250/260–300/320 300/320–380/400

Figures Chronology I: Hallstatt and La Tène Periods

Figure 2.1. Modified image from Romsauer 1981, 86: Fig. 1.

36

Chronology II: Roman and Great Migration Periods

Figure 2.2. Modified image from Kolník 1981, 114: Fig. 1.

37

Figure 2.3. Map of Slovakia showing the different regions.

Figure 2.4. Map of the Middle Danube region around 80 BC.

38

Figure 2.5. Map of the Middle Danube region around 40 BC.

Figure 2.6. Map of the Middle Danube region around 7 BC.

39

Figure 2.7. The dark gray area shows the extent of Maroboduus’ power in Central Europe around AD 6 (based on a map from Dobiáš 1964, 97).

Figure 2.8. Map of the Middle Danube showing the extent of Quadi territory by AD 50.

40

Figure 2.9. Map of the Middle Danube showing the main paths (gray dotted lines) of the Amber Route through Quadi territory (based on a map from Wielowiejski 1980, Mapa 1).

41

power, but also economic centers, controlling the manufacture of various things, serving as storehouses for foodstuffs, and as refuges for people from the surrounding countryside in time of strife (Kolník 1998b, 58).

Chapter 3 Archaeological Account of Southwest Slovakia: 100 BC–AD 166

Bratislava was one of the major centers in the area, and is thought to have been controlled by the Boii, having been founded some time during the Middle La Tène (Pieta and Zachar 1993, 145 and 147) (Fig. 3.1). Two parts have been identified: 1) a fortified acropolis (i.e., an oppidum), which now lies under the medieval castle, and was probably around 20 ha in size, and 2) a lower settlement located to the east of the acropolis in what is today the Old Town, probably covering around 31.71 ha (Pieta and Zachar 1993, 148). Evidence of fortification works have been found along with habitations, kilns, and manufacturing areas (Pieta and Zachar 1993, 149–157). Coins were also minted in Bratislava, such as the usual tetradrachms, but also larger hexadrachms; these coins had a lyre symbol sometimes appearing on the reverse (Pieta and Zachar 1993, 146; Kolníková 1998, 45). Fifteen Celtic names have been found on coins from Bratislava, such as AINORIX, BIATEC, BVSV, DEVIL, or NONNOS (Neustupný and Neustupný 1961, 152; Kolníková 1998, 45). They are thought to represent independent kings in the region, showing a fragmented Celtic political landscape (Dobiáš 1964, 348; Kolníková 1998, 45).2 The names seen most often both in Bratislava and further abroad, such as in Noricum or Pannonia, were NONNOS, MACCIVS, DEVILLIVS, and IANTVMARVS, but there are also some names seen only at Bratislava (Kolníková 1998, 45). In connection with the oppidum at Bratislava, there were important oppida established at Devín and Hainburg, which are all thought to have controlled the trade along the Amber Route in this area (Pieta and Plachá 2000, 6).

Chapter 3 provides a general introduction to the archaeological finds for the region of southwest Slovakia, with some reference to other parts of the country and Moravia. There are nine sections to this chapter. There is no recent archaeological account in English of the Middle Danube region with a focus on the Czech Republic and Slovakia.1 This chapter, therefore, provides detailed coverage of this topic. A large number of archaeological sites are mentioned in the text and three maps are used to show locations based on period and time. For La Tène sites in Slovakia refer to Figure 3.1; for La Tène sites in Moravia, use Figure 3.2; and for Roman Period sites in Slovakia and Moravia refer to Figure 3.4. 3.1. Final La Tène: 100–10 BC The Middle and Late La Tène periods (250–100 BC) were the most prosperous in southwest Slovakia for the Celts, showing an increasing concentration of settlement (Pieta 1981, 101–102). There was growing contact with the Mediterranean cultures, with an influx of better iron implements (imported Roman types or often copying Roman designs), along with the increasing usage of coins (Neustupný and Neustupný 1961, 146–147 and 152; Audouze and Büchsenschütz 1992, 231). It was during the Late La Tène that the Celts reached their greatest extent in Slovakia and one could say in the Middle Danube region as a whole before their quick decline as a major political power during the Final La Tène. Archaeologically, Celtic material by this period was spread from the region of Bratislava in the west to the banks of the Ipeľ River and beyond in the east.

Outside of the oppida, many smaller unfortified settlements have been found throughout Slovakia dating to the Late La Tène. They were usually located on sand terraces overlooking fertile lands near water sources along southern Slovakia (Kuzmová 1980, 313 and 314). In the Záhorie lowland (southwestern Slovakia), settlements were found on sand terraces with foreststeppe characteristics, and which also had high concentrations of bog-ore (Pieta and Zachar 1993, 146). No settlement has been excavated in its entirety and at many only a few huts have been examined (Benadik 1971, 482).3 Because of this situation, while the layout of Celtic settlements is unknown, some things can be said about the composition of settlements. The common hut type was the pit-house, dug into the ground up to 50 cm, usually 3m x 4m in size covering on average 13.78 m2, with two posts, one in the middle of each short-wall, while some huts also had a bench or shelf along one of the long-walls (see Chapter 7.1 for more information; Kuzmová 1980, 316 and 320; Kolník 1998b, 57). Various huts could have attached structures, while near the huts various other features have been found, such as animal stalls, storage pits, wells, kilns, work-areas, and ovens (Kuzmová 1980, 313 and 320; Kolník 1998b, 57).

At the start of LTD2 (100 BC), the widespread growth of economic-political centers continued in the southern regions of Slovakia, while in the lowlands, unfortified settlements continued along the river valleys leading into the Carpathians (Kuzmová 1980, 313). As seen in other parts of temperate Europe, the oppida had become important centers in Slovakia by the Late La Tène, with centrally located sites identified at places such as Bratislava, Devín, Plavecké Podhradie, ŠuranyNitriansky Hrádok, Divinka, Liptovská Mara and Zemplín (Kuzmová 1980, 315; Kolník 1998b, 58) (Fig. 3.1). The oppida are seen as evidence of growing social organization and complexity, with some scholars using the term urbanization, but being fortified; they also indicate possible instability in the societies (Neustupný and Neustupný 1961, 150; Shchukin 1989, 14; Audouze and Büchsenschütz 1992, 88; James 1993, 60–61; Pieta and Zachar 1993, 147; Cunliffe 1994, 420; Wells 1999, 31, 50–51). Oppida were the centers not only for political 42

In the earlier La Tène periods inhumation graves dominated, but by the second half of the third century cremation graves took over; new cemeteries appeared, such as those at Hurbanovo-Abadomb, HurbanovoBohatá, and Bodza-Holiare, but overall in terms of archaeological survey, few graves have been found dating to the Late and Final La Tène (Kuzmová 1980, 313; Kolník 1998b, 57) (Fig. 3.1). As time went on, urn-graves became shallower and grave goods decreased in amount, so that by the Final La Tène, very few items were deposited, which indicates more of a change in funerary practice than of material poverty (Kolník 1998b, 57).

Mountains and Slovak Ore Mountains, which were known for their rich veins of metal ores, such as iron, silver and copper, and known to have been accessed by the Celts. It appears that the Celts were expanding in the search for metals, following the rivers into the mountains (Pieta 1996, 25). Some of the primary centers so far identified seem to have been Liptovská Mara (the oppidum is called Havránok), Divinka and Podtureň, while settlements have been found throughout and other fortified centers have been found in the regions of Žilina, Orava, Liptova and Spiš (Pieta 1996, 26 and 29) (Fig. 3.1). The settlement patterns follow that of the Celts further to the south, in that there were oppida around which were various types of open settlements either of agricultural or manufacture character (Pieta 1981, 106). The major difference is the use of two-roomed post-built huts above ground instead of the pit-houses seen in the south (Pieta 1981, 106). The oppida were fortified with walls made of dry-laid stone and/or palisades, usually having one main fortified gate; with the entire site surrounded by systems of ditches and embankments (Pieta 1996, 69). Púchov sites have been found in northern Moravia and southern Poland as far as Silesia (Pieta 1996, 26). In the Púchov region, coins appear there with CAT minted on them, taken to be a shortened form of the people’s name and possibly referring to the Cotini (Kolníková 1998, 47).

For pottery forms seen during the Late and Final La Tène, see below (and Chapter 9.4.1). In regards to basic types, I note here graphite, gray or gray-brown, and painted vessels (Kuzmová 1980, 324). The graphite vessels have the mineral mixed in with the clay, creating a lustrous appearance; it is most likely that locally obtainable graphite was used, coming from the Lesser Carpathian Mountains (Kuzmová 1980, 324). These graphite vessels were apparently held in high regard over other pottery, because many show signs of repair work, with holes for iron and lead clamps or hoops (Kuzmová 1980, 324– 325). The gray and gray-brown vessels are found most often in settlements and have a wide range of shapes made on the wheel (Kuzmová 1980, 327). The painted vessels are a characteristic type of the Late La Tène, using red and white colors for the slips, placed on thinwalled vessels: these are a rare find (Kuzmová 1980, 330). For the Late La Tène, the most commonly found fibulae include wire fibulae with figure-eight twists on the bow, fibulae with large disc knobs on the catch-plate, disc-shaped fibulae, clip-form fibulae without segmented bow, double-knob fibulae, and Nauheim type fibulae (Kuzmová 1980, 332). In general, the Celts produced high-quality metal items, especially ornamental jewelry, such as the previously mentioned fibulae, along with bracelets, torques, rings and belts with ornaments and fastenings; they also created a wide assortment of day-today items and specialized tools for the manufacture of bronze and iron items along wth wood-working tools (Kolník 1998b, 59).

During the Final La Tène, when the ancient sources mention the conflict between Celts and Dacians, the archaeological record shows a steady infiltration of Dacian material into Celtic lands (Kuzmová 1980, 334; Kolník 1991, 432). Two phases have been identified, with the first occurring 100–60 BC and the second, 60 BC–AD 50 (Benadik 1971, 485; Kuzmová 1980, 334). During the first phase, there is a decline in the number of identified Celtic cemeteries and Dacian pottery starts to appear at various sites throughout the southern and southeastern parts of Slovakia, identified as thick walled, handmade vessels, primarily cups and pots sometimes decorated with finger impressions in a line and with incising on the rims (Benadik 1971, 485; Kuzmová 1980, 330). The second phase is primarily seen as a period in which southern Slovakia, east of the Váh River, has Celto-Dacian settlements, having characteristics of both groups (Kuzmová 1980, 330). The pottery of the second phase shares characteristics with pottery seen throughout Dacian controlled territory, being both hand and wheel made, mixed with the Celtic pottery of the period (Benadik 1971, 330; Pieta 1980, 102).

In regards to imports, finds from the oppidum at Bratislava show fine ware from the Roman territory such as black sigillata plates and trays from Campania, and two handled amphora-like vessels of well levigated yellow-brown clay; there was also a variety of bronze vessels such as strainers and bowls of the Eggers 91 type (Pieta and Zachar 1993, 177–179).

In the Final La Tène, the oppidum at Bratislava started to mint coins that were Dacian in design, imitating those from the island of Thasos, both in metal content and size (Kolníková 1998, 46). When Burebista was victorious over the Celts and expanded Dacian control all the way to the Morava River, the oppidum at Bratislava ceased to mint coins and was most likely destroyed as well, while the settlement continued in a reduced fashion to the end of the first century BC (Pieta 1981, 102 and 190; Kolníková 1998, 46).5 The oppidum at Devín then became the main center for the area, taking over the

In north-central Slovakia, the Hallstatt period Lusatian culture survived into the Early La Tène, but by the LTB2 and LTC1, La Tène elements started to appear, usually termed Pre-Púchov. The Celtic culture that appeared in the region by the LTC2 is commonly known as the Púchov culture, based on the finds first identified at the site of Púchov and there is growing evidence to link them with the Cotini mentioned in ancient sources (Pieta 1981, 104; 1996, 26, 27, and 35; Kolník 1991, 432).4 The culture was situated in what are called the Lesser Tatra 43

position that Bratislava had held; it is thought Devín came under the control of Noric tribes that had advanced into the area (Pieta 1981, 102; Pieta and Zachar 1993, 193; Pieta and Plachá 2000, 6). Supporting evidence is the fact that, after the death of Burebista and the retreat of Dacian control, Noric coins appear in greater numbers in southwest Slovakia, with the Karstein type actually minted in the area, lasting until the last two decades of the first century BC. What is more, Noric material starts to appear at Devín (Kolníková 1998, 46; Pieta and Plachá 2000, 6).

portion. For Moravia, the southern portion sees a parallel development to what was occurring in Lower Austria and southwest Slovakia (Droberjar 1997, 134). The northern area of Moravia sees an increase in elements of the Przeworsk Culture.6 In fact the earliest German material in Moravia starts to appear in the LTC2 period from the Przeworsk Culture, made up of ceramics found at the settlement of Pravčice (Droberjar 1997, 135). This incursion of Przeworsk material increases by the LTD1 phase and is seen in other settlements such as Rymice and Bořitov, as well as at the highland settlements of Staré Hradisko and Kojetín Požaha (Droberjar 1997, 135) (Fig. 3.2). Southern Moravia has no Przeworsk from this phase, but instead there is evidence for some presence from central Germany and the Elbe (Droberjar 1997, 135). At the site of Vyškov, which was a settlement in the Roman Period, a feature was found dating to the Late La Tène that had two pots of the Przeworsk culture and two pots of the Elbe Germans (i.e., the Marcomanni and Quadi), showing a mixing of the two groups in the area, which was at the southern border of Przeworsk expansion (Drobjerar 1997, 135) (Fig. 3.2).

In south-central Slovakia, the oppidum at Nitriansky Hrádok continued to exist, showing a mixture of Dacian and Celtic elements (Benadik 1971, 486; Pieta 1981, 102 and 104). During this period at Nitra, the mint there continued to operate, producing tetradrachms imitating Republican Roman denarii having dogs or horses on them, in the style seen on coins minted by Gaius Postumus, Titus Carisius, and Marcus Volteius (70–40 BC) (Kolníková 1998, 46). The site persisted until the end of the first century BC (Pieta 1981, 104). There is archaeological evidence to show a shift of Celtic settlements northward, into northern Slovakia, away from the advancing Dacian colonization during the Final La Tène period (Pieta 1981, 104). During the LTD2 period, there is evidence of work to reinforce the defensive systems at Liptovská Mara and Velínku pri Podturni, while the sites of Púchov and Divinka lack LTD2 period material and may have been partially abandoned (Pieta 1996, 29).

3.2. The B1 Period (AD 10–70/80) The B1 period for all three regions of Slovakia is a period of change in the archaeological record (Fig. 3.3). As already mentioned in section 3.1, German material culture starts to appear in southwest Slovakia in the early part of the B1a phase, c. AD 10–15, even though historical accounts indicate that the Germans did not cross the Morava River until slightly later, when Vannius was given the territory sometime around AD 21 (Kolník 1971, 511).7 In north central Slovakia there is evidence that the Celts there came into trouble not only with advancing Germans from the east, but also the Przeworsk culture moving south from Poland.8 For the parts of eastern Slovakia, where there was the Celto-Dacian horizon, changes were brought about by the arrival of first warriors and then settlers of the Przeworsk Culture (Kolník 1981, 113; Lamiová-Schmiedlová 1977, 205).

In eastern Slovakia, with the penetration of the Dacians, a Celtic-Dacian amalgamation at the site of Zemplín helped to make it a large trading center in the Final La Tène (Pieta 1996, 27) (Fig. 3.1). The site had earlier Celtic beginnings in the second century BC and is located on an elevated spit of land that sits not far from the confluence of the Boldog with the Ondava and Latorica Rivers (Benadik 1971, 484; Lamiová-Schmiedlová 1977, 203). The site was fortified, covering an area of 250 x 150 m, with a defended gate on the south side (Benadik 1971, 484). The sides facing land had stone walls with earth fill, while the side facing the river had a wooden palisade (Benadik 1971, 484). Evidence for its commercial importance includes the discovery of a large-scale production facility for the creation of high-quality Celticstyle painted wares that were imported to neighboring regions (Kolník 1998b, 58–59). Along with Zemplín, Dacian material has been found at places such as Streda nad Bodrogom, Prešov, Šebastovce, and others (Lamiová-Schmiedlová 1977, 203) (Fig. 3.1).

It is in the B1a that German material starts to appear in cemeteries, such as Sládkovičovo (graves 34, 64, and 69), Abrahám (graves 4, 40, 80, 81, 86, 100, and 138), and Kostolná pri Dunaji (graves 3, 10 and 35) (Kolník 1971, 511 and 513) (Fig. 3.4). These graves are marked by the presence of large bronze vessels of the Eggers type 131 and fibulae of the Almgren 45, 47/48, 67 and 67/68 types (Kolník 1971, 513).9 All three cemeteries are within 20 km of each other in the area of Galanta and have their start in the first century AD, B1a phase, with only Abrahám having graves dating to the Late Roman Period as well (Kolník 1971, 511; 1977a, 144; 1980, 8 and 13). Sládkovičovo has 88 graves, of which only six were inhumation and the remainder cremation; Abrahám has 226 graves, of which 12 were inhumation; and Kostolná pri Dunaji has 67 graves, only two being inhumation (Kraskovská 1978, 4, 7, and 10–11; Kolník 1980, 11, 13, 21, 91, 126 and 129).10 The cremation graves are marked by the use of large handmade German vessels.

In contrast to Slovakia, the archaeological data in Moravia shows that the Celtic settlements seen in the lowlands do not go past the LTD1 phase (Droberjar 1997, 134). During the LTD2, the only Celtic occupation can be seen in a few highland fortifications built by the Púchov Culture (Droberjar 1997, 134). Moravia appears to have a similar situation as Central Slovakia, where the northern portion sees a different development than the southern 44

Inhumation graves are rare finds, but usually have a rich assortment of Roman imports, especially those seen at Zohor and Vysoká pri Morave in the B1c phase (see below; Kolník 1971, 510; 1981, 116).

Abrahám (Kolník 1971, 514). There is also evidence of contact with the Upper Rhine, with the discovery of an Almgren type 24 fibula from a grave at Abrahám (Kolník 1971, 517; 1977a, 152). Also dating to this phase are the two green-glazed skyphoi and one green-glazed kantheros from the cemeteries of Abrahám, Kostolná pri Dunaji and Sládkovičovo (see Chapter 9.1.12 for information on imported glazed wares; Kolník 1971, 515 and 517). In terms of bronze vessels found in cemeteries during this phase, a majority are of Eggers types 159, 160 and 162 (Kolník 1971, 517). Weapons also start to appear in graves more frequently, which is interesting since in the B1a phase almost no weapons were found in graves (Kolník 1971, 517; 1977a, 149).

The graves from the B1a phase share similarities with the Dobřichov-Pičhora phase in the Czech Republic and is identified with the settlements of the Marcommani and the period of Maroboduus (Kolník 1971, 514; 1977a, 149). According to T. Kolník, the oldest material from the cemeteries at Kostolná pri Dunaji, Sládkovičovo and Abrahám are parallel in type to the materials found in the last-phase graves at Tišice and Dobřichov-Pičhora in the Czech Republic and so verifies the ancient sources in the movement of Germans from Marcomannic lands into territories east of the Morava River in this period (Kolník 1971, 519–520; 1980, 8) (Fig. 3.4). It is also of interest that many of the earliest German graves, dating to the B1a phase (the early part of Vannius’ kingdom), have Noric material, which Kolník believes may indicate that the followers of Maroboduus and Catualda were housed for a period in the province of Noricum before being settled in southwest Slovakia (Kolník 1971, 520).

A total of 356 fragmentary or whole bronze vessels have been found in southwest Slovakia, with a majority of them coming from cemeteries; of the total amount 70% of the vessels come from just three cemeteries: Abrahám, Kostolná pri Dunaji and Sládkovičovo (Krekovič 1997a, 233). These three cemeteries have a high amount of imports overall, with Abrahám having a total of 150 items, Kostolná pri Dunaji 115, and Sládkovičovo 101 (Krekovič 1997a, 234). The distribution was not even, however, with only 15.7% of the graves at Abrahám, 55.9% at Kostolná pri Dunaji, and 34.1% at Sládkovičovo having Roman imports (Krekovič 1997a, 234).

In Moravia, there is little settlement data to help identify German development during this period (Droberjar 1997, 136). The oldest German material in a settlement for this period occurs at Nejdek, where a Late La Tène Przeworsk pot of the Liana II/1 type was found with fibulae of the Almgren 19 type and Hachmann II37, which go from the Late La Tène to the B1a phase in the Roman Period (Droberjar 1997, 136) (Fig. 3.4).

The earliest German settlement found so far in southwest Slovakia is on a hillside at Bratislava-Dúbravka and dates to the B1b phase (Elschek 1995, 39) (Figs. 2.11 and 2.12). The site itself is around two hectares in size and sits within a valley on a Pleistocene terrace near a stream. Surrounding the site on three sides are higher hills and so it is only open to the north, thereby having an unopposed view of the entire Morava Valley (Kolník et al 1993, 250; Elschek 1995, 40). The site is four km from the confluence of the Danube and Morava Rivers and 10 km south of Stupava. An oval pit (feature 22/88), fitting within Kolník’s Type I huts (see Chapter 7.1), was found with material dating to the second quarter of the first century AD (B1b period) (Fig. 3.5). The primary datable material was a fragment of a vessel known as a “Netzwerkschale,” which has been found also at Carnuntum, Magdalensberg, and other sites, while in Pannonia these vessels primarily date to the Claudian period (Elschek 1995, 40). There were also fragments of a Dacian vessel that dated to the B1b period as well (Elschek 1995, 40). A second pit-house (feature 13/88), with two-post construction (Kolník Type II), may date to the same time or slightly later than feature 22/88, based on associated provincial pottery and a bronze fitting with two rivets (Elschek 1995, 40) (Fig. 3.5). Three other features, two pit-houses (features 129/93 and 74/89) of Kolník’s Type III and a pit (feature 10/88) were found to date to the B1c and B2a phases (Elschek 1995, 40–41) (Fig. 3.5).

The following two phases, B1b and B1c in Slovakia and Moravia, show a marked increase in both the number of burials found at pre-established cemeteries and new cemeteries, as well as in the discovery of isolated graves and the identification of settlements (Kolník 1981, 116; Elschek 1995, 39; Droberjar 1997, 139). These two periods fall within the time of Vannius’ reign as king in southwest Slovakia (and probably southeast Moravia). The primary region in Slovakia that shows German expansion is the area south of Trnava, but also in areas slightly to the north, as at Vlčkovce, Križovany nad Dudváhom and Žlkovce, and further afield, to either side of the Morava River, at the sites of Mistelbach, Altenmarkt im Thale, and Láb (Kolník 1971, 514; 1977, 149) (Fig. 3.4). In terms of material culture, there is an increase in the variety and amounts of Roman imports during the two phases, which continues in the following B2 phase (Kolník 1981, 116). For a detailed discussion of Roman imports, see Chapter 9 for Roman Pottery and Chapter 10 for other types of goods. In regards to German pottery, see Chapter 8 for detailed discussion of the various forms and decorations. For the B1b phase, grave goods continue to have similar items, such as fibulae of the Almgren 45/47 and 67/68 types, but also there is the growing presence of the Almgren 68 type (Kolník 1971, 514). At the same time, imported noric-pannonian fibulae (Almgren 236 and 237) occur in graves at Kostolná pri Dunaji, Sládkovičovo, and

The oldest identified Elbe-German remains in Moravia also date to either the B1b or B1c period at the site of Hrabětice (Droberjar 1997, 136) (Fig. 3.4). A fragment of 45

a vessel was found identified as a Brandenburg A1b type, decorated with meanders made by a spiked wheel, supplemented by triangles, which are identified as belonging to the B1b or B1c phases. There was also an isolated grave found at Nedakonice (grave 8) which could be late B1a or early B1b, but also cremation graves at Bzenec and Velatice dated to the B1b period (Droberjar 1997, 136) (Fig. 3.4).

Three elite graves (nos. 3, 4, and 5) found at Zohor in 1957 date to this period, c. AD 70 (so they could be considered part of the B2a phase) (Kolník 1971, 519; Krekovič 2000a, 49) (Fig. 3.4). They and other graves were discovered in a gravel quarry by workers and the graves were badly disturbed by the time archaeologists arrived on the scene (Krekovič 2000a, 49). The three elite graves were buried 1.5–2 m below the surface and were close together, probably signifying some kind of relationship to each other (Krekovič 2000a, 49). Two of the graves had similar inventory and are thought to be contemporaneous, containing bronze vessels, such as ladles, strainers, pans and buckets; one strainer had the name ILLIOMARVS on the handle (Krekovič 2000a, 49). Glass vessels were found as well, including a blue glass bowl and two shallow bowls of blue and white mosaic glass from Grave 3 and a green glass bowl from Grave 4 (Krekovič 2000a, 49). Grave 5 was the richest with eight bronze items, including a ladle and strainer, two bowls, two pitchers and two pans, along with a gold bowl, two silver fibulae, a gold armband and a bronze knife (see Chapter 9.1 for more information; Krekovič 2000a, 49).

The B1c phase sees the widespread appearance of the Pannonian-trumpet fibula (Hull type 162 and variants of Almgren 74, 99, and 102), along with the variant of Almgren 68 that has a lengthened form, tri-segmented bow, and perforated catch-plate (Kolník 1971, 518). The Almgren types 50–51 continue to be used, while the Almgren 57 type makes an appearance in this phase, but really becomes widespread in the B2 phase; one also sees the appearance of the Hofheim type fibula and other ornate provincial imports, some with niello (Kolník 1971, 518–519; 1977, 159). The further intensification in the amount of Roman imports into southwest Slovakia can be seen in the utensils and serving sets for the consumption of wine found in German graves, which is seen as an identifier of the German elite’s adoption of the GrecoRoman customs of consuming wine (Kolník 1971, 519).11

In Central Slovakia, many Púchov sites show destruction layers dating to the B1b period, in the areas of Žilina, Orava, Liptov, and Spiš (Pieta 1996, 29). The destruction includes ruined fortifications and burned houses with insitu household goods (Pieta 1996, 29). At the site of Liptovská Mara, the oppidum Havránok was rebuilt but along simpler lines with only certain parts redone, while out of the eighteen known settlements in the nearby region, only three appear to have been reoccupied; by the middle of the first century there were a total of seven settlements (Pieta 1996, 30 and 31) (Fig. 3.4). In the region of the Púchov culture as a whole, the reoccupied settlements were smaller and denser (Pieta 1996, 31). The material culture of these rebuilt settlements is a mixture of Celtic, Dacian, Przeworsk, and German materials, creating a complicated picture as to what might have been happening in the area (Pieta 1996, 31). There is also evidence for the intensification of ore extraction and processing in the region, with smelting furnaces found at Spišské Tomášovce-Čingov, and Hrabušice and the mining of metals from hematite deposits in the area of Varín and Nezbudská Lúčka, east of Žilina (Pieta 1996, 31) (Fig. 3.4).

The site of Devín itself appears to have been incorporated into the Limes Romanus for a brief period at least under Augustus, because it sits at the confluence of the Morava and Danube Rivers, not far from Carnuntum (Plachá, Hlavicová, and Keller 1990, 15 and 16; Kolník et al. 1993, 243 and 245) (Fig. 3.4). Devín has the earliest Roman material found in southwest Slovakia, with Augustan period terra sigillata and wine amphorae (unique to the Middle Danube) as well as a silver denarius of the late Republic (Plachá, Hlavicová, and Keller 1990, 15). Recent excavations under and near a later third-century structure (Building I) at Devín have yielded possible evidence for this late Augustan occupation. On top of the older La Tène strata was a stratum containing the ruins of a structure built of wattle-and-daub and filled with late Augustan artifacts (Pieta and Plachá 2000, 6). Nearby were the remains of what has been identified as a watchtower, which also had Augustan period artifacts, including terra sigillata and fibulae; all of these artifacts and remains point to a possible presence between the last decade of the first century BC and the first two decades of the first century AD (Pieta and Plachá 2000, 6–7). If there had been an actual Roman occupation of the area, it appears to have been pulled out by the 20s and 30s AD, because the Celtic settlement at Devín was destroyed by the Quadi at that time and around AD 50 a German settlement appears somewhere in the area (Kolník et al 1993, 245). German material appears on the site and nearby on the south slope of Devínska Kobyla, two German female urn-graves with various bronze goods were found (Plachá, Hlavicová, and Keller 1990, 16; Kolník et al 1993, 245, 258–259).

As to who is responsible for the destruction layers, the Quadi are good candidates, since it was precisely in this period under Vannius that there was German expansion eastward into Slovakia, and what is more, Tacitus mentions that the Cotini paid tribute to the Quadi (Pieta 1996, 30). There are other possibilities, however, because, as mentioned in previous sections, the Przeworsk culture was also spreading southward, with appearances in northern Moravia as well as in northern Slovakia. We know the Marcomanni and Quadi had problems with the Lugii who lived to the north, especially in the period of Domitian. The Dacians could also be responsible, because the archaeological material shows their expansion into northern Slovakia from the south 46

during the B1 period, which may have been a response to the arrival of the Sarmatian Iazyges along the Danube (Pieta 1996, 30).

3.3. Description of Quadi Burial Practice As is readily apparent from the above discussion, the best evidence for German settlement of southwest Slovakia in the B1 period comes from graves. Even though some description has been given above, the following section provides a brief overview of Quadi burial practice.

In eastern Slovakia during the B1 period, the CeltoDacian Culture continues for the first half of the century. A cremation cemetery from Zemplín dates to this period, as well as settlements at Šebastovce and Prešov (Kolník 1971, 524) (Fig. 3.4). The pottery displays a mixing of continuing Celtic elements, such as bowls and containers made on the wheel, and Dacian, such as handmade sprofiled pots decorated with horizontal raised bands punctured by thumb-impressions (Kolník 1971, 524– 525). There is contact with the provinces as seen through the appearance of Roman-provincial material, such as fibulae of the Almgren 67 and 68 types, and noricpannonian belts. Around the middle of the first century, urn graves of the Przeworsk type appear at Zemplín, along with new types of pottery such as sharp-profiled vessels with feet (Kolník 1971, 525; LamiováSchmiedlová 1977, 205). Elements of the Przeworsk culture not only appeared in eastern Slovakia, but also in northeastern Hungary and northwestern Romania (Kolník 1981, 122). According to the evidence available, the first Przeworsk incursions were done by warrior groups, followed by cattle-breeding farmers (LamiováSchmiedlová 1977, 205).

Cremation was the dominant burial method, with the deposition of ashes and remaining bones into an urn, which was then placed into a pit, some being quite shallow, or the remains with grave goods were placed into a pit without an urn (Kolník 1971, 509; 1998, 71). Inhumation graves were present, but in very small numbers, as already seen in the grave types mentioned above for Kostolná pri Dunaji, Abrahám, and Sládkovičovo. Grave goods were placed with both types of burials. For urn-burials, the grave goods could be placed inside the urn, on top of the bones, or beside the urn, but very rarely under the urn; the material almost always showed damage from fire or deformities and breaks done on purpose (which is primarily seen in Early Roman Period graves, but not in inhumation graves) (Kolník 1971, 509–510). The inhumation graves found in the large cemeteries usually had the bodies in a scrunched position and the axis aligned east-west, with the head usually pointed west (Kolník 1971, 510; 1998, 72). Elite burials appear usually as inhumation, sometimes isolated from cemeteries for lower classes; this situation also appears in other German territory (Kossack 1981, 312). We see this at Zohor and Vysoká pri Morave for example.

There is very little to no cultural material from the Sarmatian Iazyges in eastern Slovakia (LamiováSchmiedlová 1977, 206). While it is clear the Iazyges occupied the territory between the Danube and Tisza Rivers and for a brief period had also encroached into southern Slovakia, it is unclear as to how far north their territory reached in modern-day Hungary. Based on archaeological evidence, there has been speculation that the Iazyges split the Dacians, with one group in Romania and the other occupying East Slovakia, Galicia (part of southeast Poland and western Ukraine), and the Carpathian Ukraine (former Ruthenia) (LamiováSchmiedlová 1977, 206). The frontier between the northern Dacians and the Iazyges would have been along the Bodrog River (Lamiová-Schmiedlová 1977, 206) (Fig. 2.8).

After the Marcomannic Wars, there is a decrease in the amount of grave-goods placed with the burials and a focus instead on the appearance or type of urn used (Kolník 1971, 510). The range of vessels used broadens to include ceramic vessels that originally had other functions, bronze vessels, and even one example of an alabaster vessel (Kolník 1971, 510). By the fourth century, there are few grave-goods present (Kolník 1971, 510). By the Late Roman Period, inhumation graves are usually found alone or in small groups separate from cremation cemeteries (Kolník 1971, 510). It is interesting to note that while in the Roman world inhumation graves start to become more common in the third and are dominant by the fourth century AD, in the barbaricum, cremation urn-graves were still the most common (Kolník 1998b, 71–72).

As seen with the graves of the B1c period in southwest Slovakia, similar developments are also known in Moravia. Settlements at Brno-Chrlice and Komořany have terra sigillata from south-Gaul of the La Graufesenque type, primarily Drag. 29 bowls, which date to the second third of the first century AD (Droberjar 1997, 139) (Fig. 3.4). Bronze containers also start to appear in Moravia during this phase (Droberjar 1997, 139). There is also the appearance of the Almgren 68 type fibula, having two holes in the catch-plate, at various settlements such as Uherský Brod and Dobročkovice; this type of fibula was common in the second half of the first century AD in Noricum, Pannonia, and the barbaricum (Droberjar 1997, 139) (Fig. 3.4).

In eastern Slovakia, there is a mixture of burial practices. Urn graves appear in cemeteries and as isolated examples, but also burials are made using tumuli or in the fill of previous built tumuli (Kolník 1998b, 72). Examples of this are found at Streda nad Bodrogom and Zemplín; this burial rite is seen as a Dacian influence (Kolník 1998b, 72) (Fig. 3.4). At the same time, urn graves of warriors are also found at places such as Lesné, Kvakovce, and Zemplín that show influences from, or are part of, the Przeworsk culture (Kolník 1998b, 73) (Fig. 3.4).

47

either in terms of huts or hectares covered, include Branč, Komjatice, Močenok-Sládečkovce, Nitra-Janíkovce, Nitra-Párovské Háje, Chotín, Pobedim, Čaka, and others (Kolník 1998b, 70) (Fig. 3.4). It is uncertain if some of these settlements were occupied year round or cyclically (Kolník 1971, 508; 1998a, 155; 1998b, 70). It does appear that after huts were abandoned, new huts were not built on top, but instead nearby, causing settlements to move progressively across the landscape, thereby extending the settlement area.

3.4. Description of Quadi Settlements In regards to settlements, the archaeological evidence shows that when the Germans first arrived in southwest Slovakia, they did not live in large-scale settlements, but instead small groups of huts and isolated farmsteads; over time there was a process of agglomeration so that by the Late Roman Period one sees denser settlements, such as at Pobedim and Branč (Kolník 1971, 508; 1998b, 68). The majority of German habitations were situated near water sources and fertile land (Kolník 1998a, 154; 1998b, 68). There is no evidence that Germans had oppida, forts, or even palisades around villages, while in contrast during this period these types of architecture continued in the areas of the Púchov Culture (Kolník 1998a, 154; 1998b, 68–69). By the end of the Roman Period and at the beginning of the Great Migration Period, one starts to see German settlements and refugee areas placed on higher terrain that had more of a defensive appearance (Kolník 1998a, 154; 1998b, 69).

The archaeological appearance of German settlements for southwest Slovakia and southeast Moravia accords well with Tacitus’ description of how the Germans built their homes. He describes Germans as “[living] separated and scattered, as a spring or a field or a grove has attracted them. They do not plan their villages in our manner with buildings joined together and next to one another: each one has an open area around his home, whether as a protection against the disasters of fire or because of a lack of skill in building. They make no use at all of stones or tiles: they use unshaped timber for everything without regard to appearance or aesthetic pleasure. Certain parts they coat with greater care with an earth so pure and gleaming that it looks like painting and colored drawings. They are also much accustomed to dig underground chambers and they cover them with much dung; these serve as a retreat against winter and a storage area for crops…” (Tac. Germ. 16; translation by H. Benario 1999, 29 and 31). It is within all likelihood that Tacitus based his description of German habitation on the Marcomanni and Quadi, with whom the Romans had had close dealings, as already related in section 2.3.2.

For the Roman Period, isolated huts, small groups of huts and settlements have been found on sand terraces near rivers and streams (such as Branč, Milanovce, Nitriansky Hrádok, Bánov, Marušová, Šarovce, Phronský Ruskov, Bíňa, and Štúrovo); in plains on top of rises near rivers and streams (such as Pobedim, Pác, Šaľa, MočenokSládečkovce, Tvrdošovce, Komjatice, Ondrochov, Andovce, and Bohatá); or they were located on the southwest, south, or southeast slopes of foot-hills near water sources (such as Vráble, Čifáre, Čaka, and Plavé Vozokany) (Kolník 1971, 508; 1998a, 154) (Fig. 3.4). All of these were located in areas of highly fertile loess deposits (Kolník 1971, 508; 1998a, 154). From the size of the isolated cemeteries that have been found, such as Kostolná pri Dunaji, Abrahám, Sládkovičovo, Bešeňov, or Očkov, there were large settlements somewhere nearby (Fig. 3.4).12 Besides these, there are other smaller cemeteries that may indicate small settlements or homesteads nearby.13

In the region around Stupava-Urbárske Sedliská, a large majority of the settlement sites found so far have only one to three Grubenhäuser, usually of the third type.14 Some sites have multiple houses and may in fact be very large settlements, but further archaeological survey is required to determine the actual size. Burials near Stupava have been discovered at Bratislava-Devínska Nová Ves, Vysoká pri Morave, Malacky, Láb, and Zohor as well as single-grave sites and small cemetery sites (Kováč et al. 1998: 70) (Fig. 3.4).

Two types of structures are found at settlements: pithouses and above-ground structures, both of which had multiple functions, either serving as dwellings or specialized work areas (Kolník 1971, 508–509; 1998a, 144). Very few above-ground structures have been discovered in Slovakia, especially the long-house, which is commonly seen in Germany (Kolník 1998a, 145; 1998b, 69). For more information on German house architecture see Chapter 7.1. The German house did not stand alone, but could have nearby animal pens, fire-pits for food preparation, ovens, kilns, and storage pits (Kolník 1998a, 153–154; 1998b, 70). In settlements, German huts were spread out with usually five meters or more space between them. Almost all the huts would be oriented with the entrance facing toward the south, southwest, or southeast. In terms of settlement size, the isolated huts and small groups would have occupied only a small area (less than 0.5 ha), but the large settlements show evidence of having many tens to hundreds of huts spread out over large areas, covering tens of hectares (Kolník 1998a, 155; 1998b, 70). The largest settlements,

3.5. The B2 Phase (AD 70/80–166) There is continued prosperity and German expansion during this period in Slovakia, with increasing population (Kolník 1980, 9). The cemetery at Kostolná pri Dunaj that was in use during the B1 phase continues (Abrahám skips the B2 and has burials of the Late Roman Period; Sládkovičovo was in use only in the B1a and B1b phases), while various isolated graves and new cemeteries date to the B2, such as the cemetery at Bešeňov (Kolník 1961; Kraskovská 1978, 10). Many of the settlements that have been archaeologically investigated date to this period (Kolník 1971, 522). There is also an increasing presence of the Przeworsk culture whether through material imports coming into Quadi territory or actual Przeworsk

48

settlers in some areas, such as in northern and eastern Slovakia (Kolník 1971, 524; 1980, 9; Pieta 1996, 30). The most commonly found fibula in burials and at settlements is the Almgren 68 and its variants, along with the Almgren 57 already mentioned above (Kolník 1971, 522). During the B2a, the Almgren 68 has a perforated catch-plate and the end-knob bent upward; in the B2b, the fibula has a lengthened bow and the catch-plate is larger (Kolník 1971, 523–524).

The second grave is from Zohor, where three B1 period elite graves had been found and discussed above (Fig. 3.4).15 This second grave was found by accident in the same gravel quarry where the earlier graves were found; it was an inhumation, with the body spread out, lying northwest-southeast, with the head turned facing to the northwest (Ondrouch 1957, 40). Only three items were found, which were again placed around the head; the workers claimed that these were the only goods in the grave (Ondrouch 1957, 40). There were two pans and one rusted belt-buckle that fell apart during removal. The first pan is a Gödåcker type (Eggers 144) and the second pan is an Eggers 140, dating the grave to the middle of the second century AD (Ondrouch 1957, 42, 44, 47, 68; Kraskovská 1978, 13).

The cemetery found at Bešeňov contained 113 cremation graves, of which 106 were urn burials and seven were in pits (Kolník 1961, 262; Kraskovská 1978, 4) (Fig. 3.4). The earliest graves date to the beginning of the second century and the cemetery was in use until the fourth century (Kraskovská 1978, 5). A total of 27 fibulae were recovered from 19 graves, with 18 being bronze and nine iron (Kolník 1961, 250). The types of fibulae present included iron Pannonian trumpet fibulae, iron fibulae with cylindrical heads (part of Almgren group V, types 120–131), a knee-fibula (similar to Almgren 145), bronze crossbow fibulae (Almgren group VII, similar to type 193), iron one-piece fibulae, and fibulae similar to Almgren type 158, with spiraling on the bow before the catch-plate (Kolník 1961, 245–250). Three graves had bronze buckets and ladles, including two Eggers type 40 bronze buckets and two Eggers type 160 ladles and strainers (Kraskovská 1978, 5). There was some Roman provincial pottery, but in small fragments, including Pannonian Gray Ware and Brick-red Ware, along with one fragment of Rheinzabern terra sigillata (Kolník 1961, 262). The German pottery present was all hand made for the Early Roman Period, while some wheel-made German pottery was found in the Late Roman period graves (Kolník 1961, 258–262).

The third grave is slightly different, being a cremation urn grave found at Borský Peter (formerly Borský Svätý Peter) (Ondrouch 1957, 47) (Fig. 3.4). The grave was discovered by a farmer at a depth of around 70 cm; the grave contained burnt bone, along with two terra-sigillata bowls and part of a thin-walled bronze bowl (Ondrouch 1957, 47). Both terra-sigillata bowls are of Gaulish manufacture, possibly Lezoux; one bowl has on the register a scene of wild animals, while the other has the register divided into panels in which are figures, such as a hunter or bestiarius, a lion within a medallion, a rabbit and dog (Ondrouch 1957, 48, 50–51). The thin-walled bronze bowl fragment was too small for proper identification, but it had similarities to a similar bowl from the B2 grave at Vysoká pri Morave (Ondrouch 1957, 53). Based on the material, the grave probably dated to the middle of the second century (Ondrouch 1957, 69). As already mentioned before, many of the German settlements discovered in southwest Slovakia either have phases dating to the B2 period or solely date to this phase. For a basic idea of what these settlements were like, I provide a brief description of four sites: Branč, Láb, Veľký Meder, and Bratislava-Trnávka “Zadné” (Fig. 3.4).

Three elite burials have been discovered that date to the B2 period, with two close to each other in the Záhorie Lowlands and the third on the other side of the Carpathians in the region to the northwest of Trnava, near Piešťany. The first burial was discovered at Vysoká pri Morave and was an inhumation grave that was found 120 cm below the surface (Ondrouch 1957, 13) (Fig. 3.4). The grave was found by accident and damaged before the arrival of an archaeologist, with the body being removed and lost. A variety of grave goods were found arranged around the head, five cm above the body, and date the grave to the end of the first century or beginning of the second (Ondrouch 1957, 13, 68; Krekovič 2000a, 49). The inventory of goods include a bronze situla (Eggers 24), three bronze pans (one large Eggers 140, with the name P CIPI POLIBY on the handle, and two identical small Eggers 146, with the inscription P CIII POLIBIIF), a bronze oinochoe (Eggers 127), a patera, ladle (Eggers 162), a fragment of a thin-walled bronze bowl (possible Eggers 70), bronze fittings to a horned-cup, a bronze belt buckle, and two glass vessels (Ondrouch 1957, 13, 17, 18, 20–21, 26, 33, 34). The name from the Eggers 140 pan is of Publius Cipius Polybius, a well-known maker of metal goods from Capua, who worked in the second half of the first century AD (Krekovič 2000a, 49).

The most well-known is Branč, which sits near the Nitra River and was identified as having three phases: 1) first third of the second century, 2) second half of the second century, and 3) third century (Kolník, Varsik, and Vladár 2005, 313–314). The area excavated was 1.3 ha in size, being 230 m long and ranging between 40 and 80 m wide (Kolník, Varsik, and Vladár 2005, 313) (Fig. 3.6). A total of 130 cultural features were excavated, of which 40 were huts, most being Type III (Kolník, Varsik, and Vladár 2005, 313). The site saw its greatest expansion during the second phase, in the second half of the second century, from the B2c to B2/C1 phases, with sixteen huts dating to this period, and only eight dating to the third phase, falling in the C1b phase (Kolník, Varsik, and Vladár, 2005, 315). The banks of the Nitra River saw dense settlement of Germans during the Roman Period, with sites at Komjatice, Lipová-Ondrochov, and Bánov, among others (Kolník, Varsik, and Vladár 2005, 313) (Fig. 3.4).

49

There are settlement sites that have not been fully excavated, but surface surveys have shown material covering wide areas. At the site of Láb, near Bratislava, material dating from the second century BC to the fourth century AD covered an area of 10 hectares; one hut and an above-ground structure were excavated, dating to the second half of the third century (Elschek 2005, 262). Another site is Veľký Meder located 7 km from the Daunbe River and 15–16 km from Ad Statuas and Arrabona (Varsik 2005b, 301). The site has material dating from the mid-second to the second half of the fourth centuries AD, covering 6–8 hectares; only an area of 0.7 hectares was examined during rescue excavations, when three huts were excavated, with one hut dating to the second half of the third century AD (Varsik 2005b, 301 and 308).

Danube in this period (Mócsy 1979, 627–628). During the time from Augustus to Nero, military camps were of a temporary nature, built of earth and timber, which may contribute to the difficulty of locating them (Soproni 1980, 219). The legions were used to construct the important road network through Pannonia, starting under Augustus but not finished until the mid-second century (Mócsy 1974, 44–45, 109; Burns 2003, 208). Tiberius kept the same number of legions in Pannonia, with one at the frontier and two in the interior (Mócsy 1974, 44). Around AD 15 work was started on a castellum at Carnuntum, but it has still eluded archaeological detection (Mócsy 1979, 628; Kandler 1997, 258; Valachovič 2000, 3). It was under Tiberius’ instigation that the cliff road through the Iron Gates was constructed, and therefore it was probably also under Tiberius that the Danube river fleet was formed, but first mentioned only in AD 50 (Mócsy 1974, 45). Work continued on the Cliff Road under Claudius, Domitian, and Trajan (Mócsy 1974, 47).

Another area recently excavated is located to the east of Bratislava, toward Trnava. A series of sites were excavated before the construction of a highway, with a total of eleven find spots in the area of Vajnory, Trnava, and Ivanka pri Dunaji (Varsik 2005a, 275) (Fig. 3.7). One of the sites, usually referred to as either BratislavaTrnávka “Zadné” or just Zadné, had an area of 2,937 m2 excavated, revealing a two-phase German settlement: one of the second century and the other of the second half of the third century (Varsik 2005a, 275). Six huts were found dating to the second century, along with other features such as storage pits, fire-pits, and post-holes (Varsik 2002, 130). As seen with Branč and other sites, the huts were in clusters distributed across the terrain, usually with at least five meters between them.

Under Claudius, the numbers of legions were reduced on the Middle Danube, with Pannonia losing IX Hispana, leaving two legions stationed in the province, of which VIII Augusta was replaced by XIII Gemina (Mócsy 1974, 48). The two main fortresses in Pannonia remained at Carnuntum and Poetovio, but there was also a military camp established at Gorsium (Soproni 1980, 210) (Fig. 3.12). Even though there is indication that there was a castellum at Carnuntum under Tiberius, the earliest archaeological material dates to the Claudian period (Gabler 1979, 637; Rajtár 1996, 18; Kandler 1997, 258). A number of auxiliary forces were probably also stationed in Pannonia to help keep order both within the province and along the frontier (Mócsy 1974, 49; Gabler 1979, 637; Soproni 1980, 220). There is also evidence to show that a detachment from the XV Legion at Carnuntum was stationed at Vindobona (Börner 1997, 241) (Fig. 3.12). The primary evidence for the stationing of auxiliary forces on the Danube come from tombstones of soldiers, which primarily date to the middle of the first century AD, with evidence for forces at Arrabona, Brigetio, Aquincum, Lussonium, Teutoburgium, and Malata (Mócsy 1974, 49–50; Kolník 1998b, 63) (Fig. 3.12). It is tempting to believe that these forces were already in the province under Augustus and Tiberius, but the evidence points to the period of Claudius and Nero (Mócsy 1974, 50; 1979, 628–629).

Similar settlements of the B2 period, with huts spread out near a water source and having entrances turned toward the south, are seen in southeast Moravia, such as Blučina, Křepice, Komořany, Hrušky, and others (Droberjar 1997, 15) (Fig. 3.4). Blučina’s settlement was situated near the Cézavy River with 142 cultural features, of which six were huts dating to the second century (Droberjar 2002, 20) (Fig. 3.8). The site at Křepice had a two-phase settlement, with the first phase dating to the B2 period (Droberjar 1997, 17) (Fig. 3.9). Fourteen huts were identified belonging to the first phase, spread out along the bank of the Točka stream, with an iron-smelting workshop at the south end of the settlement (Droberjar 1997, 17–18). At Komořany, six huts were excavated dating to the second century, while the site at Hrušky had four huts and three pits dating to the second half of the second century (Droberjar 2002, 92 and 131) (Figs. 3.10 and 3.11).

The Flavian emperors did much to establish a new policy along the frontiers, but the changes they instilled was a gradual one, continued by later emperors (Mocsy 1979, 630; Gabler 1979, 637; Williams 1996, 38). A series of building programs took place to build garrison-posts first within Pannonia and probably some along the Danube, but between 89 and 102, all the forces were moved to the Danube frontier, leaving the interior of Pannonia demilitarized (Mócsy 1974, 80 and 88; Williams 1996, 19; Burns 2003, 216). Troops were transferred from the Rhine and it becomes clear that the Middle Danube was now the focus of frontier policy in Europe for the Flavian

3.6. The Roman Presence Along the Middle Danube Frontier: 10 BC–AD 166 During the period of Augustus, three legions were stationed in northern Illyricum (future Pannonia), the VIII Augusta at Poetovio, XV Apollinaris at Carnuntum, and IX Hispana at an unknown site, possibly Siscia (Mócsy 1974, 43; Soproni 1980, 219) (Fig. 3.12). It is unclear what types of fortifications if any were built along the 50

and later emperors (Millar 1981, 222). This change is also clear from the historical evidence in regards to how much fighting was done against the Marcomanni, Quadi, and Iazyges in comparison to groups along the Rhine or Lower Danube (outside of the period dealing with the Dacians). These projects first start in the early 70s in Pannonia under the governor C. Calpetanus Rantius (Mócsy 1974, 80 and 81; Gabler 1979, 637). More forces, especially auxiliary, were stationed in the province under Domitian (Mócsy 1974, 85). A third legion was also stationed in Pannonia, first the XXI Rapax, which was annihilated soon after and replaced by the XIIII Gemina stationed at Vindobona, but then Domitian also added a fourth, the III Flavia (Mócsy 1974, 85 and 86; Kolník 1998b, 63). There is also evidence for some reorganization of the river fleet, which now took the name of classis Flavia Pannonica (Mócsy 1974, 81, 109). The forts at Carnuntum and Aquincum were rebuilt in stone around AD 73 (Soproni 1980, 220). Based on archaeological evidence auxiliary forts of earth and palisades were built at Cannabiaca, Ad Flexum, Albertfalva, Vetus Salina, Ad Statuas, Ad Militare, Teutoburgium, Malata, and Gerulata (Mócsy 1974, 85 and 88; Gabler 1979, 639; Soproni 1980, 220; Rajtár 1996, 20) (Fig. 3.12). Legionary forts were also built on the left side of the Danube, such as opposite Carnuntum and Aquincum (Mócsy 1974, 110).16

After Trajan’s Parthian War and also in the period of Hadrian, the legions were reorganized again, with the X Gemina at Vindobona, XIIII Gemina at Carnuntum, I Adiutrix at Brigetio, and II Adiutrix at Aquincum (Mócsy 1974, 99). Auxiliary forces were interspersed along the frontier between the legionary bases, spaced at 20 km intervals (Mócsy 1974, 105). Along the frontier with the Quadi, the legionary camps included Vindobona, Carnuntum, and Brigetio; the auxiliary camps were Cannabiaca, Ala Nova, Aequinoctium, Gerulata, Ad Flexum, Quadrata, Arrabona, Ad Statuas, Ad Mures, Azaum, Crumerum, Solva (Soproni 1980, 221) (Fig. 3.12). The Antonine Itinerary has been of great help in locating the frontier forts in conjunction with archaeological exploration (Mócsy 1974, 104–105). From the period of Augustus to Trajan, the forts were earth-and-timber, but it appears that under Hadrian and Antoninus Pius some were rebuilt in stone and other forts after the Marcomannic Wars (Mócsy 1974, 107; 1979, 630; Soproni 1980, 221; Williams 1996, 64; Kolník 1998b, 63). The reforms instigated by the Flavians and Hadrian have been at times seen to indicate a Roman strategy of defense coupled with the concept of a closed frontier (Luttwak 1976, 60; Gabler 1979, 637; Millar 1981, 106). The ancient sources and archaeological evidence, however, do not agree with this assessement, as seen in this chapter. The frontier was never closed in the eyes of the Romans or to the Germans, except at specific periods. The concept of a permeable frontier is discussed in Chapter 4.

The reforms and reorganization started under the Flavians was continued under Trajan. In his preparations for the Dacian campaign, the four legions in Pannonia participated in the invasion, so replacement legions were brought into the province (Mócsy 1974, 91–92). By the end of the campaign, the legions in Pannonia were the XXX Ulpia Victrix at Brigetio, X Gemina at Aquincum, XIIII Gemina at Vindobona, and XV Apollinaris at Carnuntum (Mócsy 1974, 92). It was also sometime between 103 and 106 that Trajan divided Pannonia into the provinces of Pannonia Superior and Pannonia Inferior (Mócsy 1974, 92) (Fig. 3.12). The reasons behind this division are not precisely known: either there may have been a fear of having one governor in charge of four legions, or it was strategic. Mócsy makes a good point that Pannonia had a long frontier, 700 km, which fronted two different people, one sedentary (Germans) and the other nomadic (Sarmatians) (Mócsy 1974, 92; also supported by Williams 1996, 64). The Romans may have felt it necessary to split the province, so that separate tactics could be employed for each group (Mócsy 1974, 92). This idea is supported by the way in which the province was divided, since Pannonia Inferior encompassed a narrow strip of territory, no more than 50 km wide, on the west bank of the Danube from the river’s bend southward in the north to where it bends eastward again, directly facing the Iazyges; also, around 106, the Sirmian section of the frontier that had been taken from Pannonia in 86 was re-attached to Pannonia Inferior, thereby partially encircling the Iazyges (Mócsy 1974, 92). With this division, three legions were stationed in P. Superior, the XIIII Gemina, XV Apollinaris, and XXX Ulpia, and one in P. Inferior, the X Gemina (Mócsy 1974, 94).17

3.7. Overview of Roman-Style Structures in the Barbaricum Some of the more intriguing discoveries not only in Slovakia, but also in Austria, have been complexes beyond the Roman frontier that contain structures built using material and architectural design that are evidently of Roman origin. As to why they were built and for whom is still debated, with some scholars identifying them as military structures, while others see them as emporia or German elite residences built by Romans (Mócsy 1974, 91; Whittaker 1994, 115; Kolník 1991, 433; 1998b, 66). Because almost all of these structures fall outside the time period covered by this chapter, a detailed description of each will not be provided, but I provide a few observations on the topic of Roman-style structures as a whole (also see Chapter 4). The next section, however, does provide a detailed description of the Roman-style structure at Stupava, because of its importance to the overall focus of this book. These structures are found at the sites of Oberleiserberg and Stillfried in Austria and Devín (structures 1–9), Dúbravka, Stupava, Cífer-Pác, and Milanovce (today’s Veľký Kýr) in Slovakia (Figs. 1.4, 3.4, and 3.12).18 These structures are well known for their use of roof tiles, some with legionary stamps, such as LEG XIII GPF and LEG XV APOL from Carnuntum, LEG X GPF from 51

Vindobona, and LEG I AD from Brigetio (Kolník 1998b, 65). These structures do not date from the same period, but have a range of dates, with the earliest being Stupava, having its start either in the late first century, early second century, or mid second century, followed by structure II at Devín, which is either second or third century in date, and Bratislava-Dúbravka with a Roman-style bath dating to the third century, while the remaining structures date to the fourth century (Kolník et al. 1993, 246–247; Friesinger and Stuppner 1997, 283; Urban 1997, 295; Kolník 1998b, 66; Staník and Turčan 2000, 23).

comprehensive publication has been produced. Stupava is located 16 km from the Danube and around 6 km from the Morava River (Fig. 3.13). The archaeological site sits on top of a small hill next to the Stupavka stream, overlooking the Záhorie lowland and further in the distance the Morava River, both of which were part of the Amber Route. For a more detailed description of the geologic setting, see Chapter 5.1. After the Roman Period, the site was abandoned and eventually buried by natural processes. It was not until sometime in the eighteenth century (if not earlier) that the site was rediscovered by plowing, which turned up pottery, coins, and roof-tiles. These finds eventually reached academic circles, because in 1736 the site is mentioned in the Notitia Hungariae by Matej Bel, who identified the site as Roman (Ondrouch 1945/46, 1; Hečková 1986, 378; Turčan 1991, 66; Staník and Turčan 2000, 22). Local intellectuals such as I. Gond and W. Honz took interest in the site in the 1860s, collecting material from the site; and it was W. Honz who reported the site to the Austrian researcher F. Kennera and the National Museum in Budapest (Hečková 1986, 378 and 382; Staník and Turčan 2000, 22). From 1889 to 1891 an amateur archaeologist and antiquities dealer, M. Spitzer, dug at the site and discovered various items, of which only a few small finds, such as fragments of a bronze statue, were donated to a school in Bratislava (Staník and Turčan 2000, 22). The first archaeological excavations were done in 1925 by A. Gnirs of the archaeological institute in Prague (Ondrouch 1945/46, 1; Staník and Turčan 2000, 22). Gnirs only conducted short-term excavations, putting in two separate trenches across part of the site. These trenches uncovered some walls, a plaster floor, chunks of wall plaster with blue and/or green paint, stamped roof-tiles, and many small finds, along with evidence for a hypocaust system. Gnirs was unable to create any type of plan for the site, which he identified, however, as a Roman fort (Eisner 1933, 203– 205; Ondrouch 1945/46, 2; Staník and Turčan 2000, 22).

The material remains from these structures do not have a military appearance, lacking military-type fortifications, defensive constructions, weapons or other such evidence in large enough volume to make a military identification unassailable. Based on what has been found, the structures have more similarities in layout to that of villae and should be regarded within a civilian context (Wielowiejski 1970; Mócsy 1974, 91; Kolník 1981, 120; 1991, 433; 1998, 66; Rajtár 1986, 106; Vrba 1997, 97). The archaeological remains make a strong case for these structures to be seen as emporia or a combination of emporia and elite residences (Vrba 1997, 99; Kolník 1998b, 66). As already mentioned in section 2.3.2, these structures sit on or near various paths of the Amber Route, but also of importance are the locations of so many richly adorned elite graves on these routes, such as those at Zohor and Vysoká pri Morave. Based on the location of these structures and richly adorned graves, I believe that the German elite positioned themselves on the trade routes. This hypothesis is further supported by the accounts of Vannius’ taxing trade along the Amber Route, which show that the Germans were sophisticated enough to take advantage of positive economic situations within their territory. By placing themselves along the routes, the German elite would have had quick access to the wide assortment of Roman imports and also the ability to monitor the activities of Roman merchants operating on their territory. Keep in mind also Tacitus’ account of Maroboduus’ palace, where the Roman merchants were found. It seems plausible to believe that these German elites housed Roman merchants with themselves, and so one cannot separate the elite residence from the emporium, because they were one and the same (Whittaker 1994, 117).

Large-scale excavations were conducted in 1940 and 1941 by V. Ondrouch, from the Komenius University in Bratislava. In a few weeks, Ondrouch uncovered the entire site, following the walls of the various features (Staník and Turčan 2000, 22) (Fig. 3.14). He identified the different structures, such as the enclosure wall, main building (A –commander’s residence and B – quarters for other personnel), bath-building (D), and cistern (CIII) (Fig. 3.14). He also found a large variety of Roman-made items, such as pottery, glass, coins, various bronze items, and especially roof-tiles, some with legionary stamps, along with German pottery (Ondrouch 1945/46, 3–5). Ondrouch believed that the site had only one phase, AD 107–114, a dating based on the legionary stamps. He identified the site as a Roman Station (Ondrouch 1945/46, 2–3 and 46–47; Staník and Turčan 2000, 22).

It is thought that more of these types of structures will be found in the barbaricum. Various sites and surveys have made stray finds of roof-tiles with legionary stamps on them, which may indicate buried structures somewhere in the area (Kolník 1981, 120). These sites include Bratislava, Komárno, Štúrovo, Komjatice, Chotín, and Želiezovce, among others (Kolník 1981, 120 and 122; Pichlerová and Tomčíková 1994, 85–86; 1995, 96) (Fig. 3.4).

New excavations were carried out in 1974–1979, under J. Bujna, J. Hečková and L. Kraskovská (Kraskovská 1980, 117; Staník and Turčan 2000, 22). These excavations helped to understand the site better, especially to realize

3.8. The Roman-Style Structures at Stupava One of the largest of the Roman-style structures in the barbaricum is seen at Stupava. Excavations have been occurring at the site since at least the 1890s, but no 52

that Ondrouch had been wrong in identifying the site as having only one phase; instead, the site apparently had gone through at least two phases of development (Bujna 1976, 511; Hečková 1986, 391). These excavations explored the central structure in more detail and the area around it, especially to the northeast, where more walls were found just outside the building and within the enclosure wall (Bujna 1976, 510–511) (Fig. 3.15). The archaeologists also reexamined the southwest corner of the enclosure wall where Ondrouch had uncovered the east walls of a structure, which was further investigated and identified as a horreum (building S) (Bujna 1976, 510; Hečková 1976, 104; 1986, 379 and 380; Staník and Turčan 2000, 23) (Fig. 3.15).

being only scattered finds of pottery sherds and no evidence of cultural features (Staník and Turčan 2000, 23). The pottery does bear similarities to those types seen at the Bratislava oppidum, as well as at Devín and Plavecké Podhradi (Staník and Turčan 2000, 23). There is archaeological evidence for a Celtic settlement in Stupava itself as well as in the nearby area of Zohor (Staník and Turčan 2000, 23). German settlement of the region came in the B1 period, as shown in section 3.2, with elite graves found at Zohor and Vyskoká pri Morave (c. 6–8 km to the northnortheast) and a settlement at Dúbravka to the south. At the hilltop in Stupava, three German pits were found dating to the first half of the second century, showing some kind of settlement (Staník and Turčan 2000, 23). These features were later disrupted and partially destroyed by the walls put in for one of the phases of the Roman-style structure. There is also evidence in one of the German pits for a Roman military presence, with a number of nails from legionary footwear found in a thin layer at the top of the feature (Staník and Turčan 2000, 23). In conjunction with the German features, earlier excavations had also uncovered a similar German feature (2/75), containing a mixture of German and Roman provincial pottery, that had been disrupted by the construction of the east portion of the horreum (Hečková 1976, 105). According to Staník and Turčan, the Romans took possession of the hilltop in the middle of the second century and leveled it for the construction of the first phase of the complex (Staník and Turčan 2000, 23; Turčan 2002, 235).

J. Bujna believed that the site could have been started at the end of the first century or beginning of the second century, with the evidence pointing to sometime within the first twenty years of the second century (Bujna 1976, 511). His excavations did not uncover third-century remains except an ash pit filled with German pottery inside the main structure, which would indicate abandonment sometime in that century (Bujna 1976, 511). The excavations by Kraskovská in 1977 did in fact uncover some third- century material, especially bronze coins, one of which was found on the surface of the pavimentum in a corridor of the main building, but was poorly preserved and could only be identified as possibly an issue of Gallienus (253–268) (Kraskovská 1982, 36 and 46). J. Hečková’s conclusions agreed with those of J. Bujna, that the complex most likely dated to within the first two decades of the second century, because stamps of Legions X and XV were found in situ within the horreum and channel under the floor of building B (Hečková 1986, 390–391). Also, there was an absence of stamps from earlier legions and, even more important, material that could be dated to the first century AD (Hečková 1986, 391). As to the history of the complex, there were multiple phases, complicated by the location within the site of German features, sometimes inserted into buildings, ranging in date from the second half of the second century to the 3rd century (Hečková 1986, 391). In regards to the function of the complex, J. Hečková did not take a final stand one way or the other, but did point out that the construction of the horreum and some kind of structure north of the main building both interrupted the enclosure wall, which therefore was not much of a military feature (Hečková 1986, 393–394).

Work since the 1980s has enabled three separate main phases to be identified, with evidence to show that at the start of the second and third phases, previous phases were partially or totally cleared away (Staník and Turčan 2000, 23). Each phase has a different character, which shows the complex played various roles in the area (Staník and Turčan, 23–24). The first phase is seen as the creation of an outpost more in line with military needs, the second phase saw the creation of a trading-post, and the third phase saw the complex rebuilt as a villa rustica (Staník and Turčan 2000, 23 and 26; Turčan 2002, 235). The exact details of the first phase are still unknown, but two features were detected in the northern section of the enclosure, primarily under the main structure. The first feature had two parallel foundation trenches 2.5 m apart and apparently belonged to a long structure, oriented north/south. The structure had wooden walls, possibly wattle-and-daub, with good quality plaster on the outside, which was apparently applied more than once, while the roof was covered with roof-tiles (Staník and Turčan 2000, 24). The second feature was an enclosure wall that had stood at least 2.8 m high and was on average 45 cm thick, with its southern section at least 33 m long, while on the east side only 11 m were detected (Staník and Turčan 2000, 24). The foundations of this wall were made up of unshaped stones and joined with an earthen-mortar (Staník and Turčan 2000, 24).

In 1986 new excavations were started by V. Turčan and I. Staník to further explore the complex, but this time in a more controlled fashion so as to provide as much data as possible about the site (Staník and Turčan 2000, 23). These excavations are still ongoing because the site is being uncovered one portion at a time, with exploration going below the level of the foundation walls. The earliest archaeological material found on the hilltop are pottery sherds belonging to the Late Bronze Age; there were no associated cultural features (Staník and Turčan 2000, 23). There appears to be a hiatus until the La Tène period, but even then the evidence is small, 53

In the second phase, parts of the first-phase structures were torn down or incorporated into the new buildings (Staník and Turčan 2000, 24). At the center of the northern section of the complex (under the later main building) there was a narrow structure oriented north/south and partitioned into eight rooms; the walls were 65–70 cm thick and may have supported a second floor (Staník and Turčan 2000, 24). Nothing survived of the first floor itself or of the walls above the foundation courses; the walls were built of poor-quality mortar and in some places had a minimal amount of quarried stone (Staník and Turčan 2000, 24). There is evidence for another structure to the west of this central structure and in fact there may have been multiple sub-phases (Staník and Turčan 24; Turčan 2002, 235). During this second phase, a larger enclosure wall was built, surrounding an area of 65x70 m (Staník and Turčan 2000, 24–25).

north side and 70 on the south side, while interior walls were 60 cm. The corridor wall that surrounded the interior courtyard was 50 cm thick on the north side, 35 cm on the west side, 50 cm on the south, and the east wall, which was originally an outside wall, 60 cm. The entrance is unknown, but thought to be on the south side, possibly room XI. In the east, Building B was added, abutting the east side of the courtyard wall (Staník and Turčan 2000, 25) (Fig. 3.15). This structure had three rooms (I–III) on the north side facing a larger room (IV), while on the east side were another two rooms. The structure measured 17.75 x 20.85 m. Room I was 6.9 x 6.15 m in size, with room II 6.8 x 6.15 m, and the largest area, room IV, 16.25 x 13 m. Average thickness for the outside walls was 65 cm on the north side, 60 cm on the east and west sides, and 70 cm on the south side, while the interior walls ranged from 50–60 cm. The floor was preserved in room II and III and in small areas of I and IV.

The third phase is the best known, having been uncovered by V. Ondrouch, then further explored from the 1970s to this day. The construction of the third phase cleared away most of what had been seen in the second phase, leaving only the enclosure wall intact at first (Staník and Turčan 2000, 25). Within the compound various structures were built, including the main building, a bath-house, cisterns, and piping to bring water and/or heating to the bath-house (Fig. 3.15). The enclosure wall has the following lengths: 69.55 (north side) x 64.85 (east side) x 73.30 (south side) x 65.40 (west side) meters, enclosing an area roughly half a hectare; the walls varied in thickness, with the north wall being around 80 cm, while the south wall was 75 cm thick (Ondrouch 1945/46, 2–3; Bujna 1976, 510) (Fig. 3.15). The walls, possibly having reached a height of three meters, were built of hewed ashlar and mortar placed on a bed of river stones, which were laid out with no mortar, (Ondrouch 1945/46, 3; Turčan 1990, 41; 1991, 67).

Both structures, A and B, had a hypocaust system for at least some of the rooms with praefurnia attached on the outside (labeled as T1 for building B and TII for building A) (Fig. 3.15). Heating channels were located coming from TI and TII, with rooms I–III in building A heated and a channel went from TI across room IV in building B ending in room XI of building A. The interior of the structures had high-quality pavimentum, with evidence for wall painting in some rooms and corridor; the painting appears to be borders painted in red, green or blue around the rooms, but there is some evidence for painted designs as well (Bujna 1976, 507; Staník and Turčan 2000, 25) (Fig. 3.16). One of the more interesting recent discoveries was a collapsed wall from the interior corridor into the courtyard that preserved two windows and some wall-painting (Staník and Turčan 2000, 25) (Fig. 3.17). The collapsed wall section preserved four meters of original height from the north wall of the atrium and about 24 m2 of area (Staník and Turčan 2000, 25). The two windows had arched tops and were framed by a band of red paint on the white plaster. The wall and windows were removed as a whole and conserved (Staník and Turčan 2000, 25); they were placed on display at the AM-SNM in Bratislava.

The main building located in the northern part of the compound appears to have gone through multiple subphases, with additions being built (Staník and Turčan 2000, 25). The first part of the structure was the western building (A), which was a square construction (29.80 x 28.60 m) that had three main internal spaces on the west side overlooking a walled courtyard; the walls of this structure were supported by piers spaced along the outside, possibly to help sustain a second floor (Staník and Turčan 2000, 25) (Fig. 3.15). Over time, the interior was subdivided, with rooms added in the north, so that rooms were spaced around a central enclosed corridor overlooking a smaller courtyard or atrium (labeled F by Ondrouch), while to the south more rooms were added on the outside of the original construction. By the last phase, the structure had 12 rooms (two more rooms on the south side were recently discovered by Staník and Turčan) around the corridor (labeled as X), with the largest rooms being III (5.4 x 9.55 m) and V (7.8 x 9.55 m). A pavimentum floor was preserved in rooms I, III, part of IV, VII, VIII, XI, and the north section of the corridor. The corridor was 2.8 m wide, enclosing a courtyard that was 12.4 x 13.8 m in size. Average wall thickness had the outside walls either 90 cm on the west side, 75 on the

To the southwest of the main building was a bathbuilding, labeled structure D by Ondrouch (Fig. 3.15). The identification of the structure as a bathing facility was based on the water channels connected to the structure from the cisterns (EVII and CIII) and the use of a suspended floor on suspensura for the circulation of heated air and water (Fig. 3.18). The building itself is oddly shaped to accommodate the various bathing rooms and heating system, with general size of 14.3 x 18.35 m. There are six rooms, while the channel for the water coming in on the east side, turned and entered the south side of the structure, to where rooms V and VI are located, and ending in room III. The outer walls of the bath-building range in thickness from 60–70 cm. Ondrouch believes the heating unit may have been in 54

room I (Ondrouch 1945/46, 8). Room III was the central room of the building with an apsidal section on the west side. Rooms I and II were caldaria, while IV and VI frigidaria (Ondrouch 1945/46, 6–7). Room I had a partially preserved suspended floor on suspensura (pillars were made of bipedales, each one 7.1 to 7.7 x 19 x 20 cm in size, some stamped with LEG XIV and others with GIF or IVLI OF) (Ondrouch 1945/46, 9) (Fig. 3.18). Room IV, 3.75 x 4.7 m, was well preserved with brick tiles, which were heavily encrusted (Ondrouch 1945/46, 8).

tegulae?). Hollowed bricks were found in the north wall to channel the hot air up the sides. Along with this system, there is evidence for the floor having been raised up on bipedales to further circulate air (Hečková 1976, 104). The small-finds were quite poor from this structure, being primarily stamped roof tiles (Hečková 1976, 104). After the structure was abandoned, apparently a German hut or feature of some type was placed in the east section of the structure; a large amount of German and Roman provincial pottery and animal bones was found there (Hečková 1976, 105).

KV1 is the channel built to bring water from cisterns EVII and CIII to the bath building. The channel was well preserved because it was placed deeper than the foundation walls, with the channel itself 50 m long and 25 x 25 cm in size, lined with stones, and covered with tegulae stamped with LEG XIIII G (Ondrouch 1945/46, 6) (Fig. 3.19). KV2 was the channel connecting cistern EVII to KVI and was built using imbrices. Pairs were joined with convex surfaces out, and a series of pairs fitted together to create a 12.65 m long pipe within a channel (Ondrouch 1945/46, 10–11) (Fig. 3.20). Off of KVI was a third channel, KV3, which went south below the enclosure wall, but as to where it ended is unknown. Ondrouch envisioned KV3 as a regulatory channel used to release water if there was too much (Ondrouch 1945/46, 5).

Sometime in the later third century, the complex was either abandoned or attacked, with evidence of fire left on the plaster walls and floors in the main building (Staník and Turčan 2000, 25). There is evidence for a final phase during the Great Migration Period (fourth to fifth centuries), with a long feature of some type constructed between the bath-building and central building (Staník and Turčan 2000, 26). Based on the remains, outside of the few isolated instances of fire, there is no clear destruction layer over the entire site and the material remains are not of the type to imply a catastrophic event occurred, catching the inhabitants unaware (Staník and Turčan 2000, 26). Instead, it appears the site was abandoned, with a certain amount of goods being removed (Staník and Turčan 2000, 26). Two huts dating to the second half of the third century found just outside the complex (in a place called Morávkove pole) may date to this abandonment period or just after (Staník and Turčan 2000, 26; Turčan 2005) (Fig. 3.22). These two huts contained a large amount of Roman pottery and even brick and roof-tile fragments, which are all thought to have come from the complex, with V. Turčan envisioning the Germans looting the abandoned structures for whatever was usable (Staník and Turčan 2000, 26; Turčan 2005, 298–299).

Cistern EVII was 3.55 x 3.65 m in size with walls either 70 or 75 cm thick, with an interior space of 2.05 x 2.25 m (4.61 m2) (Fig. 3.15). In the southeast corner of the complex were three cisterns, CI–III. CI and CII are similar in size to cistern EVII, but no connecting channels have been found; of the two, CII was preserved best (Kraskovská 1982, 53). CIII was a larger structure which Ondrouch identified as a Castellum aquae, and so is either called a cistern or reservoir in the literature (Ondrouch 1945/46, 5; Kraskovská 1982, 63). The structure was 15.5 x 13.5 m in size with three separate sections and stood around 80 cm from the enclosure wall (Kraskovská 1982, 65) (Fig. 3.21). Channel KVI was found to connect the bathhouse with cistern CIII, entering the cistern at the southwest corner of the structure, crossing through one section and ending at the wall between the other two sections (Kraskovská 1982, 65– 66). Support pillars were found between the enclosure wall and southern wall of CIII, probably to support the structure on that side (Kraskovská 1982, 65).

The site has from almost the beginning of its discovery and first explorations been given a military identity. The evidence for a military presence throughout the life of the complex, however, is weak, consisting of pieces of iron ring-mail, a bronze fitting in the shape of a lion head that may have come from a helmet, shoe-nails from legionary footwear, the large amount of roof-tiles stamped by various legions, and the enclosure wall (Turčan 1991, 67; Staník and Turčan 2000, 26) (Fig. 3.23). The site itself does not have a military appearance by the second and third phases and, as already stated in section 2.3.7, this complex and the others share more characteristics with villa complexes seen just over the border in Pannonia (Vrba 1997, 97). In regards to the first phase being of a military nature, the verdict is still out, since excavations have not been completed and more supporting evidence may yet be found. It would appear that the site most likely had more economic importance than military, probably serving as an emporium/mansion on the Amber Route incorporated into some kind of villa rustica (Kolnik 1991, 433).

The structure, identified as a horreum, just outside the southern portion of the enclosure wall had a measurement of 11 x 13 m (Hečková 1976, 104) (Fig. 3.15). It appears that the west section of the enclosure wall in this portion was removed for the construction of the horreum. The building was poorly preserved with the western and southern portions only surviving as negatives, while the northwest side is well preserved, with surviving pavimentum and walls (Hečková 1976, 104). Underneath the structure was evidence for a hypocaust system, with canals, 30–40 cm wide and 60–70 cm deep, lined with large brick plates (40 cm diameter, possible roof-tiles or 55

The complex did not sit alone, but was apparently integrated into some kind of settlement in Stupava. Excavations in 1974–1976 uncovered German features outside the enclosure wall on the south side (Kraskovská 1977b, 178). One of these features was a pit containing German pottery, Roman provincial pottery, broken fragments of Roman brick, and a large amount of animal bone. The other feature appears to have been an aboveground structure identified through post-holes and a firepit ringed with stone at the center, along with German pottery and animal bones (Kraskovská 1977b, 178). Another German feature of some type was uncovered on the east side of the complex as well, containing German pottery and animal bones (Kraskovská 1977b, 178). Around 60–100 m to the northeast of the complex, more German cultural features were found (Hečková 1976, 105). Also, as stated in Chapter 1, V. Turčan had excavated a hut at Urbárske Sedliská, which lies 800 m to the northwest, and in part two of this book I shall present the data from two more huts found there (Turčan 1985; 1991, 67) (Fig. 3.22).

K. Kuzmová in her 1997 publication analyzing and listing all the terra sigillata found in southwest Slovakia provided a summary of what terra sigillata had been found at Stupava up to the year of publication (Kuzmová 1997, 47–50). Table 3.2 below is based on Kuzmová’s table 12, summarizing the forms and types of terra sigillata found at the site (Kuzmová 1997, 49; Tabelle 12). A total of 159 pieces were found at the site up to the year 1997, and a few more pieces have been found since. Based on the figures, Rheinzabern Ware is clearly the most common, followed distantly by Westerndorf, Central-Gaulish (Lezoux), and Pfaffenhofen Wares. Central-Gaulish dates to the second century and Rheinzabern Wares date between AD 148–245, which fit well with the other evidence for the complex being founded in the early second and probably reaching its third phase by the middle of the century; and with the presence of Pfaffenhofen Ware, showing that the complex lasted at least until the middle of the third century (see Chapter 9.1.13 for more information on terra sigillata in Slovakia). A comprehensive study still needs to be done on the other types of Roman Provincial and German pottery. Overall, both groups of pottery primarily date to the second and into the first half of the third century (Hečková 1986, 391). Various types of Roman provincial pottery have been found, including Pannonian Brick-Red Ware, Pannonian-Striped Ware, Coarse-Gray Ware, Smooth-Gray Ware, Glazed Ware, Thin-Walled Wares, Pannonian-Stamped Ware and mortaria (Ondrouch 1945/46, 37–42; Bujna 1976, 499–500; Kraskovská 1982, 68–69). The German ware is a mixture of fine and coarse wares, with Nail/Finger Impressions, Comb Lines and Incised Lines being the most common decorations (Ondrouch 1945/46, 42–45; Bujna 1976, 500 and 502; Kraskovská 1982, 70). There is also German pottery from the Late Roman Period, including one vessel that may have some type of runic writing on the body (Staník and Turčan 2000, 26) (Fig. 3.25).

Roof tiles with legionary stamps have been found within the complex and even outside the complex (Figs. 2.31 and 2.33). Table 3.1 provides a list of the various stamps and the amounts found in various parts of on Kuzmová’s table 12, summarizing the forms and types of terra sigillata found at the site (Kuzmová 1997, 49; Tabelle 12). A total of 159 pieces were found at the site up to the year 1997, and a few more pieces have been Rheinzabern Wares date between AD 148–245, which fit well with the other evidence for the complex being presence of Pfaffenhofen Ware, showing that the complex lasted at least until the middle of the third founded in the early second and probably reaching its third phase by the middle of the century; and with the found since. Based on the figures, Rheinzabern Ware is clearly the most common, followed distantly by Westerndorf, CentralGaulish (Lezoux), and Pfaffenhofen Wares. CentralGaulish dates to the second century and the compound based on the data provided by J. Hečková in her 1986 article, combining the finds of Ondrouch’s excavations and those of the 1974–1979 seasons (Hečková 1986, 383 and 390; Tables 1 and 2). Since 1979, many more stamped roof-tiles have been found, but not tallied. As can be seen from the table the most common stamp was from LEG XIII G, which was known to have been stationed in Carnuntum around AD 100. Tiles from this legion were not alone, with stamped tiles coming from other parts of Pannonia as well, where other legions were stationed. Along with the legionary tiles were those of second century private firms such as IVLI OF, C VAL CONST, KAR, C.I.IVL, and FL SEX(T), but in very small numbers (Ondrouch 1945/46, 14–15; Bujna 1976, 510).

A large variety of small finds have also been found at the site, including fibulae, glass, lamps, belt decorations, bone hair-pins, buttons, iron implements, and other items, as well as those already mentioned above (Ondrouch 1945/46, 26–29; Kraskovská 1982, 70) (Fig. 3.26). A small gold earring was also found (Ondrouch 1945/46, 27). The glass fragments found by V. Ondrouch during his excavations primarily came from cistern EVII, one channel in the bath-build and another channel just outside (Ondrouch 1945/46, 27–28). The forms included bowls, cups, and unguentaria (Ondrouch 1945/46, 28). Two glass-paste beads were also found by V. Ondrouch, both blue (Ondrouch 1945/46, 28–29). Two spindle-whorls were found during the 1974 season, but one came from a German feature just north of the enclosure wall and the other came from the later German feature present in Building B (Bujna 1976, 502).

56

Table 3.1. Stamp types found on roof-tiles and their find-spots at Stupava Structures Stamp A B CII CIII D S EVII KVs Unknown LEG XIIII 1 LEG XIIII G 71 41 2 1 6 LEG XIIII IIG 1 LEG XIΛ IIG 2 LEG XIIII GMV 1 4 4 LEG XIV 1 1 1 1 LEG XIV G 1 10 LEG XIΛ G 1 1 1 LEG X GPF 8 3 1 LEG X GEM 3 LEG X PF 2 LEG XV APOL 5 8 1 1 2 LEG XV AP 2 3 3 LEG XV POL 1 LEG XIII GE OLI 1 LEG XIII GEM 1 G.I.F 1 39 2 1 C . I . IVL 2 2 CI . IVL 1 KAR 2 FL SEX 3 IVLI OF 5 OF. PLS 1 IIIXIIIIG 2 C VAL CONST 1? Table 3.2. Terra sigillata found at Stupava Type of Terra Sigillata Form SG? MG Rh Rh/We We We/Pf Pf Total Drag. 30? 2 1 1 2 6 Drag. 30/37 1 1 Drag. 37 9 48 4 17 4 15 97 Drag. 18/31 4 11 6 1 1 23 Drag. 32 6 6 Drag. 33 1 2 1 4 Drag. 35/36 1 1 Drag. 38 1 1 Drag. 39 1 1 Drag. 43 3 3 Drag. 44? 1 1 Drag. 46 1 1 Drag. 54 7 7 Unknown 3 3 6 Total 1 18 83 15 20 5 17 159 SG = South-Gaulish Ware MG = Central-Gaulish Ware Rh = Rheinzabern Ware We = Westerndorf Ware Pf = Pfaffenhofen Ware The coins found from the site support a middle of the second century date for the period in which the site saw its fullest extent, and with the presence of so many third century coins, it is apparent that Stupava was still important in that century as well. Bronze coins have been found that were issued during the reigns of the following

emperors and wives: Hadrian, Antoninius Pius, Faustina, Iulia Domna, Elagabalus, Maximinus Thrax, Gordian I and III, Gallienus, Trajan Decius, Aurelian, Licinius, and Constantius II (Hečková 1986, 382 and 391; Rajtár 1986, 106 and Tab. 4). Silver coins have been found from Antoninus Pius, Elagabalus, Gordianus III, Alexander 57

Severus and Aurelian (Hečková 1986, 382; Rajtár 1986, 107: Tab. 4).19 One gold coin has also been found, an aureus of Vespasian (Staník and Turčan 2000, 26).20 From data up to 1986, a total of 13 coins had been found within the complex and another 35 around it, making a total of 48 coins found in relation to the site, but this number has gone up since then because of more recent excavations (Rajtár 1986, 107: Tab. 4). The problem with many of the coins is their circumstance of discovery, so that many of them could not be used to help date parts of the complex or be associated with specific building phases (Rajtár 1986, 106). What the coins do show is the start of the complex in the late or early first century AD, with the greatest activity at the site in the third and fourth centuries. Since the Roman-style structures show abandonment or at least decline in the mid- to late-third century, it would appear, based on the coin evidence, that activity moved nearby.21

constructed on top of an earlier German settlement some time during the principate of Trajan. This complex was eventually composed of an enclosing wall, a central building of two structures to either side of a courtyard, a bath-building, cisterns, and a horreum. The complex went through multiple phases in layout and focus, with the first phase seen as a military occupation, the second as an emporium, and the third a villa rustica. The structures were built of good quality Roman mortar and stone, with roof-tiles manufactured at the legionary forts along the Danube as well as by private firms. The main building and bath have evidence for sub-floor heating. Based on the evidence, the complex was abandoned sometime in the later third century and looted by local Germans. The information presented in this chapter shows that southwest Slovakia was an integral part of the Roman world. The archaeological evidence shows that the Germans of southwest Slovakia had a deep interest in Roman goods, which were imported in great abundance. At the same time, the elites of the Quadi were embroiled in the politics of the Roman Empire, either being put into or removed from power by the Romans, while also participating in Roman power struggles. For the Early Roman Period, one of the most important sites was the complex of Roman-style buildings at Stupava. Whether it was built by Romans for their own use, for the use of German elites, or a mixture of the two, the complex was an important symbol of how important the Germans of southwest Slovakia were to Roman interests, both economic and political. This importance is seen in the placement of the complex on the Amber Route, the quantity of Roman goods within it, and, if built for German elites, a clear sign of the close relationship that the Romans had with their client kings. These issues are explored in more detail in Chapter 4, which focuses specifically on the theoretical aspects of the Roman frontier, the concept of Romanization, and how both affected the Germanic cultural identity of the Quadi in southwest Slovakia.

3.9. Conclusion The archaeological evidence shows that the Germans entered southwest Slovakia during the first quarter of the first century AD, a date based on graves found at various sites in the region. German expansion was quick and by the beginning of the second century, the Quadi had occupied the territory from the Morava River in the west to the Ipeľ River in the east. At the same time as they expanded across southern Slovakia, the Celto-Dacian culture known as the Púchov Culture, and sometimes identified with the Cotini mentioned in ancient sources, continued their occupation of north-central Slovakia. Over time, this cultural group came into conflict with either the expanding Quadi or the also expanding Przeworsk Culture, who were arriving from southern Poland. The Púchov Culture saw a Germanization of their matieral culture and old Celtic strongholds were abandoned or shrunk in population. In the east, the Dacians were pushed out of northern Hungary by the arrival of the Sarmatian Iazyges, but the Celto-Dacian culture around the oppidum of Zemplín prospered as a powerful trading center.

Endnotes

During the period of German expansion into Slovakia, the Romans solidified their control of the region of modern day Hungary, creating the province of Pannonia, which was later subdivided into Pannonia Superior and Inferior. At first, the Roman legions were stationed throughout the province in temporary forts, but over time, units were moved to the Danube River, which was becoming increasingly important as a frontier with the Germans in southern Slovakia. By the period of Trajan, all Roman military forces that had been spread out through Pannonia, were now stationed along the Danube, with legions garrisoned at permanent forts, between which were smaller forts manned by auxiliary forces. Many of these forts were rebuilt in stone during the principate of Hadrian or after the Marcomannic Wars.

1. For Slovakia, the multi-author book titled Archaeological Research in Slovakia (1981) is still the primary archaeological account in English, even though it is out of date. T. Kolník’s account of the Roman Period in the book provides an overview that does not go into enough detail so that it has limited use. One of his most important articles in Slovak and which is still cited in numerous publications in Slovakia is “Prehľad a Stav Bádania o Dobe Rímskej a Sťahovani Národov.” (1971). A German version of the article was published in 1977, but does not go into the same detail as the Slovak original. Even though there have been many discoveries since 1971, the basic information and descriptions of the various periods in the article still hold true. Kolník’s article was the starting point for the archaeological section of Chapter 3.

In southwest Slovakia, in the modern town of Stupava, a complex made up of Roman-style buildings was 58

2. By the Final La Tène, it appears that the Celts of the Czech Republic and Slovakia were unified under the Boii king Critasirus, who had control of the Taurisci in Noricum as well (Strabo 7.5.2).

10. Abrahám is also interesting because of another find. When a surface survey was conducted on the left bank of the Gidra stream near the cemetery, a variety of artifacts were found of La Tène, German, and Roman origin, amongst which was a fragment of cameo glass, with blue translucent glass on which was a layer of opaque white glass (Bartík and Varsik 1994, 63). The fragment came from a drinking cup, most likely a skyphos, and shows a seated woman, seen in three-quarter view, looking to the left; she seems to be leaning on the chair, wearing a tunic without sleeves, while a piece of cloth is thrown over her left shoulder and her hair is combed back and knotted (Bartík and Varsik 1994, 64). There is some dispute about the fragment’s date, but because of its location and the weathering it is thought to be ancient, and from the style of the figure, the glass is dated to the first century AD (Bartík and Varsik 1994, 65 and 69).

3. While this statement was originally made in 1971 by Benadik, it still remains relatively true to this day. The only Celtic site that has been excavated extensively and even partially reconstructed is the Havránok oppidum at Liptovská Mara, belonging to the Púchov Culture (see Pieta 1996). 4. For a detailed description of the Púchov Culture, see K. Pieta’s Die Púchov – Kultur (1982). 5. In the areas identified as having been part of the oppidum and settlement at Bratislava, at least seven separate areas show layers of ash with chunks of burnt wood, with some also having weapons and human bones, dating to sometime between 58 and 44 BC (Pieta and Zachar 1993, 185–188). These layers have been dated by the fibulae found, especially the Feugere 21a1 ‘Alesia’ type (Pieta 1993, 188).

11. The presence of these serving sets and the drinking of wine by elite Germans is seen as part of their wellknown banqueting, which had significant social ramifications. Tacitus reports that banqueting was important to the Geramns, during which not only was there much drinking and fighting, but also various types of tribal and inter-tribal business would be taken care of (Tac. Germ. 22). Later German medieval poems, such as Beowulf, the Nibelungenlied, and Edda further relate the importance of banqueting (Owen 1960, 137). Feasting or banqueting, which also could include the redistribution of wealth, was an important part of tribal societies throughout the world (see Chapter 4.3 for more information on this in relation to German identity).

6. The Przeworsk culture originated in the central and southern portions of Poland, with its earliest elements appearing in the La Tène and lasted into the Great Migration Period (Droberjar 2002, 268–269). The culture spread out from Poland into northern Moravia with settlements there appearing by the Final La Tène, followed by expansion into northern Slovakia, Ukrainian territory beyond the Carpathians between the Odor and Bug Rivers, and also into the Silesian parts of northern Czech Republic (Droberjar 269 and 270). It is thought their spread and control of trade was for the acquisition of metals (Droberjar 269). In regards to the ancient sources, the Przeworsk culture is sometimes identified with the Vandals and Lugii.

12. Očkov has 267 cremation graves (Kraskovská 1978, 7). 13. Other cemeteries include Ivanká Pri Dunaji, Šopron, Bratislava-Devínska Nova Ves, Kostolná pri Dunaji, Sládkovičovo, and many others (Kraskovská 1965, 163; 1977, 487; 1978, 10–11; Krekovič 1979, 413).

7. There is also a burial dating to 50 BC found at Cífer (Kolník 1971, 511). The material indicates a German warrior and was identified as a mercenary fighting for the Boii against the Dacians during that conflict. A reappraisal of the material culture indicates that the burial was mis-dated and should actually be placed sometime in the early B1a phase instead (Kolník 1971, 511).

14. Sites that fall under this category are BratislavaDúbravka, Vysoká pri Morave, Láb, Zadné, Stillfried, Kuchyňa, and Zohor. 15. This grave was found in 1949, while the other three were found in 1957 (Ondrouch 1957, 40; Krekovič 2000a, 49). The 1949 grave is identified as Grave 1 and the 1957 graves as Graves 3, 4 and 5. The grave identified as Grave 2 was the cremation grave, which had a bronze vessel (Eggers type 38) and inside it a bronze fibula (Kraskovská 1978, 13).

8. See footnote 6 on the Przeworsk Culture. 9. Other Roman imports include bronze vessels such as pans of the Eggers 137 type, Poggendorf style bowls (Eggers type 92), buckets (Eggers type 6), noricpannonian belts (Garbsch types B1, B7, variant 5, E3a, G1c, G2d, Ka, R1, and R2), hairpins (MotykováŠneidrová types 23/1965, 24/1965, and 57/1965) (Kolník 1971, 513–514). There is one inhumation grave from Reca that had a fibula of the Almgren 19 type (Kolník 1971, 514).

16. The legionary fort opposite Brigetio at Izá-Leányvár had been originally thought to have been constructed first under the Flavians, most likely Domitian, with an earth and palisade construction. Excavations had uncovered sun-dried brick walls under the later stone fortress, which were identified as barracks of the Flavian period (Rajtár 2000, 34). New excavations have expanded the 59

understanding of these earth and timber buildings with the discovery of associated coins dating to AD 179 (Rajtár 2000, 35–36). It appears now that the fort, identified as Celemantia, was actually started as an earthand-timber fort under Marcus Aurelius and was then rebuilt in stone by Commodus, under whose period the fort at Brigetio was also rebuilt with stone (Rajtár 2000, 35 and 37). 17. Krekovič makes a good point that the number and types of troops stationed in Pannonia Superior indicate that the Romans saw the Germans as the greater threat (Krekovič 2001b, 107). 18. There is also the site of Mušov in the Czech Republic, but this one has the characteristics of a legionary fort and dates primarily from the period of the Marcomannic Wars into the early third century (Droberjar 2002, 193–196). 19. The antoninianus of Alexander Severus was discovered by the author during the 2000 field season within building A. 20. The aureus of Domitian was found in the back-fill pile during excavations in the 1970s by V. Turčan and so cannot be used for accurate dating purposes (communicated to me by V. Turčan). 21. The possibility of structures at Stupava of later date needs to be explored further. In fact, satellite photographs of the site show promising outlines of what could be two structures near the excavated Roman-style structures. It is hoped these will explored at some point in the near future.

60

Figure 3.1. Map of Slovakia showing La Tène period sites, including Puchov and Celto-Dacian (Based on a map from Pieta 1981, 99: Figure 1). 1. Bratislava, 2. Devín, 3. Plavecké Podhradie, 4. Nitriansky Hrádok, 5. Divinka, 6. Liptovská Mara, 7. Zemplín, 8. Hurbanovo-Abadomb, 9. Hurbanovo-Bohatá, 10. Bodza-Holiare, 11. Podtureň, 12. Nitra, 13. Púchov, 14. Streda nad Bogrogom, 15. Prešov, 16. Šebastovce, 17. Hainburg

Figures

61

Figure 3.2. Map of Moravia, showing the La Tène settlement area with German incursions, both from the Elbe and Przeworsk groups. 1. Bořitov, 2. Brodek u Prostějova, 3. Břest, 4. Držovice, 5. Kojetín, 6. Količín, 7. Malé Hradisko, 8. Pravčice, 9. Rymice, 10. Štramberk, 11. Vyškov, 12. Dubany, 13. Hustopeče (based on a map from Droberjar 1997, 135: Abb. 56).

Figure 3.3. Archaeological groups of the Middle Danube c. AD 50.

62

63

64

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. C. Gerulata D. Brigetio

25. Komjatice 26. Cífer 27. Bíňa 28. Krížovany nad Dudváhom 29. Trnava-Hrnčiarovce 30. Šarovce 31. Horná Streda 32. Janova Ves 33. Bratislava 34. Bratislava-Devín 35. Krakovany-Stráže 36. Vysoká pri Morave 37. Mlékojedy 38. Nejdek 39. Tišice 40. Dobřichov-Pičhora 41. Vlčkovce 42. Žlkovce 43. Mistelbach 44. Altenmarkt im Thale 45. Láb 46. Hrabětice 47. Nedakonice

A. Vindobona B. Carnuntum

Stupava/Urbárske Sedliská Dúbravka Bratislava-Trnávka Sládkovičovo Abrahám Kostolná pri Dunaji Kráľová pri Senci Devínska Nová Ves Bohdanovce Báhoň Kuchyňa Branč Pobedim Blučina Křepice Horní Věstonice Mikulov Mikulčice Hrušky Rajhrad Křižanovice Komořany Mušov Zohor E. Celemantia

48. Bzenec 49. Velatice 50. Liptovská-Mara (Havránok) 51. Spišske-Tomašovce Čingov 52. Hrabušice 53. Varín 54. Nezbudská Lúčka 55. Žilina 56. Zemplín 57. Šebastovce 58. Prešov 59. Brno-Chrlice 60. Uherský Brod 61. Dobročkovice 62. Streda nad Bodrogom 63. Lesné 64. Kvakovce 65. Milanovce (Veľky Kýr) 66. Nitriansky Hrádok 67. Bánov 68. Marušova 69. Pohronský Ruskov 70. Štúrovo

71. Pác 72. Šaľa 73. Močenok-Sládečkovce 74. Tvrdošovce 75. Ondrochov 76. Andovce 77. Bohatá 78. Vráble 79. Čifáre 80. Čaka 81. Plavé Vozokany 82. Bešeňov 83. Očkov 84. Nitra-Janíkovce 85. Nitra-Párovské Háje 86. Chotín 87. Malacky 88. Borský Peter 89. Komárno 90. Veľký Meder 91. Oberleiserberg 92. Stillfried 93. Želiezovce

Figure 3.4. Map of sites mentioned in Chapter 2 and 7-9 (both in text and figure-captions), with Roman forts indicated along the Danube River by letter.

Figure 3.5. Site-plan of Dúbravka (modified image from Elschek 1995, 45: Abb. 1).

65

66

Figure 3.6. Site-plan for the settlement of Branč, showing huts for all three phases (from Kolnik 1998a, 146: Abb. 2).

Figure 3.7. Site-plan for Bratislava-Trnávka “Zadné” (Varsik 2002, 130: Abb. 2).

Figure 3.8. Site-plan for Blučina, showing the marked cultural features (Droberjar 1997, 17: Abb. 4).

67

Figure 3.9. Site-plan for the settlement at Křepice (Droberjar 2002, 144).

68

Figure 3.10. Site-plan for Komořany (Droberjar 1997, 18: Abb. 5).

Figure 3.11. Site-plan for Hrušky, outline is of the excavated area (Droberjar 1997, 18: Abb. 6).

69

Figure 3.12. Map of Pannonia by AD 200, with the various forts identified and the division into Superior and Inferior (based on Mócsy 1962, 649/650: Abb. 5; 1974, 93: Figure 16; Figure 59). Sites in Quadi Territory: A. Mušov D. Stillfried B. Oberleiserberg E. Stupava C. Niederleis F. Devín Forts along the Danube: 1. Cannabiaca 2. Ala Nova 3. Aequinoctium 4. Gerulata 5. Ad Flexum 6. Quadrata 7. Arrabona 8. Ad Mures 9. Ad Statuas 10. Celamantia 11. Hatvan 12. Azaum 13. Crumerum 14. Solva

G. Dúbravka H. Cífer-Pác I. Trenčín

J. Milanovce

15. Pilismarót-Szob 16. Visegrád 17. Cirpi 18. Ulcisia Castra 19. Albertfalva 20. Campona 21. Matrica 22. Vetus Salina 23. Intercisa 24. Annamatia 25. Lussonium 26. Alta Ripa 27. Alisca 28. Ad Statuas

29. Lugio 30. Contra Florentiam 31. Altinum 32. Ad Novas 33. Teutoburgium 34. Cornacum 35. Cuccium 36. Bononia 37. Casteffum Onagrinum 38. Cusum 39. Acumincum 40. Rittium 41. Burgenae

70

Figure 3.13. Map showing of location of Stupava with other important Roman Period sites indicated.

71

Figure 3.14. Plan of the site at Stupava according to Ondrouch’s excavations (Hečkova 1986, 379: Abb. 1).

72

Figure 3.15. Plan of the site at Stupava after the 1970s excavations.

73

Figure 3.16. Wall-painting fragments from the site. W = white, Y = yellow, O = orange, R = red, and B = Blue (modified image from Bujna 1976, 507: Obr. 8).

Figure 3.17. Photo of the windows in-situ at the site (modified image from Staník and Turčan 2000, 24).

74

Figure 3.18. Reconstruction of the heating system and suspended floors in Building D (based on Ondrouch 1945/46, Obr. II).

Figure 3.19. Reconstruction of canal KV1 (based on Ondrouch 1945/46, Obr. II).

Figure 3.20. Reconstruction of canal KV2 (based on Ondrouch 1945/46, Obr. II).

75

Figure 3.21. Plan of CIII from Stupava (based on Kraskovská 1982, 52: Obr. 1).

Figure 3.22. The site at Stupava with nearby German huts marked.

76

Figure 3.23. Lion-head ornament from a helmet (Staník and Turčan 2000, 26).

Figure 3.24. Examples of Legionary stamps from Stupava (Bujna 1976, 506: Obr. 7).

77

Figure 3.25. Late Roman German pottery, lower one with possible runes on the body (Staník and Turčan 2000, 26).

Figure 3.26. Various finds from the site at Stupava (Staník and Turčan 2000, 25).

78

and counter-maneuver communism, and which regarded the frontier as a physical barrier, strategically conceptualized to be an aspect of a policy designed to defend “us” and to keep the “other” weak (see Cahnman 1949 to understand the concerns of that period; Wells 1999, 126).

Chapter 4 Frontiers, Romanization, and German Identity

Scholarship of the last twenty years has shown that the Roman frontier was more complex, going beyond simple definitions of open or closed, or the idea that the frontier was merely a linear boundary made up of forts and defensive works. The Roman frontier, while having military defenses and having a boundary within it, was at the same time permeable (Hanson 1989, 60; Whittaker 2004, 9). There is ample evidence to show that Romans participated in various actions on a regular basis beyond the frontier, not only in a military context but also civilian, as documented in Chapter 2, while those beyond the border had access to markets and work within the empire. Even more so, the frontier has come to be seen as a zone that straddles both sides of some physical boundary, where the culture there is an amalgam of the larger cultures that share the border, so that the zone is something distinct (Wells 1999, 126; 2005, 50; Whittaker 2004, 5–6).

The following chapter is divided into three main sections: 1) a discussion about frontiers, 2) a look at Romanization, and 3) an overview of German culture, what is cultural identity, and how it manifested among the Germanic Quadi. In relation to frontiers, I shall also deal with the term limes, client kings, and Dio’s “neutral zone.” 4.1. The Roman Frontier Even though the Romans had vast frontiers and most of their resources were tied to maintaining and defending them, there has been a lengthy and at times fierce debate about whether they had a preconceived notion of what frontiers should be and whether there was a defined policy about them. There have been attempts to show that the Romans did in fact have a strategy, such as E. Luttwak’s 1976 publication The Grand Strategy of the Roman Empire, and more recently, but differently, in S. Mattern’s Rome and the Enemy (1999, xii, 81–82). Luttwak sparked a lively debate, because he argued that the Romans had a “grand strategy” for the creation and maintaining of the frontier that evolved over time from a mobile, offensive military line to that of a static linear defense (Luttwak 1976, 4–5, 57, 60, 192–193; James 2002, 29; Whittaker 2004, 6; Luttwak was not alone in this view, see James 2002, 10, 11–12). Ever since his ideas were put forward, various scholars, such as J. Mann, F. Millar, B. Isaac, and C. Whittaker, have tried to show that the Romans did not have a predefined strategy as defined by Luttwak for maintaining the frontier and that the Romans were not entrapped within a mindset of defense (Mann 1979; Lintott 1981; Millar 1982; Whittaker 1989, 64; 1994, 17–18, 35, 62–70; 2004, 28– 29; Isaac 1993; Mattern 1999, 89, 91, 101, 109, 112, 115–116; James 2002, 29).1 In connection with the concept of a grand strategy has been the discussion of whether the frontier was closed, acting as a barrier/linear defense to keep the “barbarians” out, or was open, meaning the frontier was more a marker of where absolute Roman control began and permeable. Early modern scholarship on this topic, such as Haverfield’s in Britain and Mommsen’s in Germany, was steeped in the traditions of nineteenth century European nationalism, imperialism, colonialism, and their frontiers, thus favoring the idea of a closed frontier for the Roman Empire (Whittaker 1994, 2–4; 2004, 6–11; Wells 1999, 125–126; Hingley 2000, 36–37; see also Williams 1996, xviii–xix for a continuation of this view in regards to Rome).2 Some scholars in the middle of the twentieth century, such as E. Luttwak, were influenced by the Cold War, with its clearly defined boundaries, such as the “Iron Curtain,” strategic thinking to compartmentalize

4.1.1. The Meaning of the Term Limes Connected to the idea of the Roman frontier not being a closed linear defense, is the reassessment of the meaning of the word limes. In the nineteenth century, when frontier studies developed, the term came to be accepted as meaning a system of defense along the border of the empire, with permanent forts and road systems that created a hermetically closed border (Isaac 1988, 125). This definition or view of the term limes was influential into the twentieth century in the interpretation and understanding of how the Roman frontier worked. This misconception of the term limes colored the discussions about frontiers as static and too focused on the defenses.3 In fact, B. Isaac and G. Forni’s research showed that the term was used in the first–third centuries as indicating a road network or demarcated land border, with references to military troops only in direct association with the road network; the term was not used to refer to military structures, such as forts, walls, dykes, or towers (Isaac 1988, 146; Hanson 1989, 55; Wheeler 1993a, 27). In the fourth century, the term was broadened to include the designation of a frontier district under the command of a dux, which connotes an administrative concept, not military (Isaac 1988, 146). Isaac went on to show that there was no evidence to support the idea that limes indicated the modern idea of a defended border, with forts and military organization, and this interpretation also redefined the term limitanei, which should be seen as troops under the command of a dux in the frontier district whose duties could take them anywhere in the district and not just having them stationed on the actual frontier itself (Isaac 1988, 146).

79

Wheeler’s critique of Isaac goes too far in my opinion, making a jump in logic when he states, “Isaac, however, wishes to divorce this term from any connection with defense, since limes is not directly equated with military structures in ancient sources. Hence political borders are irrelevant: boundary stones marked provincial borders, but none showed where the Empire ended;…” and he goes on to argue that if we accept Isaac then we also must accept the idea that there were no definite lines of defense (Wheeler 1993a, 27). Wheeler’s mistake is to believe limes, frontiers, and borders are the same creature and that the political boundary was the same thing as the military boundary. I am not advocating accepting all of Isaac’s views, but we do need to acknowledge that the term did not refer to a line of forts and frontier defenses as defined by nineteenth and early twentieth century scholars. This reinterpretation enables us to reject the restrictive view that the Romans saw the frontier as a static boundary of forts and defenses, and it further helps to bolster the view of the frontier as a zone straddling both sides of a boundary. In fact, in terms of military thinking, one can see the position of the legions at the frontier as the most flexible, being equidistant to potential problems further within the empire and beyond.

What needs to be made clear is that “frontier” is different from border or boundary. For too long the terms were used interchangeably, or authors were confused as to which they should have been using in their studies (Hanson 1989, 55; Whittaker 2004, 5). While a border or boundary can be a fixed thing, such as a river, mountain range, or even an agreed line on a map, the frontier extends beyond the border on both sides (Wells 1999, 126). An older definition of the frontier defines it as “a territory or zone of interpenetration between two previously distinct societies” (Thompson and Lamar 1981, 7). Another view is that the frontier was a series of overlapping zones, each different, represented by Roman soldiers, Roman civilians, local natives, and barbarians, who each had their own boundaries made up of political, social, ethnic, religious, linguistic, economic, and military elements (Elton 1996, 4 and 5). In a recent publication, Whittaker sees the frontier as the edge of society, where the boundaries of different societies interact through exchange within social and economic systems, but territory still played an important part, making the frontier a dynamic but ill-defined zone of power (Whittaker 2004, 5–6).

to define how it functioned or how the people within it interacted. While the frontier may be seen as a zone of interaction between different peoples, some scholars still see the frontier as eventually having one group become dominant over the other; this concept is too grounded in a colonialist perspective influenced by a core-periphery approach (Thompson and Lamar 1981, 7; Lightfoot and Martinez 1995, 471). This viewpoint rests on the supposition of the frontier being a place of expansion, where settlers arrive, intermingle, and eventually dominate the indigenous groups.4 The frontier of the United States was the impetus for such a viewpoint, but this concept has been expanded to include the frontiers in South Africa as well as Spanish colonization in the Americas. The American frontier was one which saw settlers moving into a region, where a complex process of interaction took place between them and the indigenous population, eventually leading to the disintegration and impoverishment of American Indian societies (Drummond and Nelson 1994, 5; Elton 1996, 1; Russell 2001, 2; Whittaker 2004, 13). Once the European settlers became dominant at the frontier zone, where the indigenous people may have been exterminated, expelled, incorporated, or a stalemate created, the frontier was considered “closed” and lost its frontier connotation, thereby moving the frontier to territory where this process was still taking place (Thompson and Lamar 1981, 7–8; Russell 2001, 2). Lightfoot and Martinez believe this type of frontier perspective is too heavily influenced by the core-periphery approach, which sees cultural innovation and policy decisions coming from the core to the periphery, which was a passive participant with the indigenous population too often seen as taking a subsidiary role (Lightfoot and Martinez 1995, 475–476).5 They also rightly criticize the over-emphasis of the “dominance and resistance ideology,” which if taken too far can result in one group as being defined as merely subordinate players who respond to another group through passive or active forms of resistance that reenforce the misleading view of emphasizing the separation of groups at the frontier (Lightfoot and Martinez 1995, 483). It can be argued as well that merely labeling somebody in the resistance role automatically creates the existence of a “dominating” partner and creates a bias in the analysis because of the unequal view of the two groups, sometimes also helping to vilify one side (Mattingly 1997, 10; James 2002, 34). A further problem with the idea of resistance has been its use in conjunction with Romanization, sometimes further reinforcing the failed theoretical underpinnings of Romanization (Woolf 1998, 22–23).

Elton and Whittaker’s concepts can be seen as a modified form of Lightfoot and Martinez’s idea that the frontier is a zone of “cultural interfaces in which cross-cutting and overlapping social units can be defined and recombined at different spatial and temporal scales of analysis” (Lightfoot and Martinez 1995, 472). Lightfoot and Martinez argue that the frontier is an interaction zone “where encounters take place between peoples from diverse homelands” (Lightfoot and Martinez 1995, 473). The issue of frontier is complicated by the various ways

The approach advocated by Lightfoot and Martinez and similarly supported by Whittaker, Wells, and Elton, is to see the frontier as a place where multiple ethnic, class, and social groups come into contact and whose varied backgrounds, interests, and motivations provide a fertile ground for dynamic interplay between members of all the groups that “cross-cut traditionally defined boundaries” between them (Lightfoot and Martinez 1995, 483). In conjunction with this new view is the reassessment of what role material culture could play in the interaction of

4.1.2. The Concept of Frontiers

80

all these different groups. The use of specific types of personal adornment, dress, style of house or use of space, for example, can be done in new ways, because of the new culture-contact situations and the identities being either formed or re-enforced in the frontier zone (Lightfoot and Martinez 1995, 485). While new meanings can be created from how specific items may be used, it is also important to keep in mind that some people may have purposefully reinforced “traditionalist identities,” rejecting input from other groups present at the frontier because of complex reasoning, both personal and political (Lightfoot and Martinez 1995, 485). The issue of identity will be taken up later in this chapter (section 4.3).

Even though there were defined boundaries for the Roman Empire, such as the Rhine and Danube Rivers in Europe, the Hadrianic and Antonine Walls in Britain, the fossa in Africa, or the Euphrates River and Via Nova Traiana in the East, we have to keep in mind that the frontier existing along the boundaries extended beyond into non-Roman lands. For the Middle Danube, we should not restrict ourselves to social interaction only at the physical border, which would have been the line of forts and vici along the Danube River, but instead should realize that social interaction was taking place well beyond the border both in the province of Pannonia and in the territory of the Quadi.6 The Romans themselves did not believe their line of forts to be some kind of barrier for themselves, since as clearly shown in the previous chapters, provincial traders operated among the Germans, the military conducted campaigns whenever necessary, and even more important, there was manipulation of Quadi politics to ensure continued peace along the frontier.

Even though the views of Lightfoot and Martinez were in response to frontier studies within colonial America and other regions, the core ideas can be used to help us understand the Roman frontier. It should be noted, however, that the Roman expansion and creation of frontiers was a different process, because there was no massive exodus of Romans from Rome or Italy to the frontier territories comparable to what happened in the United States, and the elimination/expulsion of the indigenous population was not a policy (Whittaker 2004, 13). Instead, the Romans took control of new territory and incorporated the local population into the empire and over time integration of some sort took place between the indigenous population and the “Romans” (see section 4.2 on Romanization; Drummond and Nelson 1994, 33; Woolf 1998, 7–16; Wells 1999, 94 and 122).

As we saw with the use of the term limes above, so too we need to keep in mind the multiple meanings of other terms used by the Romans. Whittaker makes a good point that the term provincia can be both in reference to an area with geographical dimensions, and also to a sphere of power without geographic boundaries. In conjunction with provincia as “sphere of power,” Whittaker adds that when Suetonius said “…nec aliter universos quam membra partisque imperii curae habuit,...” (“… He never failed to treat them all with consideration as integral parts of the empire,…” Aug. 48) in reference to client kings, the author meant that these kings were a part of the empire’s sphere of power, but not of the administered provincia (Whittaker 1989, 65; 1994, 17; 2004, 40).7 Therefore, we need to keep in mind the Romans may have had two ideas of empire, an “empire of administration” and an “empire of control,” whereby the one of control could extend well beyond the established boundaries of the empire (Whittaker 1989, 65; 1994, 17).8 With the well documented Roman actions beyond the Danube in Slovakia, we can see that the frontier region in southwest Slovakia, which included the Quadi, was within the “empire of control” for the Romans. The frontier region was, in the eyes of the Romans, their territory, but which they allowed to remain under the administration of non-Romans. Whittaker is of the opinion that the Romans believed in “three bands of space:” the administered territories, the unadministered territories under Roman rule, and the outer periphery (Whittaker 2004, 40).9 We see these distinctions in Greek authors such as Plutarch, when they use the term α̉ρχή (hereafter arche) for instances of direct rule and η‛γεμονία (hereafter hegemonia) for indirect rule when discussing the workings of the Roman Empire (Whittaker 2004, 40). This view is also supported by Braund, who believed that frontier kingdoms were both outside of and part of the Roman Empire (Braund 1984, 182).

Roman control of newly acquired territory was carried out by different means depending on the situation. If there was a power structure in place that could be realigned to that of Roman needs, then it was kept, such as having client kings, which was done during the Republic with Hiero II of Syracuse, and by the time of Augustus, Juba in Mauritania and Herod in Judea for example (Braund 1984, 182; Drummond and Nelson 1994, 222; Burns 2003, 190–191). In other instances, territories could be reorganized under the direct control of Roman governors and their staff (Hanson 1997, 72). There was also the preferred two-tier system, with a Roman governor at the top who used indigenous elites to maintain control at the local level (Hanson 1997, 75; Burns 2003, 144–145). However, archaeology has also shown that even if local populations were technically within the empire, sometimes their material culture did not show a marked change, thus putting into question just how much integration actually took place (Woolf 1998, 8, 136, 158; Wells 1999, 94–95, 131). There was no Roman program of forcing locals to adopt Roman values, religion, or even fashion, as seen in the Americas (Saddington 1991, 413; Wells 1999, 130). The core idea from Lightfoot and Martinez that needs to be kept for use with the Roman frontier is to see the frontier as a place where people of varied background, language, class, belief, and ethnicity intermingled, whereby the old social boundaries were malleable and could be cut through to create new social links and identities.

The boundary of the arche was where one saw the beginnings of Roman defenses, in order to protect the Roman administered territories, but also to regulate the 81

movement of peoples from beyond into the provinces. This defense and movement control can be seen in the creation of Hadrian’s Wall, the fossa in Africa, and even in the density of forts along the Rhine and Danube. Wheeler in his critique of Whittaker and Isaac uses the existence of these fortifications as evidence that the Romans did have a set idea for the borders of the empire and also refers to Pliny’s comment that Pompey fixed the eastern border of the empire at Oruros east of the Euphrates (Plin. NH. 6.30.120; Wheeler 1993a, 29). If we accept the existence of arche and hegemonia however, then Wheeler is confusing the two types of territory that made up the entire empire. The territory that fell outside, within the realm of the hegemonia was left to be defended by the rulers there, but the Romans felt it within their right to interfere whenever they wanted to either militarily or politically. The Romans were well aware of the breakdown of relationships that could happen between themselves and those in charge of the lands along the borders of the empire and also the breakdown between the ruler and his own subjects in those lands, so that defenses were needed. Wheeler argues that Whittaker and Isaac go too far in demilitarizing the frontier and creates the impression that they and their followers deny or ignore the military aspects of the frontier (Wheeler 1993a, 16–17). Not so; Wheeler has confused the concept of frontier and boundary.10 No one can deny the militarization of the borders of the Roman Empire, but these need to be placed within the context of frontiers and how the Romans viewed not only the territories they directly controlled, but also the lands that bordered them. Again, this view reinforces the idea that the placement of the legions at the border was the most efficient method to deal effectively with both internal and external problems for the empire. The legions at the frontier were not a symbol of a hard external shell merely positioned facing outward (see Williams 1996, 38–39), but instead the legions were poised looking inwards and outwards (especially true when we keep in mind how many times the frontier legions marched into the empire to secure the power of new emperors or to crush other legions in mutiny).11

usually seen in modern scholarship within the framework of patron-client relationship (which actually led to the creation of the term “client-king”) (Lintott 1981, 61; Pitts 1989, 46). The term “client-king” and seeing kings within the patron-client relationship may be misleading, since there is no instance in Roman literature of these kings being referred to as cliens (Pitts 1989, 46; also see Braund 1984, 6–7, 23–27). Pitts makes a good point that Maroboduus viewed himself as equal to the emperor in power, which was noted by Velleius; but I would add it is clear that Rome did not view Maroboduus as an equal, as shown in Chapter 2 (Velleius himself thought the German king was grossly mistaken to view himself as an equal) (Vell. Pat. 2.109; Pitts 1989, 46). Even though some kings may have been treated as true equals (such as Ptolemy Philadelphus or Attalus I), there are many instances where it was clear Rome was in the superior position and the king was an inferior (Antiochus IV, Attalus II, Prusius of Bithynia, or Vannius) (Jones 1970, 109; Lintott 1981, 62; Braund 1984, 66–67).12 By the mid-first century AD, many of the Republican period client-kingdoms were gone and their territories absorbed into the empire, while those still in existence were even more dependent upon Rome’s favor (Lintott 1981, 62). This situation continued in the imperial period, with the power of client-kings more and more an illusion and with the Romans behaving as if these kingdoms were mere provinciae, so that if the king fell out of favor or was ruling poorly in the eyes of Romans, he could be replaced as one would replace a bad governor of one of the administered provinces (Lintott 1981, 63; Braund 1984, 66–67, 85).13 We see this result with the fall of Maroboduus and Vannius along the Danube, for example, and in the third century, when Zenobia became too independent and was removed by Aurelian. These kingdoms along the borders of the empire should not be seen as simple “buffer-states” but as part of the Roman imperium and one aspect of imperial administration (Braund 1984, 91, 182; Kolník 1991, 432; Hanson 1997, 70).14

4.1.3. Client Kings

4.1.4. Roman Authors, Views on Imperialism, and the Instructions of Augustus

Client kings were not imperial phenomena; examples go back into the time of the Republic, such as Ptolemy Philadelphus, Attalus I, Demetrius of Pharos, and Masinissa (Lintott 1981, 61; Braund 1984, 9, 10). Kings could have been named as friends of Rome already ruling an established kingdom (such Ptolemy and Attalus I); they could have been kings defeated by the Romans in war and then treated as clients (such as Philip V of Macedon and Antiochus III of Syria); or they were created by the Romans as kings (such as Demetrius of Pharos, Masinissa, or Vannius) (Lintott 1981, 61). These kings were called reges amicique populi Romani or reges socii et amici populi Romani; the title of friend was officially given by the senate (Lintott 1981, 61; Braund 1984, 23–24). The relationship between these kings and the senate during the Republic and later the emperors is

The concept of there being an empire of control, the hegemonia, fits in well with what the Roman authors of the Republican and Imperial periods were writing in reference to imperial aspirations. For example, in the period of the late republic, Cicero stated, “the world is already contained within our imperium” (Cic. Rep. 3.15.24; Whittaker 1994, 28). This belief that the Romans rule the whole world, or should rule, is reinforced by the Augustan-period authors: consider Ovid’s statement in the Fasti, “Gentibus est aliis tellus data limite certo/ Romanae spatium est Urbis et orbis” (“To other people land may be given with a fixed limit/ But for the city of Rome its space is the same as that of the world”), or Vergil in the Aeneid having Jupiter assure Rome that it will have “imperium sine fine”—an empire without end (Ov. Fast. 2.667–684; Verg. Aen. 1.278–279; Lintott 82

1981, 53; Whittaker 1994, 29; Williams 1996, 5; Mattern 1999, 169–170).15 These and other types of statements show that the Romans did not confine their idea of empire within territory they expressly controlled, but considered it to exist even over remote lands, which is further seen in various geographical works such as the Cosmograpia of Julius Honorius or the Cosmographia of Pseudo-Aethicus, who not only list real Roman provinces but also regions of the world such as Africa below the Sahara or India as provinces of the Roman Empire (Whittaker 1994, 14–15; 2004, 40).

action between the Romans and Parthians across the span of many principates. It is true that Hadrian relinquished control of the Mesopotamian provinces and Armenia, but this act should be seen as a decision influenced by the circumstances of that region, where a comparable imperial power, the Parthians, were regaining their strength and Mesopotamia had been a traditional part of their territory, while Armenia had usually existed as a client kingdom.17 4.1.5. Germans and Soldiers on the Middle Danube Frontier

These views are important to keep in mind when considering the passage from Tacitus where Augustus left instructions to Tiberius stating “…that the empire should be kept within its boundary” (Tac. Ann. 1.11).16 This statement has been interpreted by many scholars to mean that Augustus advocated the abandonment of traditional Roman procedure, which held the frontiers as open, expansionist, and zonal; with Tiberius onwards, emperors supposedly followed a new policy that saw its apogee with Hadrian when the frontier evolved into that closed, linear barrier already discussed (the view supported by Luttwak and his followers) (Whittaker 2004, 6 and 25). Again, the example of the Middle Danube frontier from AD 10–166, as discussed in the last chapters, shows that emperors did not believe the frontier was a linear barrier, with Roman activity in Slovakia occurring continuously from Augustus to Marcus Aurelius. Whittaker argues that Augustus was not referring to some military frontier or to the whole empire, but instead to the boundaries of the organized provinces, in effect stating, “keep the arche where it is,” implying no restrictions on those territories seen by Rome as her sphere of influence, which was already mentioned as falling within the hegemonia (Whittaker 1994, 25; 2004, 41). Whittaker sees Augustus as providing tactical advice not strategic, that Tiberius should not seek to increase the empire, while many provinces were not sufficiently “Romanized” to allow further expansion (Whittaker 1994, 29–30, 35). Augustus may have been remembering the Pannonian rebellion of AD 6 and the disaster of Varus in the Teutoburg Forest in AD 9 (Whittaker 1994, 35). Varus may have been the foremost in mind, especially when we remember Velleius, remarking that Tiberius returned to Rome in 8 BC for his second triumph having left Germany almost as a tribute-paying province and that Varus in AD 9 was exacting tribute and holding assizes as if it was a pacified province (Vell. Pat. 2.97.4; 2.117.3–4; Lintott 1981, 60). Whether one agrees with Whittaker’s view or not, it cannot be denied that Augustus’ advice was not followed by succeeding emperors, as we see below.

In regards to the Middle Danube, while the Romans saw southwest Slovakia as part of their sphere of influence, the Germans living there saw Pannonia as a place of opportunity. Dio records that in the first truce (ca. 169/70?) established with the Quadi during the Marcomannic Wars, the Romans forbade the Quadi from coming to the provincial markets out of fear that Marcomanni or Iazyge agents may pass themselves off as Quadi and reconnoiter Roman positions for future attack (Dio Cass. 72.11.3). This passage indicates that the Quadi had access to the provincial markets on a regular basis before the wars; literary evidence also shows, however, that German movements in Roman territory was restricted and supervised (Thompson 1982, 12). Even though Roman traders did operate amongst the Germans, we have to imagine that some goods were probably only available at the markets or were available there for a better price/trade. As to whether Germans used coinage, we know that they did not mint coins as the Celts had done earlier, and some Roman coins found in German territory had holes drilled in them for apparent stringing, which would indicate their being used as commodities of exchange through their weight in metal (or they were simply used on necklaces for ornamentation). Some scholars see the evidence of gold and silver Roman coins found on German territory as evidence of a monetized economy, but not everyone accepts this viewpoint (Drummond and Nelson 1994, 109). So it would seem that barter was the mainstay of the German economy, and probably as well as when they came to the provincial markets or dealt with traveling Roman merchants, but again there is some evidence to show that Roman traders preferred payment in coin, so Germans may have been forced to use coins for purchases along the frontier as well (Drummond and Nelson 1994, 110). Dio also mentions in regards to the Marcomanni that when they had agreed to peace and fulfilled their obligations, Marcus Aurelius established the places and days for the Germans to meet/conduct their trading with each other (Dio Cass. 72.15). This passage would indicate that the Germans conducted market-days where one could barter for goods, but in this case the Romans had established fixed places and days probably for security reasons, allowing the Germans to be monitored. While bartering/trading was important, we have to consider also the role of gift-giving in the context of a parallel redistributive economy, where the war-leaders and kings would have distributed largess to the people through

Future emperors that fall within the period covered by this book (10 BC–AD 166) certainly did not believe in being satisfied with the borders as they were, with Claudius taking Britain, Domitian campaigning against the Dacians and Germans, Trajan adding Dacia, Arabia, Assyria, Armenia, and Mespotamia as provinces, and Lucius Verus campaigning against the Parthians after Vologaeses IV placed their own choice of king on the throne in Armenia, which was a place of contentious 83

feasting; at the same time gift-exchange probably took place between households as well (this topic is taken up in section 4.3 in relation to German identity).

venison (Drummond and Nelson 1994, 85). Approximately 5,000 tons of meat were consumed annually by the armies in Europe (calculated at minimum consumption of one pound of meat per week per man; Drummond and Nelson 1994, 86); with the Pannonian forces consuming 1,210 tons of that total. In order to appreciate the number of animals needed to meet this requirement, I present here some figures. At the frontier, the legionary camps probably had a mix of animals procured from Pannonian sources as well as German ones beyond the frontier. German animals were around half the size of their Roman counterparts (Kreuz 1999, 82 and Fig. 3.5) and this difference needs to be taken into consideration when creating estimates for amounts of animals consumed. M. MacKinnon calculated the average meat weight for Roman cattle at 200 kg, pigs at 50 kg, and sheep/goats at 27.5 kg (MacKinnon 2004, 190), so that German counterparts would be 100 kg for cattle, 25 kg for pigs and 13.8 kg for sheep/goats. 18 I split the tonnage in half, with the first half being cattle (605 tons) and the remaining half split between pigs (303 tons) and sheep/goats (302 tons). Of these amounts, I assume that a third of each animal type was procured from German sources, so that of cattle, 404 tons came from Pannonian farms and 201 tons from Germans ones, 202 tons and 101 tons respectively for pigs, and 202 tons and 100 tons respectively for sheep/goats. In terms of actual numbers, the legionary forts possibly consumed annually 1,832 cattle from Pannonia and 1,819 from Germania, 3,673 pigs from Pannonia and 3,673 from Germania, and 6,623 sheep/goats from Pannonia and 6,667 from Germania. What is immediately apparent is that the numbers are almost equal because of the smaller size of the German counterparts, so that in fact, while German animals made up a third of the meat weight, they in fact made up half the procured animals. These numbers are large when we also consider the consumption of the civilian population attached to the legionary camps, as well as the population in the countryside. There is ample evidence to show the development of agricultural production in Pannonia itself, but we should also look to the territory along the frontier in Slovakia as also being part of the supply network for the legions (Drummond and Nelson 1994, 61, 63–64, 114, 119–120).

In regards to what was being purchased/bartered for at the provincial markets, we would probably expect the Germans to bring skins, food-stuffs, other raw materials, and slaves to be used to acquire finished products, such as bronze cooking vessels, jewelry, pottery, and food items not made by Germans such as olive oil, wine, or even garum (Cunliffe 1988, 9–10; Drummond and Nelson 1994, 11–12, 110; Wells 1999, 255; Todd 2004, 95; Whittaker 2004, 17). The Germans were an important component of the frontier economy along with the legionary soldiers themselves (Whittaker 1989, 67; Cunliffe 1988, 10; Drummond and Nelson 1994, 11–12, 29, 95, 110–111; Wells 1999, 134, 235, and 257). We have to remember that by the time of Domitian, the legions, auxiliaries, and their associated vici were all positioned along the Danube. By AD 106, there were 48,392 men stationed in Pannonia along the Danube, of which 28,392 were in Pannonia Superior and 19,760 in Pannonia Inferior (Krekovič 2001b, 105). These numbers, however, do not include the slaves attached to the forts (2,000 are postulated for Vindonissa in Switzerland for example), as well as the camp followers and associated merchants/businessmen who set up shop in the vici outside the legionary and auxiliary bases (Whittaker 2004, 91). All of these troops and their dependents needed to be fed and clothed, and this fact would have caused strains on the provincial farms and population, so that the agricultural production of the Germans of southwest Slovakia would have been important to the maintenance of the legions. A single legion needed more than two thousand tons of wheat per year for example (Wells 1999, 142). The number of soldiers stationed in Pannonia given above is close to the estimated 50,000 stationed in Britain during the later first and early second century AD (Breeze 1982, 27), and so comparative rates of consumption can be expected. Calculations done by Drummond and Nelson have shown that the force in Britain needed a total of 530,000 bushels of grain a year for men and horses; this amount would have required 800 square miles of land cultivation and 10,000 draft animals for plowing (Drummond and Nelson 1994, 52–53). The legions also needed constant supplies of iron, bronze, wood, and leather to outfit and maintain the soldiers’ weapons and gear and the fort. P. Wells provides a figure of 54,000 calves needed just to supply the leather required by a single legion for their tents used on campaign (Wells 1999, 145). Drummond and Nelson provide a figure of 750,000 calves needed for tents by the frontier armies stationed in Europe (approx. 200,000 men; Drummond and Nelson 1994, 80), so that the Pannonian troops would have needed 181,470 calves, which does not include cattle used to supply the leather needed for other equipment.

We know that along the Rhine, German settlements, such as Feddersen Wierde, beyond the frontier appear to have realigned their food production to supply the legionary bases, providing cattle and grain (Drummond and Nelson 1994, 114; Wells 1999, 244–245; Whittaker 2004, 17– 18). All of these requirements and the ready availability of money made the forts important economic factors at the frontier (Drummond and Nelson 1994, 9, 29; Wells 1999, 134, 143–144). It is completely possible that the density of German population and their apparent prosperity in southwest Slovakia was in part a result of their integration into the frontier economy and the maintenance of the military bases that ironically were there to keep these Germans in check. Chapter 10 provides a discussion of animals raised by the Germans along the Middle Danube frontier, and shows that cattle dominates the assemblages, making it seem that here too,

Meat was also an important part of legionary diet, comprising not only beef, but also pig, goat, and some 84

Germans were raising cattle for Roman markets (but there are interesting hints that pigs were important as well). Dio records that part of the peace treaties signed between the Romans and Quadi included the transfer of cattle and horses to the Romans (Dio Cass. 72.11.2).

the empire, and they brought all this back to the barbaricum (Wells 1999, 255). These phenomena, however, encroach on the topic of Romanization, which is taken up below.

At the same time, legionary bases were also centers of production in specialized items along the frontier (Krekovič 1997a, 234). The soldiers had multiple duties and only a small part of their time was actually spent fighting. Soldiers could take part in road building and construction, while some were specialized craftsmen, such as blacksmiths, sword-makers, coppersmiths, carpenters, leather-workers, tanners, or potters just to name a few (Wells 1999, 133–134; Whittaker 2004, 90). Much of what the legions needed in terms of manufactured goods was made on the spot at the fort or contracted to local producers (Wells 1999, 142–143). By the second century AD, Pannonia had important Roman/provincial pottery production where the legions were based, such as at Brigetio, Carnuntum, or Aquincum, the pottery of which not only went to the provincial markets but beyond into Quadi lands (see Chapter 9.1 for more information on Roman provincial pottery in Pannonia and Quadi territory). Legions produced roof tiles used not only on military structures but also on civilian ones as well, and it appears these tiles were allowed to be used on structures beyond the frontier (as seen in the previous chapter).19

4.1.6. Comments on Dio’s “Neutral Zone” In connection with the forts along the Rhine frontier there was apparently a strip of land in what would technically be the barbaricum that the Romans claimed as part of the forts’ area. This zone was not allowed to be occupied by Germans, but instead to be left open for the use of the forts in maintaining herds of cattle and horses needed by the soldiers. Tacitus provides evidence for this when he recounts that the Frisians moved their population to the Rhine bank around AD 58, settling in the clearings reserved for the use of the troops – “… inbellem aetatem per lacus admovere ripae agrosque vacuos et militum usui sepositos insedere…” (“…those not of military age by the lakes. Here they settled in the clearings reserved for the use of the troops,…”); later he then adds that after the Frisians were forced to vacate the land, the Ampsivarii moved into it and an advocate for the tribe, Boiocalus, made a speech to the Romans stating, “…Why should such an extent of clear ground lie waste, merely that on some distant day the flocks and herds of the soldiers could be brought over to it? By all means let them keep reservations for cattle in the midst of starving men, but not to the extent of choosing a desert and a solitude for neighbors in preference to friendly nations!” (Tac. Ann. 13.54, 55, trans. by J. Jackson).21 Whether such a zone existed as a contiguous space along the entire Roman frontier is unknown, but has been postulated for the Danube frontier at least (Potter 1992).

Germans could have also been hired as day laborers or migrant workers for various jobs in the provinces (Whittaker 2004, 17). The best known examples of employment, however, are the Germans who served in the Roman military, but being peregrini, they would have been in the auxiliary forces (Fitz 1980, 135; Wells 1999, 135, 238, 257). These foreign elements in the Roman armies would have been sent to other parts of the empire for their service, far from their homeland to ensure loyalty. An example on the Middle Danube frontier is the auxiliary force stationed at Gerulata, which had a fort and vicus, located around 10 km from Carnuntum. The force stationed at Gerulata in the first quarter of the second century was the Ala I Cananefatum, which was a cavalry force made up of Germans from Cananefati, near Ryna, Holland (Kolník et al., 1993, 222).20

We do not know how large or continuous this strip of land was, but it appears to have been an important part of the fort’s zone since the Romans made sure these lands were kept empty of Germans (Potter 1992, 272–273). The possibility of this zone being large is supported by the fact that small German tribes wanted to live there, implying substantial amounts of land (Potter 1992, 273). It cannot be discounted that these lands had a dual purpose of not only serving the herds and flocks belonging to the forts, but also as a defensive measure, keeping land cleared in the forward defense zone of the fort. Keeping a strip of land empty of settlement would have helped Romans enforce agreements with the local populations, such as restrictions to attend markets or to use the waterways (Pottery 1992, 273).

Most of the best-known examples of individual Germans serving in the Roman military come from the later periods, such as the third to fifth centuries AD, but there are known instances in the first to second centuries as well (Todd 2004, 59–61). A well known example was Arminius, who gained Roman citizenship and equestrian status while probably serving as commander of an auxiliary unit; he then returned to his homeland on the Rhine and incited rebellion in AD 9 (see Chapter 2; Wells 1999, 230–232). Arminius is an extreme example and most Germans who returned to their homelands settled down peacefully. However, it is important to keep in mind that these Germans had lived in various parts of the Roman Empire, picked up new ideas, some Latin or even Greek, had weapons and accoutrements produced within

I mention this strip of land because, in regards to the Middle Danube, Dio has a well known passage where the Marcomanni (and based on later wording, the Quadi too) were given back half of the neutral territory along the Danube, allowing the Germans to settle within five miles of the river (Dio Cass. 72.15–16). It is possible then that the “neutral zone” could range in size from 5 to 10 miles in depth. In Dio’s account of Commodus’ principate, he goes on to add that the emperor made peace with the Marcomanni and Qaudi, abandoning all the posts amidst 85

them that had existed beyond the strip of land along the frontier that had been neutralized (Dio Cass. 73.2.4). L. Pitts argued that this neutral zone was a short-lived creation, formed by Marcus Aurelius and probably abandoned sometime after Commodus (Pitt 1989, 50–51). D. Potter believes the opposite to Pitt, basing his argument on the Tacitus passage given above, that the neutral zone was an established part of Roman frontier policy and permanent (Potter 1992, 269, 270–271). He also makes a good point that within this zone, various watchtowers could have been set up to watch the frontier and document the movement of Germans, while boundary stones were laid out to inform the Germans when they crossed into this zone; he also makes a compelling case that the maintenance of this zone was vital to the Romans to project the image of power, whereby they indicated their strength by being able to maintain such a zone (Potter 1992, 272 and 274).

Xanten on the Lower Rhine the land is postulated to have been 11 km by 3 km (Whittaker 2004, 15 and 101).23 We have little information about this type of attached land and it is clear that not all forts had it (Whittaker 1994, 83–84; 2004, 101). In other instances, where the territorium did exist, it may have been rented to locals and not left fallow (Whittaker 2004, 101). The passage from Tacitus is most likely merely referring to the territorium of one particular fort and not necessarily to a whole zone along the Rhine River. Since this land was used to help feed the fort, keeping it clear of German settlers may have been important to the Romans. It is entirely possible then that there was no strategic thinking in the use of these territoria for neutral zones, with the land having only agricultural value not military. The forts along the Danube probably had territoria, but they would have existed separately from each other, being specifically set aside for each fort. I believe it is possible that in the creation of the first peace treaty with the Marcomanni and Quadi, Marcus Aurelius and his advisors possibly united the territoria of the various forts to create a temporary neutral zone, providing time for the preparation of the later major campaign that ended with the second peace treaty, in which the zone was reduced by half.

The archaeology, however, does not support the views of Potter. It appears that the neutral zone mentioned by Dio was something different from the zone discussed by Tacitus. While forward forts have been found opposite Carnuntum, Brigetio and other sites, these are seen as forward protection for river-crossings and the legionary forts on the Roman side. No watchtowers or boundary stones have been found in southwest Slovakia within five to ten miles of the Danube. This is not to say they may not be discovered at some future date, but none has been found so far.22 Also, German sites have been found within this netural zone, such as Devín (right on the Danube), Devinska Nová Ves (2.79 miles), Dúbravka (2.5 miles), Bratislava-Trnávka “Zadné” (3.7 miles from the Danube), Ivanka pri Dunaji (4.9 miles), and Štúrovo (right on the Danube) to name just a few.

4.2. Romanization In recent years, the term Romanization and its uses have come under closer scrutiny by various scholars from varied backgrounds. Much like with the concept of frontier, it has become clear that the traditional definition and views of Romanization were grounded in the nineteenth and early twentieth centuries, influenced by the nationalism and imperialism of various European powers and the United States (Mattingly 1997, 8). The term “Romanization” has been used traditionally as an all-inclusive word to describe the process whereby indigenous people in various parts of the Roman Empire and even beyond were “incorporated or aligned” with the empire (Barrett 1997, 51). The very structure of the word implies a “unilateral transfer of culture” and leads to over-generalizing, as in the statement “the people of Britain were Romanized” (Mattingly 1997, 8–9; Wells 1999, 126). There has been a belief of uniformity and homogeneity within the empire, so that when one looked at indigenous people in Britain and Syria for example, Romanization for each was seen as the same process (Freeman 1997, 43; Hingley 2000, 117). In fact, Romanization has been seen as an integral part of Roman expansion, as if the primary goal of the Roman Empire was Romanization (Freeman 1997, 27; James 2002, 34, for the role of the Roman legions in this view).

The mention of the neutral zone and its reduction occurs in a peace settlement from around 174/175, but after Marcus Aurelius had previously defeated the Marcomanni in a different battle, around 172 (Dio Cass. 72.3,5; 72.8.1). The peace terms after the first Marcomannic defeat are not named, but may have included the creation of the 10-mile-wide neutral zone, which was then reduced in the second peace settlement. We cannot say, therefore, with any conviction as to when this neutral zone was created, whether it existed right from the start of the frontier’s inception along the Danube under Augustus or was it something particular to the period of Marcus Aurelius. Also, can we be sure that the neutral zone mentioned by Dio is actually the same as the land mentioned by Tacitus? Some forts did have land attached to them, usually referred to as territorium or prata (Millar 1981, 121; Drummond and Nelson, 89; Whittaker 2004, 15). This land was used for agricultural purposes and from the Vindolanda tablets we have the title of some of the men who worked this land, such as bubulcaris or ad porcos, who obtained rations from the camp (thereby indicating they were employed by or part of the fort) (Whittaker 2004, 15). The territory could be quite large, such as thirty to seventy km2 for Caerleon in Wales, while for

These concepts have come under serious challenge and Romanization is now seen as a complex process that definitely was not unilateral, uniform, or even a primary goal of the Romans (Mattingly 1997, 9). In terms of material culture, the presence of Roman artifacts has been seen in the past as a sign of Romanization if found in 86

non-Roman context, but it is now understood that many of these items were used within the context of the indigenous culture, providing new meaning to the artifact (Jones 1997, 130 and 134; Woolf 1998, 15).24 At the same time, the very word “Roman” has come under scrutiny (Barrett 1997, 51; Wells 1999, 127). Too often in provincial or frontier context, interaction is seen as taking place between “Romans and Barbarians,” but this kind of thinking totally ignores the complicated situation outside of Italy (and even there too; Woolf 1998, 7). The provinces and frontiers had many types of people, not just a few representative individuals who actually came from the city of Rome or Italy, but more often indigenous people, who were now considered part of the empire, interacting with people from other parts of the empire seen in the form of soldiers, merchants, priests, or just workers to name a few examples.25 Roman interaction in the provinces is in reality a complex multi-ethnic/cultural interplay. What has to be realized is that people could identify themselves as “Roman” in some context out in the provinces, but the same persons in Rome itself would probably be seen as provincials or barbarians, because, as Woolf rightly reminds us, while the adoption of Roman culture could be a unifying process there were also what he terms “cultures of exclusion” in operation at many levels of society, wherein the lack of appropriate grammar or pronunciation of Latin or not wearing the latest style could be used to differentiate and exclude (Woolf 1998, 19). Perspective is important when one talks about Romanization, especially in the context of identity, because a person’s identity is often defined by what others perceive and not just what the person thinks is being projected as the persona. Romanization has come to be seen by some as an inadequate term to define what is happening at the frontier, and for others even for the provinces as a whole (Woolf 1998, 7).

absorption; 2) Roman expansion was seen as favorable during the Early Empire, leading to peace, consolidation, prosperity and orderly government throughout the lands incorporated; 3) Emphasis on common themes seen in all the provinces, such as the spread of Roman institutions, coinage, architecture, material goods, which created the view of homogeneity (Freeman 1997, 30–31; Hingley 2000, 113). These ideas helped to create the view of Romanization described in the previous section. Mommsen’s views were grounded in the political realities taking place in the nineteenth century, with the unification of Italy under Emmanuel, Germany under Bismarck, and the expansion of empires (Freeman 1997, 33–34; Wells 1999, 125–126; Hingley 2000, 22). In England, while some scholars rejected the views of Mommsen, others embraced them, especially H. Pelham (Freeman 1997, 36–37).27 Pelham expanded or worked off the ideas set forth by Mommsen (Pelham 1897; 1906; Freeman 1997, 37). The reason Pelhman is important is his mentorship of F. Haverfield, one of the most influential British archaeologists in the early twentieth century, with some seeing him as the founding father of modern Romano-British archaeology (Freeman 1997, 37– 39; Hingley 2000, 12).28 Haverfield applied the views of Mommsen to archaeology in Britain, continuing the ideas of defensive imperialism, Roman expansion was beneficial, and there was homogeneity in the Roman world (Hingley 2000, 114; see also James 2002, 16–17, on how British military studies were also heavily influenced by German tradition). Haverfield also introduced the concept of “binary opposition between barbaric native…and civilised Romans,” whereby Romanization was the process by which the uncivilized Briton achieved civilization (Hingley 2000, 4). Haverfield explained his ideas in numerous publications, but one of his most important was “The Romanization of Roman Britain” (1905), which had multiple publication runs in book form, and became influential in solidifying the traditional definition of Romanization (Hingley 2000, 114, 130–143). This view of Romanization was reflected in the attitudes toward British imperialism and its beneficial effects on indigenous peoples (Bryce 1914, 25, 544-55; Chamberlain 1972, 65, 68–69, 85; Hingley 2000, 4; James 2002, 8–9; see also footnote 1 of this chapter and the quote from Winston Churchill). Romanization then was seen as a natural product of imperialism, which can be viewed as the “practice, theory, and attitude of a dominant metropolitan nation or people in establishing control over and ruling another nation or people” (Hingley 2000, 7).29 An interesting irony is that Haverfield himself did not approve of using classical scholarship and Roman examples to reinforce contemporary imperial attitudes (Freeman 1997, 46), but the way in which he wrote (see footnote 23 above) and how his works were used were rife with examples of imperial thinking; Hingley disagrees with Freeman’s view and believes Haverfield did in fact support the use of Roman examples to better run British imperialism (Hingley 1997, 83; footnote 11). Comparisons were definitely made between modern and ancient empires in the late 19th and early 20th centuries, with Lord Cromer

4.2.1. The Origins of Romanization P. Freeman and R. Hingley have shown quite effectively the origins of Romanization and its use in many historical and archaeological works of the late nineteenth and early twentieth century (Freeman 1997; Hingly 2000). The way in which the word has been used and conceptualized has to be placed within a modern context (Mattingly 1997, 17; Hingley 2000, 111). Three nineteenth and early twentieth century scholars, T. Mommsen, H. Pelham, and F. Haverfield, were the principal players in conceptualizing and transmitting the traditional views on Romanization.26 T. Mommsen in his works Römische Geschichte (1854– 1885) and The provinces of the Roman Empire (1885, the fifth book of his Römische Geschichte) provided an influential foundation in conceptualizing the growth of the Roman Empire (Freeman 1997, 29). He created three concepts: 1) Defensive/Reluctant Imperialism, in the idea that Rome did not want an empire, but was forced to expand by circumstance, whereby Rome became embroiled in the politics of other regions not by choice but by fault of her neighbors, which led to their 87

and James Bryce’s essays the most prominent (Cromer 1910; Bryce 1914).

provides a very important example when he mentions a common belief that natives adopted Roman wheel-made pottery because these vessels were superior to native ones. This assumption in fact ignores the evidence showing many Roman pottery forms common in the provinces were based on native prototypes (Hingley 1996, 42) (for more detail see the discussion of Roman Pottery and Celtic influence along the Middle Danube frontier in Chapter 9). We also need to keep in mind how the native society possibly redefined the function, value, and role of foreign goods (Hingley 1996, 42; Jones 1997, 134). Since the empire was so large and filled with many cultural groups, there was no one standard Roman image or even unified material culture package (Hingley 1996, 42; 1997, 85–86; Woolf 1998, 7). Many items taken to be symbols of Rome were in fact manufactured in the provinces, such as the Gallic and North African versions of terra sigillata or various bronze vessel shapes that originated in N. Italy but were then copied and produced in Pannonia (Wells 1999, 127–128).

The static linear-defense view of the Roman frontiers was an outgrowth of the nineteenth- and early twentiethcentury view of Romanization (Hingley 1996, 38). The frontier was the wall against the barbarian, so that within this protection civilization could flourish, which is expressly meant by Haverfield when he says, “The greatest work of the imperial age must be sought in its imperial administration – in the organisation of its frontier defences which repulsed the barbarian, and in the development of the provinces within those defences … In the lands that [Rome] had sheltered, Roman civilisation had taken root” (quote reproduced from Hingley 2000, 37, taken from Haverfield 1905, 185–186).30 4.2.2. The Concept of Romanization Today With the rejection of the traditional view and the introduction of a post-colonial approach, it is possible to still use the term Romanization, but within a more holistic framework. I use the term holistic to mean that while Romanization may in fact have been taking place for some people, we need to keep in mind that other cultural events were taking place, such as Germanization at the Middle Danube frontier in parallel with Romanization. This approach fits with Woolf’s view of Romanization as shorthand for a “series of cultural changes that created an imperial civilization, within which both differences and similarities came to form a coherent pattern” (Woolf 1998, 7). For the provinces, where there was such a mixing of different groups of people, one can argue that after the Roman conquest, the culture that was created was neither wholly Roman or unaffected indigenous, but a mixture of the two (Wells 2005, 50 and 95). This holistic approach fits in with the three interrelated aspects of the new approach to Romanization discussed by Hingley (1996, 40). He introduces three elements of post-colonial study to be used in conjunction with Romanization: 1) de-centering, 2) showing the “complex and varied range of responses to colonial contact,” and 3) acknowledging the “overt and covert opposition to domination by colonial powers” (Hingley 1996, 40).

When we contemplate the varied reactions to colonial contact, we need to avoid the pitfall of assuming that the locals were passive or that they all had the same reaction to the Roman presence. There has been a long-held belief that the local elites adopted Roman goods and concepts for their own needs and were then passed on to a passive lower class (Hingley 1996, 41–42; 1997, 84). The problem of this viewpoint has been the archaeological focus on elites and ignoring the lower classes (Hingley 1996, 43; 1997, 85). Hingley makes a good point that all levels of native/provincial society need to be explored because each group probably had a different reaction to the Roman presence. For the study of the Germans in southwest Slovakia, the archaeological data that have been accumulated are a mix of all levels of society. The cemeteries are a mixture of rich and poor, but the settlements are almost totally from the lower strata of society, with little evidence for elite residences (outside of the Roman-style structures, which are seen as possible elite residences). When new territories were incorporated into the Roman Empire or groups along the frontier were exposed to Roman goods, we should look for signs of opposition or at least be aware that they may exist. The opposition need not have been outright revolt through arms but more subtle, such as the purposeful maintenance of traditional pottery, house construction, or food preparation (Hingley 1996, 42–43; 1997, 96–97).31 The indigenous people need to be seen as active participants along the frontiers, as “decision-makers,” and much as in the provinces, so too beyond the frontier, some elites welcomed the Roman presence as a way to increase their “status, power, and wealth” (Wells 1999, 95).

By de-centering Hingley means providing equal voice, both to the colonialists and indigenous peoples. Hingley sees de-centering as exploration of the “differing views of the colonial or imperial situation on the part of both native peoples and members of the colonizing power” (Hingley 1996, 40). There needs to be an emphasis on seeing native society “constraining or modifying Roman material culture” and concepts to fit their own needs, especially the tribal elites (Hingley 1996, 41; Jones 1997, 134). With regard to the frontier society and especially the Germans beyond the border, the use of Roman material has been seen as evidence of Romanization (Bouzek and Ondřejová 1990). What needs to be studied is how these goods might have been redefined by the Germans within the context of their own culture. Hingley

These three avenues of study taken from post-colonial study fit in with the approach to the study of frontier discussed in previous sections. Much in the manner of Hingley’s comment that in studying the varied native reaction to the Roman presence, we need to keep in mind the differences in power, wealth, age, gender, identity, 88

and geography, so too in looking at the frontier, we need to keep in mind all the different people interacting there. With the emphasis on viewing the frontier as a mixing of different peoples who could form new identities and cut through social barriers to from new social contacts, the ideas put forward by Lightfoot, Martinez, Whittaker, Wells, and Elton incorporate the concept of “decentering,” providing a voice for the multiplicity seen at the frontier, and acknowledging the potential existence of opposition. At this point, it may be helpful to look at German culture and identity and then toward the end of the chapter bring back Romanization into the discussion to see if there is any evidence of its having taken place.

4.3.1. A Brief Overview of Ancient German Culture Based on the few written sources and the structure of their culture in Late Antiquity and the Early Medieval Period, the ancient Germans of the first century BC– second century AD were in a state of flux in regards to their organization as a society. It appears that the ancient Germans began as loosely connected groups, possibly as small as bands, but probably closest to tribes in form, and did not view themselves as a unified people. If asked who they were, a German would respond by stating the name of their tribe (Wells 2001, 116; Todd 2004, 9). The term “Germanus” has unknown origins, but according to Roman tradition was the name of a tribe that had crossed the Rhine into Belgium and their name was applied to the people as a whole first by Celts and then Romans (Tac. Ger. 2.5; Wolfram 1997, 4; Wells 1999, 113; Todd 2004, 9). The tribes were not permanent in their structure or existence, with war or other stresses causing some tribes to breakup into smaller groups or to have one group break off to form a new tribe, but in other instances these same pressures could cause tribes to unify (Todd 2004, 29). It appears that by the time of the Late Republic and Early Empire, contact with the Celts and Romans had caused tribal agglomeration among the Germans, whereby tribal confederacies started to emerge (Wells 1999, 116–117). Again, these confederacies were not permanent and the various tribes within it could change (Todd 2004, 29).

4.3. Ancient German Culture and the Topic of Identity There has been much debate about whether there really was a cultural group that could be identified as “German” during the Early Iron Age. Some scholars, such as R. Hachmann, believed that there was no distinction between Celts and Germans (Todd 2004, 9). Archaeology has in fact shown that the description by Caesar of the Rhine as a boundary between Germans and Celts is not true and that the material culture on both sides is very similar, implying no cultural distinction (Whittaker 1989, 66; Wells 2001, 117–118).32 We also have to keep in mind that all of the ancient written descriptions and given names for tribes were from the Roman perspective for reasons embedded within the context of the writer’s needs and aspirations (Wells 2001, 116–117). While many of the points have validity, it has been shown that the rejection of an ancient German cultural identity was a backlash against the writings of German archaeologists caught up in the nationalist fervor that started in the early nineteenth century, further enhanced by German unification in 1871, and later fueled by the research and ideas of Kossina, reaching a culmination with the National-Socialist party under Hitler (Trigger 1989, 149, 163–167; Hides 1996, 40–41; DíazAndreu, 1996, 56; Härke 2002, 17; Haßmann 2002, 77; Veit 2002, 49–50; Wolters 2002, 114–115; Todd 2004, 9).33 After 1945, there was a rejection of all archaeological thought created in Germany during the nationalist period and so too the notion of an ancient German culture (Haßmann 2002, 122, 123; Veit 2002, 58; Wolfram 2002, 184; Todd 2004, 9). By the 1980s, the pendulum started to swing back again and new studies, such as those by L. Kilian (1983 and 1988), attempted to show that there was a separate group from the Celts that later became identified as German (Todd 2004, 10). Even though archaeology has shown that the Rhine was not the divide between Celt and German, there were in fact smaller-scale societies of less complexity east of the lower Rhine on the North European Plain and east of the Elbe that fit the model of what could be conceived as German starting around 500 BC (the period of the Jastorf culture) (Wells 2001, 117–118; Todd 2004, 11). Below, I present a view of what German culture was like by the late first and early second century AD.

In tribes until the first century BC, it appears that the tribal assembly had governed the affairs of the people. In times of danger, a war-leader (referred to as dux by Roman writers) could be elected from the council to lead the tribe, but then most likely relinquished power afterward (similar to dictator during the early Roman Republic) (Cunliffe 1988, 175; Wells 1999, 233). It is apparent that there was also the position of king, who was chosen from noble families, but his role was merely to lead the tribal assembly and perform religious duties (Tac. Ger. 7.1, 10.2; Wallace-Hadrill 1971, 3; Cunliffe 1988, 175; Wolfram 1997, 16). After the first century BC, with the creation of larger tribes and the confederacies, the tribal assembly started to become unwieldy and lose their power in the face of the rise of war-leaders who did not relinquish their power and kings who started to absorb the power of dux (there is no indication that a king could not be a war-leader as well, which may be reflected in the position of Maroboduus) (Wallace-Hadrill 1971, 3,15; Todd 2004, 30, 32).34 Based on the little evidence extant, by the time Germans entered into ancient written accounts, the tribal assembly did not initiate action, but considered plans and proposals put to it by kings or other leaders (Todd 2004, 29). The consolidation of power in the hands of war-leaders (duces) and kings is a complex issue, and I cannot go into detail here. As already mentioned with the rise of the tribal confederacies, it is within reason to believe that the growing contacts and interaction with the Roman Empire, including warfare, caused the rise of these leaders and broke the structure of the tribal system, pushing the Germans toward what has sometimes been regarded as chiefdoms and by the early medieval period, kingdoms 89

(Service 1962, 151; Cunliffe 1988, 177; Wolfram 1997, 8; Wells 1999, 116–117; Todd 2004, 33).35 The Romans preferred to do business with a single ruler when dealing with other peoples and possibly re-enforced the rise and power of German leaders (Wells 2001, 113–114; see also Ferguson and Whitehead 1992). There is some question as to whether the Romans differentiated between the two types of leaders and just referred to all of them as kings.

While the king was apparently weak in German society at first, or perceived as ceremonial, it is clear that by the late first century BC, his position started to change. We see men such as Maroboduus from noble lineage becoming king, but then also acting as war-leader, thereby creating an autocratic position. We do not know how certain families came to be identified as “noble,” but it could be based on having had successful war-leaders in their past who accumulated enough loot, followers, and cattle to establish the family’s reputation in the tribe. This possible progression is likely when we consider that after Maroboduus fell from power, there are references later in history to rulers who traced their lineage to him (Todd 2004, 32). We do know there was differentiation and ranking in German society based on the grave evidence, as already shown in the previous chapter and in other parts of the barbaricum; also in northern Germany, settlements show that some houses were larger and setoff from the others indicating a family of different rank (Todd 2004, 33–34, 66).

The war-leader’s power was based on his own abilities as a warrior, skill in leading men to victory, usually measured in obtaining loot and distribution of gifts, and also in eloquence (Tac. Ger. 14.2, 15.2). When the tribal council met, decisions were based on the speeches given by the war-leaders or kings and whether anything was agreed upon was dependent on the abilities of the speaker to persuade the council; approval was given by cheers or banging weapons together (Tac. Ger. 11.2). The position of war-leader was tenuous, totally dependent on successful action, since it is inherent in human nature not to want to follow a loser (Todd 2004, 32). It does appear that some young men could become dux based on their birth and if their fathers had done great things (Tac. Ger. 13.1). The dux held a powerful retinue of loyal warriors (termed comitatus by Roman writers) who could come from various tribes, seeking honor and adventure (Tac. Ger. 13.2–3; Owens 1960, 159; Todd 2004, 30). For the followers, the links to their tribes could be weakened, especially if the leader was very successful and popular, so that the retinue could follow the will of the leader beyond the wishes of the various tribes (Cunliffe 1988, 177; Todd 2004, 30). Some leaders obtained very large retinues, which could break off from the tribes and create their own tribal identity. The power of the retinue was seen in Chapter 2, when those of Maroboduus and Catualda were not allowed to return to the Marcomanni, but instead were settled in southwest Slovakia under Vannius and the Quadi. In contrast to the war-leader who obtained goods through warfare, the king received cattle, other agricultural produce, and items from the people, along with gifts from the leaders of other tribes, such as horses and weapons (Tac. Ger. 15.2).36

In conjunction with the rulers, there were also priests, who were in a position of authority and apparently ranked outside of tribal affiliation. Tacitus refers to such a priest as sacerdos civitatis, “priest of the civitas” (Tac. Ger. 10.2). They were the ones who meted out punishment through chains, flogging, or death (Tac. Ger. 7.1). These priests held their power through tending sacred groves where the sacred tokens and images were kept (Tac. Ger. 7.2, 9.2). These groves were consecrated to the gods, who had no anthropomorphic representations in German religion (Tac. Ger. 9.2). While no clear examples of these sacred spaces have yet been found in Slovakia, examples do exist in the Czech Republic (Libenice and Tuchoměřice-Kněžívka; Droberjar 2002, 147) and Germany (Kossack 1981, 316). Because cult sites were open groves, their identification and interpretation are difficult. Literary and archaeological evidence points to two types of cult activity: one oriented toward fertility and life-force, with the second toward war (Kossack 1981, 316–317; Droberjar 2002, 202). The tribe itself was composed of individual households, patriarchal in organization. In terms of marriage, monogamy was the most common, but those of higher rank and wealth might have more than one wife, practicing resource polygamy (Tac. Ger. 18.1; Todd 2004, 30–31). For the marriage to take place, the groom had to present goods to the bride’s family, termed bridewealth, which gave rights over the woman to the husband (Tac. Ger. 18.1–4). Tacitus states that the young men experienced love late and that girls were not hurried into marriage (Tac. Ger. 20.2), possibly a reflection of the need to provide the bridewealth, something that may have taken time for the man to accumulate. Even though it appears the Germans had patrilineal descent groups, with inheritance going from father to son, the brother of the wife, the “maternal uncle,” had a strong role in the lives of his sister’s children (Tac. Ger. 20.3). It appears that the bonds to both a father’s and mother’s families were strong and kinship was used in enforcing rules as well as strengthening family identity (Todd 2004, 31). In relation

Warfare was an important part of German society.37 The maturity of a young man was decided by the tribal council, where the youth would be presented armed by the family (Tac. Ger. 13.1). The council would then decide to accept the youth as a warrior and, if accepted, he then usually attached himself to the retinue of a warleader (Tac. Ger. 13.1). Fighting as part of the retinue was seen as a great honor, and for the war-leader, the larger the retinue the more power he had. Valor through war was seen as the height of male pursuit. For the warleader, warfare was the only way to obtain the necessary amount of goods to pay his retinue, along with the requisite fame and glory. Based on examples from around the world, warfare was common among chiefdoms and tribes, where true conquest was not the objective, but booty, such as cattle and horses, or to drive the enemy out of a favored zone, or to prevent them from expanding (Service 1962, 115; Earle 1987, 293).

90

to kinship, feuds were very important in German society, whereby it served as a regulating device for righting perceived wrongs within a society that valued valor; at the same time feuds could be detrimental to the tribe, eventually leading to disunity and increasing conflict (Todd 2004, 31). These feuds were through the paternal side of the family, so that the son was required to take up his father’s feuds (Tac. Ger. 21.1). It is thought that because settlements within a tribe or chiefdom were selfsufficient residential groups, who saw it a private right to protect themselves, disputes then generated feuds, which could be settled between families, but at other times the tribal council and later chief may have had to step in to settle the dispute (Service 1962, 114). Feuds could be settled by payment in cattle or sheep (Tac. Ger. 21.1).

this may have been the case when Caesar wrote, Tacitus only briefly mentions clans and it is possible that by the first century AD, the power of the clan had waned. The evidence for clans outside of Caesar’s comments has sometimes been looked for in the long-houses found in northern Germany; which because of their size and ability to hold multiple families have been taken to indicate clans.38 The problem with this view is the wide variety of settlement types found in the barbaricum, some not having long-houses. As we have seen in southwest Slovakia, the long-house does not appear to have been very common, and instead the single-family Grubenhaus was the usual structure. This type of settlement may indicate the absence of clans. Another explanation could be in the type of Germans who settled southwest Slovakia. The Quadi were part of the larger Suebic tribe, so that as an off-shoot its members may have been from other tribes, thereby weakening previous clan ties. Also, when Vannius was made king, he ruled over a large body of men who had been the retinue to Maroboduus and Catualda, thereby indicating that these men could have come from different tribes and families, thus weakening or eliminating the role of clan, and emphasizing the independent single-family unit. It is then possible that the absence of long-houses or at least their rare discovery in southwest Slovakia is in part caused by the disappearance of the clan as a powerful social force for the Germans along the Middle Danube region. There is another use for the long-house or at least the leader’s home, which is given in more detail below.

The prosperity of families appears to have been measured in the number of cattle they owned, according to Tacitus, but probably other livestock were included as well (Tac. Ger. 5.1; Todd 2004, 76). Land was owned by the tribe as a whole and distributed to the members based on their rank (Tac. Ger. 26.2), as would be expected in tribal societies or chiefdoms, in which the concept of privately owned land is unknown, based on examples from other parts of the world (Gluckman 1965, 36; Earle 1987, 294). The women, elderly, and those too weak to be warriors did the work in the fields and houses (Tac. Ger. 15.1). Tacitus makes some comments about the men not participating in work around the homes or fields, believing it a sign of weakness to work for what could be gained by war (Tac. Ger. 14.3, 15.1). He may have been confused by the presence of the retinue among the Germans. As already mentioned, the followers of the dux could come from outside the tribe, so we have to imagine a large group of landless men, dependent upon the spoils obtained through warfare (Cunliffe 1988, 176). Men who belonged to the same tribe as the dux probably did not relinquish their ties to the land and participated in working on the farms. Farming, especially in the form of tending cattle, was the lifeblood of the Germans, so to think that men never did any work on the farms is hard to believe and we have to remember that warfare did not take place all year round. This is not to say that women did not work on the land. In fact, with men gone to war during parts of the year, we have to imagine that women had a lot of power in the home and were in charge when the husband was absent. The power of women even outside the home is implicated by Tacitus, who mentions women as being present on the battlefield in support of their husbands, tending wounds, but even sometimes acting as a cheering squad on the sideline, exhorting the men to keep fighting (Tac. Ger. 8.1). Women could provide advice freely as well (Tac. Ger. 8.2).

As already mentioned above, feasting and gift-giving were an important part of German society. Tacitus provides some negative comments about Germans’ always drinking, becoming easily drunk, and participating in banquets from dawn to dawn, during which various things could be discussed, such as reconciling enemies, marriage, or even choosing a leader (Tac. Ger. 22.1–2). The primary drink of choice was either beer or, if nearer the frontier, wine; mead may have also been important even in this period, since it figures prominently in later German and Viking epics (Tac. Ger. 23; Owen 1960, 135 and 137; Todd 2004, 79). The banquets were hosted by the leaders in their homes, and, if we look at the later epics, these were the long-houses, sometimes referred to as mead-halls, which were specifically built to accommodate the retinue of the dux or rex (Owen 1960, 139). While Tacitus reports that during the banquets, various important matters would be discussed, this time was also important for bonding and apparently boasting (Owen 1960, 137). The epics such as the Edda, Beowulf, or the Nibelungenlied are filled with tales of the hero or other warriors making great boasts about their prowess, which sometimes led to tragic outcomes (Owen 1960, 137). It is entirely possible that some raids into the Roman provinces started as drunken boasts of warrior prowess during these banquets in the war-leader’s mead-hall. The banquet was important for the leader, because it was through the providing of food and drink that he was upholding part of his end of the arrangement, providing his men with sustenance. It was also during these banquets, that the dux or rex handed out

There has been in the past a discussion about the role of clans in German society, which is mentioned by Tacitus, but Todd believes there is little evidence to support the view that they were as important as some scholars think (Tac. Ger. 7.2; Todd 2004, 31). The previous belief in the power of clans was linked to statements made by Julius Caesar in his account of the Germans when relating his campaigns in Gaul (Cunliffe 1988, 175). Even though 91

gifts of weapons, horses, and other items to his retinue, re-enforcing the men’s loyalties (Bouzek and Ondřejová 1990, 23–24; Todd 2004, 34).

The role of Roman imports not only for elites but for other classes is taken up in more detail below.

The acquisition of the gifts given out by the dux was either done through various raids on other tribes or incursions into the Roman provinces. To the Germans, the provinces were a rich source of material and for the dux an easy place from which to obtain the loot to keep his retinue happy (Thompson 1982, 3 and 5). The many minor raids conducted at various times throughout the imperial period, and even some of the larger wars fought, may have started as simple raids by a dux to keep his retinue loyal. For example, as Wallace-Hadrill points out, Ariovistus was not at war to acquire land for his tribe per se, but instead for all those who followed him and had left their homes for good, that is, the retinue (WallaceHadrill 1971, 5). But also we have to keep in mind that the Romans or various provincials may have given gifts to the duces in order to cultivate closer ties with them, foster allegiances, and create goodwill to keep the Germans from doing their raids (Todd 2004, 34).

4.3.2. Cultural Identity and the Role of Artifacts With a very basic picture of German society having been given above, we need to take a look at cultural identity and the role of material culture. The creation of archaeological cultures based on discrete groups of artifacts has been a part of archaeology practically since its inception (Jones 1997, 15). Because the study of prehistoric cultures is almost totally dependent on material remains, the connection between culture and artifact is very strong in the mindset of archaeology and has been for many years. There are many instances when distinct tools, art, or weapon types have been assumed to reflect unique cultures (Trigger 1989, 162; Ucko 1994, x, xv; Veit 1996, 45; Jones 1997, 15). Linked to this approach was the tendency to assume that when one saw within these cultures the appearance of material from other cultures, the new material was evidence of either conquest or assimilation, as seen in the overuse of the term “Romanization” to describe the appearance of Roman material among non-Roman groups. This view was partially influenced by the idea that there was a monolithic cultural identity that could be traced back in time from the present and that culture was static, so that change was slow and more easily explained by outside influence (Sklenář 1983, 91; Trigger 1989, 162; Ucko 1994, xv; Jones 1997, 16, 33; Wells 2001, 20). It is now clear that culture is not static but constantly changing and that when ideas of assimilation are expressed it should be in terms of the culture’s absorbing the foreign material into its own system, not the other way around. In recent years, there have been studies to show that the same artifacts were used by different cultures or subgroups within larger groups, and so the interpretations of the past in regards to archaeological cultures may have in fact covered over or hindered the identification of other cultures or sub-groups (Ucko 1994, xi).

The presence of elites within settlements has proven difficult to detect in Slovakia for the first and second centuries AD. The archaeological evidence shows a general uniformity in material remains within huts both in southwest Slovakia and southeast Moravia (see Chapters 8.4 and 9.3). The huts are relatively the same size in settlements and the material remains are usually similar in types of pottery and imports. We know that there were elites (leading families) based on the written sources and the various rich burials found in Slovakia and Moravia, as discussed in the pervious chapters. There is as well a similar situation in Germany itself for the same period (Kossack 1981, 309). We also know that there were isolated farmsteads, and it is within reason to believe that the absence of elite residences in the larger settlements can be seen as evidence for the elites living separately on those isolated farmsteads. The situation at Stupava seems to point toward this kind of situation. The Roman-style structure can be seen as an elite residence, which sat at a distance from other German huts in the area. If one could do a wider survey and exploration at various German settlements already excavated in Slovakia, for example at the site of Branč, I believe one would find associated elite residences on the outskirts. The later third- and fourthcentury developments in Germany (Kossack 1981, 310– 311) and even Slovakia (at sites such as Cifer-Pac and Milanovce) of fortified mansion-like farms, shows the culmination of this pattern. One way in which to view the lack of elite evidence in German settlements for the first and second centuries AD is to keep in mind how wealth could be displayed. As already mentioned, cattle was one way to measure wealth, so that while a hut was not a measure of status, having a large herd would have been important. The rich burials found at places such as Zohor and Vysoke pri Morave show that burial was another way status could be displayed. For the average person living in those settlements, while their material culture may not have differentiated all that much, it is difficult to believe that there was not competition among these households.

Connected to the study of culture and material remains is the concept of identity. Identity is an inherent part of culture, in that a group of people with a shared culture will have an image of themselves expressed through various ways. In archaeological terms, we need to identify the markers of how a group articulated their identity through material culture. Wells makes a good point, however, that an ancient group “probably did not think of identity as a category of feeling and behavior,” and that our conception of identity was “embedded in the way [the ancient group] viewed their world, interacted with others and went about their everyday lives” (Wells 2001, 21). Anthropologists looking at how ethnic groups hold on to and reinforce their identity have done much to help us understand how this process works (see Jones 1997, 51, 53, and 56–65). In studying ethnic identity, scholars realized that membership in a group was “dependent on the agreement between the individual claming to be a member and 92

acceptance by the group” (Ucko 1994, xviii). One method by which this agreement could work was based on the defined boundaries of the group—meaning how they identified themselves in opposition to other groups, thereby helping to ascertain if the individual belonged (Jones 1997, 52–53; Barth 1998). It is in the nature of a group to want to feel distinctive, so that as other groups or cultures may have changed, so too did other groups in order to maintain their distinctiveness (Wells 2001, 8).39

congruence of culture and values,” but in fact the differences persist and the two groups stay separate, thereby indicating that there are structures to interaction, which allow differences to persist (Barth 1998, 16). While Barth and his followers may argue for identity to be self-defining, I am of the opinion that the “other” also plays an important part in creating identity for a group. A group can be unconsciously self-aware of how they view themselves, but they are also aware of how others view them, and so they may be influenced by how that other group views them within their own idea of identity. For example, did the Roman view of the Germans as “barbari” affect the way in which Germans created their own identity? Identity is a fluid thing, constantly needing to be re-enforced or reformed in the face of changing circumstances. At the same time boundaries may have existed, there is no denying that some people changed identities, and Barth sees this change in terms of advancing personal economic and political interests or to minimize their losses, a view which has been identified as the “social constructionist” position (Jones 1997, 74; Wells 2001, 23).41 This switch can be seen in terms of situations where a person may either suppress or emphasize their identity depending on the situation if they can either gain a social advantage or minimize a disadvantage (Jones 1997, 74, 76; see also Cohen 1978). An example may be the instances of Germans’ joining the Roman military structure, where they possibly suppressed their “German” identity to fit in, but once they left the military and returned home, they may have then emphasized their “German” identity again to regain trust amongst their peers and suppressed those aspects of “Roman” they may have picked up while in the service.

Ethnicity has acquired over the years two definitions: one is an alternative to the term “culture,” meaning tribes or peoples, and the other assumes a genetic basis for the ethnic group. This development can be seen, according to Jones, as a tension between specificity and generality, whether one should develop specialized or generic definitions (Ucko 1994, xviii; Jones 1997, 52, 57).40 In anthropology and archaeology, the study of ethnicity has focused on both aspects; that is, ethnic studies have been used to help understand the workings of culture, and there has also been a focus of study trying to identify specific ethnicity in both living groups and groups only seen now in the archaeological record. In his discussion of ethnicity, S. Shennan presents the views of Bentley, who stated “ethnic identities are anchored internally in experience as well as externally in the cognitive distinctions in terms of which that experience is ordered” (Bentley 1987, 36 in Shennan 1994, 16). Again, the role of outside influence, such as other groups, comes into play, because the “experience” is related to “a shift in conceptions of personal identity” as new modes of interaction are instituted in response to “changed environmental circumstances” (Shennan 1994, 16; Jones 1997, 28; Wells 2001, 44–45). Ethnic identity then should have specific cultural features used as diacritical markers, something which could be detected in the archaeological record (Shennan 1994, 16; Jones 1997, 28).

These viewpoints fit into what Jones states as the views of Keyes, who believes we need to look at “both the cultural and social aspects of ethnic identity in dialectical relation to one another, particularly when they are rendered problematic in situations of change” (Jones 1997, 81; see Keyes 1976 and 1981). This view is reinforced by Keyes when he says a tension exists “between cultural meanings that people construct to differentiate their primordial identities from those of others, and the patterns that emerge in social interactions as individuals and groups seek to pursue their interests” (a statement which fits into the position of primordialists) (Keyes 1981, 14 in Jones 1997, 81). We need to be aware of the interaction of different types of identity, not only ethnic, but also class, gender, and so forth, along with how other elements, such as beliefs and practices, had an affect in the creation of identity (Jones 1997, 85–86; Wells 2001, 24). In a recent work, Wells provides an approach developed by Jones using the idea of habitus from Bourdieu that is helpful in understanding identity, in that “identity [is] formed through the processes of interaction between the individual’s (or group’s) habitus – basic assumptions, beliefs and practices – and the larger social and political context of interaction with the cultural landscape” (Wells 2001, 24).42 It is also clear then that Barth’s conceptualization of the boundary of an ethnic group, or my focus on identity in general, is part of a

One type of marker is the boundary set up by ethnic groups, as already mentioned in Chapter 1 in the work of Fredrik Barth. These are not necessarily physical boundaries but cultural and are created to define the ethnic group (Barth 1998, 15). Barth goes on to say, “If a group maintains its identity when members interact with others, this entails criteria for determining membership and ways of signaling membership and exclusion” (Barth 1998, 15). This view emphasizes the role of the individual in categorizing themselves and others for the purpose of interaction, creating the observation that defining an ethnic group is a “self-defining system” (Jones 1997, 60; Barth 1998, 13–14). I believe Barth’s concept can be applied to archaeological material by saying that if a group maintains its identity even after adopting material of another group, then there are other criteria for determining membership. The use of another group’s material culture does not imply the adoption of that group’s cultural identity or a change of the social boundaries. Barth also states in his discussion of ethnic identity that if two groups are in contact, then one would expect that the persisting cultural differences would be reduced since interaction requires and “generates a 93

larger whole that makes up how identity works. This formulation clearly shows in terms of the Middle Danube region that the frontier with its interactions, the process of acculturation/resistance, and the formulation of identity all occur at the same time and are all interrelated.

Slovakia and if we also include sites in southeastern Moravia, thereby recreating the territory occupied by the Quadi, then the number of goods jumps even higher. Three types of Roman imports can be postulated for southwestern Slovakia. 1) Trade items—which are the day-to-day objects, such as cooking pots or basic cosmetic items such as combs and functional jewelry such as fibulae and hair-pins. These basic items are found within the parameters of the lower classes, especially in settlement context. The items were acquired through the mechanism of actual trade along the frontier, whether through Roman merchants operating out of mobile emporia among the Germans or by Germans visiting Roman markets at the frontier. The type of trade was probably the Germans supplying raw materials in exchange for basic day-to-day items. 2) Gift-exchange items—this covers high-status Roman goods found in elite contexts within German territory. These contexts are most often elite burials, whether found alone or within cemeteries. The items include bronze and silver vessels and plates, glass, weapons, gold and silver jewelry, and other things as well. 3) Looted items—German raids into the empire, though infrequent in the early empire, occurred more and more often as the centuries progressed. There are many Roman accounts of Germans carrying off loot back into the barbaricum. While it is easy to say that looted items would have included gold and silver or other obviously precious items, the taste of individuals varied and even odd items may have been taken back or items that on the surface do not appear valuable, such as cooking pots.

In archaeological terms, while we should attempt to be as holistic as possible in studying all aspects of interaction along the frontier, in acculturation/resistance, and identity, there are times when it is easier to focus on one or two portions of the greater picture. Here, I am more focused on how artifacts might have been used to help identify the boundaries of a cultural group; a focus which may help us understand the bigger picture. At the same time, if no discernible pattern can be identified, we can assume that identity was being maintained another way. The validity of thinking that material culture may show identity rests in the idea that material culture is “an active constitutive dimension of social practice in that it both structures human agency and is a product of that agency;” even further, the “social practices and social structures involved in the production, use and consumption of material culture become embodied by it, because such processes occur within meaningful cultural contexts” (Jones 1997, 117–118; see also Hodder 1986; Wells 2001, 22, 25). What is important to keep in mind is that material culture may be part of multiple social practices at the same time and therefore have multiple meanings during its use, depending on who is using it and in what context (Jones 1997, 116, 126; see also Shanks and Tilley 1987; Wells 2001, 25). We cannot assume specific material remains will always correlate with a specific identity, which has led some scholars to express that there is little validity in calling a certain type of fibula a “Celtic Fibula” (Wells 2001, 25). The next section presents how cultural identity manifested itself among the ancient Germans living in southwest Slovakia during the Early Roman Period.

The problem facing an archaeologist is to identify Roman imports found at a particular site within the parameters of the three categories. One of the problems is the possible overlap among the categories. While basic day-to-day items may have been acquired primarily through trade between Roman merchants and German lower classes, there may have been gift-exchanges among these Germans for the items. However, one can argue that the items were first acquired through trade and thereby whatever happens afterward is of secondary importance. There is also the question of individual tastes while looting, whereby a regular German may take day-to-day items from the empire back to the barbaricum. However, I posit that if a day-to-day item, such as a cooking pot, can be easily acquired at a frontier market, or from a traveling Roman merchant, or exchanged for with a neighboring hut, it is hard to imagine a German looting in the heart of the empire that same object and lugging it all the way back to the barbaricum. It would seem that the warrior would probably take special items not easily acquired at a frontier market or from a traveling merchant. It can be said that if a high- status item or one that is not common is found in a German lower-class context, then that item was most likely acquired through looting (either from Romans or other Germans).

4.3.3. German Identity and the Middle Danube Frontier Similarly to Wells’ conclusion that one needs to reject the idea of Romanization in regards to explaining what was taking place amongst the Germans along the Upper Rhine, so too should the concept be rejected regarding the Germans of the Middle Danube (Wells 2001, 124–128; Whittaker 1994, 125). I do not believe that the term Romanization can be used when talking about a group beyond the frontier of the empire, or more accurately I should say in regards to groups falling outside the arche.43 For the understanding of German identity in southwest Slovakia, we need to develop a way in which to view the Roman material appearing beyond the frontier. The concentration of Roman imports in southwestern Slovakia is one of the densest along the frontiers of the empire (Krekovič 1997a, 233). Few other regions in the barbaricum can compare to the sheer volume of material recovered from the hundreds of sites in southwestern

Another problem is trying to separate gift-exchange items found in elite context with looted high-status items that would also occur in elite context. For example, when an 94

elite burial is found and there are 15 bronze vessels, five silver ones, a silver plate, various Roman gold and silver pieces of jewelry, and a range of terra sigillata, how much of this was acquired loot and how much gift-exchange? Germans did gift-exchange amongst each other and raided each other’s territories as well, so that Roman items obtained one way may end up at a later date in another context altogether. Going back to the burial example, of those items, can it ever be said that the highstatus Roman goods were obtained from actual Romans or were they merely acquired through various social mechanisms amongst the Germans themselves? Then too, there is the question of whether or not an elite German traded for a valuable Roman object, as is possible, but does not seem to fit into the social mechanisms of German society in the barbaricum, i.e., in terms of the roles played by gift-exchange and looting.

acquisition of these goods was used to show their prosperity, much as in having a large herd of cattle. We have to imagine that this took place not only among elites, but also among the lower classes. Competition among households was probably a very important component rolled into the prestige and honor seen in the warrior lifestyle. We need to keep in mind the idea of merely liking an object for perceived inherent beauty and therefore creating a desire for the item. We see this in relation to the Celtic graphite vessels that remained in circulation even after having been broken and repaired (and which were in circulation even among Germans). Also, there are examples of German fine-ware vessels imitating Roman glassware and metalware designs (see Chapter 9 for more detail). These copies reflect German enjoyment of certain forms or designs and should not be seen as a reflection of slavish copying because they wanted to be like Romans. If a potter sees that people are bartering for certain items and willing to give a lot of things for it, then the potter may start to imitate the desired item to increase the potter’s own trading power. For example, if a potter can get a dozen eggs for one pot and two dozen eggs for another pot, then the potter will more likely focus his energies on creating pots worth two dozen eggs.

As a modus operandi, I believe that when, during the excavation of German huts, one finds Roman imports of an average sort, such as cooking vessels or amphorae, it is reasonable to infer that these items were acquired through trade (even if that item was later given as a gift between huts does not change its original mode of acquisition). If one finds a unique and valuable Roman import in a non-elite context, then that item was likely to have been acquired through gift-exchange or looting. And finally, if one finds high-status items in an elite context, then both or either gift-exchange and looting could explain the context and there is no way of determining which took place. Again, as in realizing an item obtained through trade could be later given as a gift, so too an object obtained through a raid could later be given as a gift (or vice versa). The elite burials mentioned in Chapter 3.2 and 3.5, such as at Zohor, should not be seen as a static collection of Roman imports, meaning that all were obtained at the same time, such as through one massive gift from some Roman provincial. Instead, we need to realize that these items might have been collected during the lifetime of the individual through different means, with a mixture of all three categories of items. We need to remember the important role of banqueting during which time gifts could be given, so that the Roman imports we see might have all come through German channels with Romans being involved only at the time of the item’s creation and first trade/gift/looting.

We cannot minimize, however, the impact of elite behavior in terms of acquiring Roman goods. The ownership, use, and redistribution of high-valued items was used to reinforce the control exercised by the elites (Kossack 1981, 311; Whittaker 1994, 125). We see this through the dux who maintained his power through the distribution of loot and gifts, along with hosting banquets. The way in which one can explain this behavior is through the projection of power. The dux, and the elite in general, maintained their power through the creation of an image of success. Because the Germans were in a complex relationship with the Romans, who were seen as a powerful force by the Germans, then one would expect German elites to try to copy certain things from the Romans to re-enforce their own image as being successful and powerful. Therefore, the elite burials with Roman goods or the Roman-style structures, such as at Stupava, Cífer-Pác, or Milanovce, can be seen within the context that the elite were imitating what they perceived to be ways to show power and success. I am not saying that they were trying to be Roman, but merely using what they thought necessary to project power in a way that would be understood by other Germans.

The Roman goods we see in southwest Slovakia need to be seen within this context: they increased the prestige of their owners (Whittaker 1994, 125–126). By having even mundane cooking pots from Pannonia, the German who owned them may have been seen as being part of a successful retinue or that through the largess of the dux that German was able to go and obtain those items. We cannot talk, therefore, about any type of Romanization, since the various pots, terra sigillata, and bronze vessels need to be considered within a purely German context; much as Hingley argues the same from Roman material in British contexts (Hingley 1997, 87). It is not that the Germans thought Roman items were inherently superior or that they wanted to live like Romans; instead the

We have to realize as well that many Germans were willing to serve in the Roman armies because they would accumulate material wealth which they could bring back to their homeland and thereby increase their social standing. A poor man from the lower rung of a tribe might come back wealthy enough to become a dux. In some ways, we could consider the Germans going to join the Roman armies as believing that they were going to serve in the retinue or comitatus of the emperor through which they would receive largess to improve their own social standing at home in the barbaricum. The Romans 95

then were not only perceived as a threat but also an opportunity. As we know from the written sources, Germans defeated one moment were a few pages later serving as auxiliary forces in the Roman Armies. The Germans were not somehow showing an allegiance to the Roman Empire, but instead they served in the Roman armies through the mechanisms they understood, i.e., the comitatus. If their war leader was defeated, then the retinue would dissipate and some may then have joined the retinue of the one who won the victory, in some cases the Roman armies.

perceive their own frontier as a wall, because they operated in non-Roman territory adjacent to the provinces whenever they chose, while allowing non-Romans to have access to the towns along their frontiers. Whittaker argues that we should conceptualize two types of imperial control – the administered provinces, which fall into the territory termed the arche and the nonadministered provinces left in the hands of local rulers, such as Vannius over the Quadi, which fall in the territory termed hegemonia (Whittaker 2004, 40). Even though the Romans did not exert a continued direct control over the territory bordering the provinces, the Romans still felt it within their right to interfere in the workings of these lands whenever they felt it necessary.

The ways in which German identity were preserved then was the continued power of the retinue as an idea within the mind of Germans, even when they went to serve in the Roman armies. There were other ways too that German identity was expressed or preserved. We see in the period from first contact around 10 BC to Late Antiquity in the fifth century within southwest Slovakia, that German habitation did not change. Even though Germans were fully aware of how provincials lived in the empire and some elites may have created villae rusticae, most continued to live in Grubenhäuser. We can then argue that the German hut was a marker of identity, which was maintained to show difference (Wells 2001, 25–26).44 Another way was the continued use and circulation of traditional German pottery, whether handmade in the Early Roman Period or wheel-made in the Late Roman Period, they had a known repertoire of designs and decorations. German pottery then could also be a hallmark of identity, in that while households wanted to have some Roman pots for prestige, they still identified themselves as German through the continued use of traditional pottery. Another way in which German identity was maintained was through the continuation of the banquet/feasting, where they would gather and interact, a behavior that reinforced their identity as different from that of Romans. Feasting would have been important for the stories that would be told during it, especially when we remember that this was a society without writing. Story-telling was an important part of pre-literate societies, as we see through the transmission and survival of some epics (such as the Nibelungenlied and Beowulf). Religion is another facet of society in which identity was preserved. We have yet to find actual structures in southwest Slovakia that could be identified as temples. It is therefore highly likely that Germans maintained the worship of deities through sacred groves and non-anthropomorphic representations of their deities. This continuation of traditional practice could be seen as a way of differentiating themselves from the Roman provincials, whose temples and sacred places they would have seen at the various vici and other sites in Pannonia.45

By viewing the frontier as a zone of interaction, giving voice to the non-Romans within this zone, we have come to see that the term Romanization cannot be applied to the ancient Germans living along the frontier, such as in southwest Slovakia. There was a much greater complexity to the interaction taking place then merely local indigenous groups adopting Roman goods and becoming more and more Roman. It is now understood that the use of Roman artifacts by Germans beyond the frontier, and even local groups in the provinces, was done within the mechanisms of their own culture. It is also apparent that even the use of the term Roman is highly problematic in a frontier context. This is true both in calling everybody in the provinces Roman and all material from there Roman. A majority of the population in the provinces were the same groups that had lived there before conquest. These groups maintained some aspects of their identity and some of their material culture was actually preserved, such as Celtic designs appearing in what is viewed as the Roman provincial pottery of Raetia, Noricum, or Pannonia (see Chapter 9.1.1–13). Also, people who were actually Roman or of Italian origin in the provinces were a very small group in comparison to the number of indigenous people still living at the frontier; what is more, the legions and auxilia would have brought people from other parts of the empire. The frontier, therefore, was a place where there was a mixture of people from all parts of the empire. Connected to this view of the frontier and rejection of Romanization is the concept of identity. Much in the way the frontier may be seen as a multi-leveled interaction, so too identity is now seen as multi-leveled, created through the interplay both of elements within a group and from outside the group. For example, at the frontier, Germans may have been perceived as barbaric by the provincials creating an identity applied to the Germans, while the Germans themselves may have maintained traditional dress and hair-style (such as the Suebian knot) to reenforce their difference and “Germanic” identity while visiting the Roman frontier. Both the “barbaric identity” and traditional “German identity” were in play at the same time and in fact the choices made by the Germans possibly re-enforced the barbaric identity, as would have specific reactions by the provincials. At the frontier, the boundaries between groups, such as elites and lower classes or German and Roman could become blurred as

4.4. Conclusion The frontier should not be seen as a static boundary, but instead as a place of multifaceted relationships between a wide variety of people. The frontier was a zone of interaction extending both into the Roman Empire and beyond into the barbaricum. The Romans did not 96

well, thereby allowing the creation of something different than what one would find deeper in the empire or German lands. Identity was not static and a person could take on multiple identities whenever necessary. A German who had served in the Roman armies may take on a Roman persona when visiting frontier markets so as to obtain better deals, but at home among other Germans, this Roman persona may be suppressed and a German-warrior persona projected instead. It is therefore quite likely that a German visiting the market one day may use Latin and dress as a provincial, but the next day come back as a warrior, belonging to a dux’s retinue, and participate in pillaging. The barbarian persona which has become fixated in the minds of people today, that rough simpleton too stupid to understand or appreciate civilization and too simple to have multiple identities, is not true. The Germans were sophisticated in their own way, different from the Romans, but in no way inferior. A comment made by Dio Cassius shows that Germans were fully aware of identity, when in a peace treaty the Quadi were forbidden from attending Roman markets out of fear Marcomanni and Iazyges agents would pretend to be Quadi and thereby reconnoiter Roman military positions (Dio Cass. 72.11.2–3).

in serving wine, but again, other Germans he interacted with would not, and so that oinochoe again serves a new function or is given a new meaning depending on how its German owner used the vessel. When Roman objects, such as terra sigillata or cooking wares, were purchased by Germans at frontier markets or from traveling Roman merchants, again these objects did not have to maintain a “Roman” identity. Just because a terra sigillata cup or bowl was seen as fine ware in the empire and used for dinners did not necessarily mean it was used in that fashion by Germans. The cup or bowl may have been purchased as a symbol of the household’s prosperity, or was used in whatever way the owner saw fit. It is not as if everything came with instructions on how to properly use Roman-made goods, so as to maintain a proper “Roman” identity, and therefore the owner could become Roman. The belief that the use of a foreign-made item automatically changes a person’s identity is a silly one and if true, then all Americans would be Chinese since so much of their material goods are made in China. In Part II of the book I present the findings from the archaeological project and go into greater detail regarding pottery and other artifacts. Part III will bring the finds from the project into the discussion of frontiers and identity as presented in this chapter.

The German cultures bordering the Roman Empire maintained their identity as a separate people through the reinforcement of various aspects of their culture that they thought were representative of being German, such as traditional housing and pottery, but also in continuing traditional behavior, such as feasting and the worship of their gods in sacred groves. At the same time, identity could be enhanced or changed through the manipulation of outside material (i.e., Roman goods) within the mechanisms of the group; also they could reinforce their difference through these outside materials depending on the situation. The Germans saw the Roman Empire as a place of opportunity to improve their social status at home in southwest Slovakia. The behavior connected to the attempt to improve social status could manifest as participating in raids into the empire to acquire loot and to create a “warrior” identity, which, as we have seen, was important to the Germans. Looted Roman objects, therefore, took on an identity of showing German warrior prowess, or if that loot was distributed by a dux, then that object was a symbol of recognition by his leader. The Roman goods therefore lost their “Roman” meaning, and so, for example, whether or not a German knew how to use a bronze oinochoe in serving wine was unimportant, but displaying it as loot so that he could point to it when among friends to tell the story of how he obtained the object becomes the object’s new “German” identity. At the same time, a German may have improved his social status by serving in the Roman armies as a way to acquire material wealth so that when he returned he could either purchase greater status through cattle and gift-giving or show-off their wealth as a way to demonstrate his own success and therefore elevate his position in the group. The Roman items this German brought back with him then become a symbol of his success in foreign lands and as potential gifts to others to create a retinue or alliances. He may in fact understand how to use a bronze oinochoe

Endnotes 1. E. Wheeler makes an interesting argument that the response to Luttwak has been more against Luttwak than against the existence of strategy itself (1993a, 12). Wheeler has brought up very good points about the approaches advocated by Whittaker and Isaac, but his line of reasoning is weakened by a blatant bias against and dislike of Marxist archaeology or what he calls “leftist traditions” (Wheeler 1993a, 13). Wheeler rejects the view that the Romans had no strategy and believes the arguments by Whittaker and Isaac go too far in creating the impression that the Romans had no concept of boundaries and the military at the frontiers was there to deal with locals not the barbarians outside (Wheeler 1993a, 10–13, 24–25). He acknowledges that the ideas of Luttwak went too far, but that this does not mean we need to reject the idea that Romans did not know what strategy was or apply it to the maintenance of the empire (Wheeler 1993a, 3). He also rejects the need to differentiate between frontier and border (Wheeler 1993a, 25–26). I believe Wheeler makes a mistake in denying the differences between the two, which I discuss in detail in the following sections of this chapter. Wheeler agrees that there may have been no “grand strategy” as defined by Luttwak, but that there was still strategic thinking on a large scale and strategic planning in regards to war (Wheeler 1993b, 216–217, 234, 238– 240).

97

2. In the need to govern the vast territories of the imperial powers, colonial administrations (especially the British in India) used the ancient authors and the Roman Empire to find examples of how frontiers were to be maintained, emphasizing the use of natural boundaries, as seen in Curzon’s work in the very early twentieth century, as well as in Lord Cromer and James Bryce’s essays (Cromer 1910, 26–27, 31; Bryce 1914; Whittaker 1994, 3; Hingley 2000, 36–37). There was also a search for “scientific frontiers,” in India, which used natural boundaries along with strategic positions and artificial barriers to better define the territory controlled by the British; the concept was first put forward by Lord Roberts, and supported by Lord Curzon and James Bryce (Bryce 1914, 18; Whittaker 1994, 60). It was assumed then that the Romans also searched for scientific frontiers (Bryce 1914, 18; for more on Bryce, see footnote 29 below).

institutions and even character of the United States were shaped by the frontier (Turner, 1920; Thompson and Lamar 1981, 4; Wittaker 1994, 5). The ideas of Turner were added to by others such as W. P. Webb (1931) to create an image of the American frontier as “not a line to stop at, but an area inviting entrance,” thereby making the frontier a process not a boundary (Wolfskill and Palmer 1981, x; Whittaker 1994, 5). This concept creates the image of the frontier as a “line continually pushed forward (or back) by heroic frontiersmen, the pioneers. Inside the line is culture, beyond it is nature” (Russell 2001, 2). The weakness of this approach, when first conceptualized by Turner and others, was its total disregard for the non-Europeans on the frontier, such as the American Indians, and even the role of Mexicans and blacks (Thompson and Lamar 1981, 4). 5. The core-periphery approach has also been applied to the Greco-Roman world as well, with B. Cunliffe’s Greeks, Romans, and Barbarians (1988) a good example.

Along with the emphasis on actual boundaries was the connected concept that the frontier was there to keep the barbarian or unwanted out, so that within the empire the civilizing effect could take hold on the local population (Hingley 2000, 37). Examples for this type of policy were sought in the Roman Empire to justify the actions of British imperial policy within her colonies (Hingley 2000, 37).

6. This view is sharp contrast to the way which Mócsy argued one should view the Middle Danube frontier in a 1979 article, in which he states the Danube frontier was one of defense because Augustus had no goals beyond the river, unlike the territory beyond the Rhine (Mócsy 1979, 628). He uses this argument in conjunction with the lack of military evidence in Pannonia dating to before Claudius to show that the Danube River was not seen as the “military” border of the territory during the early Julio-Claudian period (Mócsy 1979, 629).

An excellent example of imperialist thought, both modern and harking back to the Romans, was provided by C. Wells in his The German Policy of Augustus when he quoted Winston Churchill saying, “What enterprise … is more noble and more profitable than the reclamation from barbarism of fertile regions and large populations? To give peace to warring tribes, to administer justice where all was violence, … to draw richness from the soil, to plant the earliest seeds of commerce and learning, to increase in whole peoples their capacities for pleasure and diminish their chances of pain – what more beautiful ideal or more valuable reward can inspire human effort?” (Churchill 1933, 9; C. Wells 1972, viii).

7. Braund makes a similar argument (Braund 1984, 188). 8. This territory beyond the established border has been termed also the Vorlimes (Whittaker 1994, 122). 9. It is interesting to compare these zones to the economic ones created by B. Cunliffe, who identifies a market zone along the frontier, ranging 100–200 km beyond, comparable to Whittaker’s unadministered territory under Roman rule, and then Rich Burial and Warrior Burial Zones, which are comparable to Whittaker’s periphery (Cunliffe 1988, 186 and fig. 71). These divisions are also paralleled by Well’s definition of Roman goods circulating beyond the frontier within a model he had created in an earlier publication, stating that “easy access to goods from another culture…results in their integration into the local everyday milieu and mitigates their use as objects signifying special status. When imports are not easily accessible, as in the hinterland far removed from the frontier, then such objects can be adopted by elites for purposes of status display” (Wells 1992, 185). This topic is taken up in more detail in the sections dealing with Romanization and German Cultural Identity.

3. See James 2002 for a detailed review of how nineteenth and twentieth century military philosophy also affected the development of Roman military archaeology and especially of its impact on Eric Birley and his Durham School, which had great influence on RomanoBritish military studies (James 2002, 9, 17–26). James also argues for the rejection of the phrase “The Roman Army,” which he views as a modern concept, while the Romans when they made generalities about the military, always referred to the “armies” (James 2002, 38). He calls for scholars to speak of “Roman armies” (James 2002, 39). 4. Frederick Jackson Turner is considered to be the father of frontier studies in the United States, starting with his 1893 paper “The Significance of the Frontier in American History” and followed by other writings that argued the

The concept of three distinct zones along the frontiers is also seen in some of the writings of Czech and Slovak archaeologists. An article by Bouzek and Ondřejová

98

(1990) discusses the frontier region of the Middle Danube as the “third zone,” whereby the first zone is the Roman Empire, the second zone is the lands far beyond the frontier where various German and non-German groups live, and the third zone is made up of those peoples living along the Danube between the first and second zones (Bouzek and Ondřejová 1990, 29, 30–31).

fails to realize that because the Romans always viewed the territory inhabited by the Quadi as part of their “imperium,” these Germans were in the eyes of the Romans under their control. Reality may have been different, as we see with the frequent Quadi incursions into Pannonia and rejection of Roman appointed kings, but this should not detract from the fact that the Romans believed they had the right to exercise control over these Germans whenever they wanted (see also Braund 1984, 93–95).

10. Even though he does explicitly state the existence of them as separate entities later, he then states that having distinctions among border, frontier, and limes is semantic (Wheeler 1993a, 24).

15. The text from the Aeneid reads, “His ego nec metas rerum nec tempora pono: imperium sine fine dedi.” (Verg. Aen. 1.278–279)

11. The idea that the legions were positioned for dealing with internal and external problems is not new, and S. Mattern makes a good point that the position of legions was because of their role as occupiers, not as a defensive force (Mattern 1999, 101). Mattern goes on to create the concept that the position of Roman troops was for the terror they could inflict, thereby ensuring peace; the aim was to punish, avenge, and to terrify, which she defines as, “to reassert a certain state of mind in the enemy, or even a certain moral equilibrium or status relationship between Rome and the barbarian (Mattern 1999, 115– 116, 117).

It should be mentioned that there are different ways of translating this line, and the term imperium need not mean empire. For example, L. Lind’s translation of the line reads, “For these I shall set no limits of time or possessions; I have given them endless power” (Lind 1962, 10). 16. The Latin text reads, “Quae cuncta sua manu perscripserat Augustus addideratque consilium coercendi intra terminus imperii, incertum metu an per invidiam” (Tac. Ann. 1.11). This statement came from the third document left for Tiberius after Augustus’ death. These were written by Augustus, with the first document detailing his funeral arrangements, the second a record of his achievements, and the third providing information on resources, strength of the armies, taxes, and this last statement given above. Suetonius referred to this third document as a “breviarium totius imperii” (Suet. Aug. 101).

12. Some of these kings were even brought to trial in the senate for misdeeds, such as Antiochus of Commagene, Archelaus of Cappadocia, and Rhescuporis of Thrace (Millar 1982, 4; Braund 1984, 166–167). 13. Braund states rightly as well that client-kings continued to exist during the imperial period and into the Byzantine; the argument that during the period of Vespasian’s principate, client-kings disappear does not hold true in the face of the evidence (Braund 1984, 187– 188).

17. Archaeological evidence shows, however, a continued Roman presence on the Euphrates, and the client kingdom of Mesene remained loyal to Rome as well, so that we should recognize that even though outright control may have been relinquished, for the Romans the territory was still part of their “imperium” (Whittaker 2004, 44).

14. L. Pitts states that “…the relationship between Rome and the Marcomanni and Quadi developing with time as Roman influence over, and interference in, the area increased, although never to the point of exercising direct control over the tribes” (Pitts 1989, 53). I take issue with this statement, especially with the idea that the Romans never exercised direct control. The Romans always believed they had the right to exercise direct control over territory that bordered their empire. This control may not have been the same as in the administered territories, but it was still direct interference in how the tribes operated. We see this belief when the Romans placed in power their choice for kings, rejecting local choices for whom the Quadi wanted king (as we see with Marcus Aurelius during the Marcomannic Wars – Dio Cass. 72.13–14), restricting access to the Roman markets in the provinces (Dio Cass. 72.11.3), creating “neutral zones,” or restricting the places and days for the open markets in German territory. All of these instances would seem to indicate a direct involvement by the Romans in the day to day existence of these tribes. As Pitts herself agrees, there was an ever increasing influence and interference, but she

18. These numbers are highly speculative and only provide an approximate idea. These differences, however, are within reason when we consider that the average German cow weighed 236 kg (Droberjar 2002, 228), of which the meat weight would have been 118 kg. The figure of 118 kg comes from a formula provided by D. Wulf in an article calculating how much meat one should expect from a market animal (Wulf 1999). In his calculation, a modern cow on average weighs 1,200 lbs., but the final meat collected is affected by dressing percent and carcass cutting yield, so that the formula is as follows: (dressing percent x carcass cutting yield) x live weight The average dressing percent is 62% and if the cow was lean, well muscled, then the average carcass cutting yield

99

is 80%, if bones are left in and fat is not trimmed. Based on these figures a modern cow, weighing 1,200 lbs. would yield 624 lbs. (283 kg) of meat.

archaeology and Romanization, which believed that: 1) “not all races had an equal potential to participate in civilization;” and 2) “a faith in the absolute validity of the values of European culture, seen as the heir to the civilization of the classical world” (Woolf 1998, 4–5). This attitude is nicely summed up in Julian’s view of the Gauls as being attracted to what the Romanized southern Gaul had to offer and that they were “members of the European civilization, even if new arrivals” (Woolf 1998, 4).

19. Roof-tiles produced by the legions need not have been solely for military use. There are examples from other parts of the empire where bricks and roof-tiles made by the legions appear in non-military structures. In Britain, bricks produced by the 6th Legion (the stamp was LEG VI VIC TPF) were used in a villa at Gayton Thorpe, while at Pentree Farm, Oakenholt, legionary stamped tiles were used at a bath (Collingwood and Wright 1992, 165, 177, 180, 185, 188; Vrba 1997, 98–99). In Dalmatia, stamped tiles with LEG III FF and LEG VIII AUG were found at the Claudian municipium Asseria (Wilkes 1979, 185; Vrba 1997, 99).

27. Mattern also includes Tenney Frank as an important proponent in English for Mommsen’s views (Mattern 1999, 214). 28. Haverfield also had personal contact with Mommsen later in his career, when he invited the German scholar to edit the inscriptions from Roman Britain (James 2002, 16).

20. This ala was a cavalry force of 480–500 men (Vrba 1997, 109, footnote 18). The word Cananefatum is the accepted spelling for the term, but various inscriptions at Gerulata have a variety of spellings, including Canafates, Cananefates, Caninafates, Caninefates, Caninfates, Cannanefas, Cannanifates, Cannefates, Cannenafates, Cannenefates, and Canninefates (Hošek 1985, 116).

29. Imperialism as a concept is also seen as a product of the nineteenth century, within the context of the British Empire and once again F. Haverfield is seen as an important figure (Webster 1996, 10; Freeman 1996, 18). There has been some debate about whether one can label the ancient Romans as imperialists, since this concept was supposed to be modern (Freeman 1996, 20). However, I would agree with the viewpoint that just because they may have not had the vocabulary or a codified concept, they could still behave as imperialists (Freeman 1996, 21).

21. The Latin reads, “Quo tantam partem campi iacere, in quam pecora et armenta militum aliquando transmitterentur? Servarent sane receptus gregibus inter hominum famem, modo ne vastitatem et solitudinem mallent quam amicos populos.” (Tac. Ann. 13.55). 22. As mentioned in Chapter 2, a possible watchtower was found at Devin dating to the Augustan period, but there is some disagreement about interpreting the remains.

Another important British author of the early twentieth century dealing with imperialism was James Bryce, who wrote two essays titled “The Ancient Roman Empire and the British Empire in India” and “The Diffusion of Roman and English Law throughout the World” (1914). The two essays compared the two empires, showing the similarities and differences, and which further illustrate the attitude toward imperialism. Bryce in his opening pages remarks, “Europe—that is to say the five or six races which we call the European branch of mankind— has annexed the rest of the earth, extinguishing some races, absorbing others, ruling others as subjects, and spreading over their native customs and beliefs a layer of European ideas which will sink deeper and deeper till the old native life dies out” (Bryce 1914, 2). The way in which this was said within the context of what else Bryce wrote clearly makes it obvious that this spread of European ideas and the extinction of native ones was seen as correct. This occurrence also happened in antiquity with the Romans, when Bryce states, “Yet it is not altogether a new thing…[sic] the conquests of Rome brought all these regions…under one government…Thus the ancient world, omitting the barbarous North and the semi-civilized heathen who dwelt beyond the Euphrates, became unified, the backward races having been raised, at least in the upper strata of their population, to the level of the more advanced” (Bryce 1914, 3).

23. The way in which the territory was calculated for Xanten has been criticized by Whittaker, since the theory depends on the territory being defined by wherever stamped bricks of the legion are found (Whittaker 2004, 101). 24. As an example, Okun mentions how the Helvetii adopted the use of terra sigillata through their own imitations, but they did not necessarily adopt the Roman eating habits implied by the use of the fine-ware forms (Okun 1991, 437). 25. Okun provides a good example when discussing the population of Germania Superior, which included the local people, some Roman citizens, non-local merchants, traders, craftsmen, and Italian and non-Roman, foreign soldiers (Okun 1991, 435). 26. In his opening chapter of Becoming Roman, Greg Woolf also lists Camille Julian as playing a role in the creation of Gallo-Roman studies similar to Haverfield’s in Romano-British studies (Woolf 1998, 4, 5). Woolf also mentions an important point about late nineteenthand early twentieth-century works dealing with Roman

100

30. Woolf provides another quote from Haverfield that further shows this attitude of the frontier as a wall when he says, “Our civilization seems firmly set in many lands; our task is rather to spread it further and develop its good qualities than to defend its life. If war destroys it in one continent; it has other homes. But the Roman Empire was the civilized world; the safety of Rome was the safety of all civilization. Outside roared the wild chaos of barbarism” (Woolf 1998, 4).

probably chosen from the council. The peace chief “kept a pulse on public opinion, settled domestic disputes, organized community works, rituals and ceremonies, and negotiated with others,” while the war chief’s duties were only to organize and lead during war, deal with prisoners, and kill suspected witches; but both offices were “sources of prestige rather than power, for no chief had to the authority to do anything more than act as a spokesman” (Fagan 1995, 465). This description has a striking similarity to the positions of dux and rex among the Germans, even in their need to be able to persuade the council to follow their ideas, so that I believe we should see both as having been part of the tribal political makeup, but whose roles changed during the Roman Period.

31. We need to keep in mind the critiques already mentioned before by Lightfoot, Martinez, and Mattingly of taking the idea of “opposition” too far, placing one group in an overly unequal framework, thereby weakening their voice (see also Woolf 1998, 19–23).

35. See Earle 1987 on chiefdoms. As Wells shows, this view can be seen also as “tribalization,” where a group of people who did not have a clear sense of identity, distinct territory, or even specified leader come into contact with a more complex society which then causes the group to create a more self-consciously defined identity, with a sense of territory and a political structure focused on a single leader (Wells 2001, 32). One can say that the Germans were originally a people of loosely organized bands that once in contact with Celts and Romans, started a process of reorganization and re-identification leading to the formation of tribes. 36. The position of the king fits in well with the role of leaders in chiefdoms that practiced redistribution, but within the context of the new outlook on how it worked. The old ideas about redistribution believing that staple goods were brought to the chief, who then evenly distributed them back to the populace has been rejected, primarily because it has become commonly recognized that communities were already highly generalized and self-sufficient (Earle 1987, 292). Now, redistribution is seen as a “system of finance, a means to mobilize staple goods to provide for public feasts and to feed chiefs’ attendants” (Earle 1987, 292). Another important aspect of the chief is his control over prestige goods, which is discussed in more detail in relation to feasting and giftexchange (see also Earle 1987, 296).

32. While it is true that material culture does not necessarily distinguish a specific group all the time, so too, multiple groups may in fact share material culture and yet maintain separate identities. The situation along the Rhine may in fact have been as Caesar described it, even though the material culture of the Celts and Germans was the same. Wolfram makes a good point when he states that one thing Caesar did know about his enemy, the Suebian king Ariovistus, was that Celtic was not his native language (Wolfram 1997, 60). 33. This nationalist fervor had in fact started earlier, in the late eighteenth century, when Germans rallied against Napoleon and the French conquest (Haßmann 2002, 68; Veit 2002, 50). The production of the play Die Hermanschlacht (1809) by Heinrich Wilhelm von Kleist helped to give the Germans hope: on the surface the play told the story of the German victory under the leadership of Arminius over the Romans at the Teutoburg Forest, but in fact the Romans represented the French, who were occupying Germany at the time the play was written (Wolters 2002, 113). The play was nationalistic and glorified Arminius as a great German hero, and was a later favorite of the Nazis. 34. H. Wolfram provides another explanation, offering the idea that the traditional position of a royal king was replaced by the dux (Wolfram 1997, 15). It is reasonable to believe that some ambitious men who became traditional kings during this period may have seen the inherent power within the position of dux and wanted to absorb this so as to expand their own power. Wolfram also acknowledges this idea in his statement that some members of the old royal families, when they sought the kingship, were really trying to create one in the new form (Wolfram 1997, 18). The major difference in views is my belief that the position of both rex and dux had existed before as extensions of the tribal council (Wallace-Hadrill alludes to something similar in his views, 1971, 15–16). Even though comparisons between lands far apart in time and distance may sometimes prove misleading, I believe the following may help to clarify this point. As with the tribes in the Southeast of the United States, the Iroquois had two chiefs, one for war and the other peace and were

37. The elevation of warfare to an important part of German society can be attributed to the expansion of their population across northern Germany and their coming into contact with the Celts and later Romans. We can imagine that when the Germans had easy access to land and associated resources, warfare had not been as endemic. However, once the Germans abutted the territory of the Celts and then the Roman Empire, their semi-nomadic lifestyle and concept of expansion came into conflict with the sedentary Celts and Romans. The Germans tried to expand into Gaul and other lands, but were repulsed, so that the Germans were forced to compete more amongst themselves for the land they already occupied. With rising population, the access to resources, such as open land for cattle or salt, became increasingly important, which led to more conflict between tribes. The need to fight became an important

101

aspect of tribal life, leading to the “pathos of heroism” and the identity of a successful male being linked to that of brave warrior (Wolfram 1997, 8). This belief was further reinforced by the number of German men who entered into service with the Roman armies, where they were further instilled with a martial mindset. We can see this also within the rise of the dux, the absorption of power from the dux by the rex over time, and finally the weakening of the tribal council in authority. I am not saying, however, that warfare was the only impetus for change, or that I agree fully with the theory of endemic warfare leading to state formation as proposed by Carniero (see Carniero 1970, 1981). See also Earle 1987, 297 for an explanation of the role of warfare in a chiefdom.

many diverse factors and be as all-encompassing as possible (Wells 2001, 23–24). 43. See Bouzek and Ondřejová (1990) for a traditional use of the term Romanization in reference to the presence of Roman goods amidst the ancient German settlements in the Czech and Slovak Republics. 44. A similar type of situation was identified by Hingley for Britain, where the traditional roundhouses continue well into the Roman period and are either as symbols of elite power over others or continued as a way of expressing resistance to the Roman presence, depending on the situation (Hingley 1997, 95–96). 45. There are other examples for the continuation of traditional ways. Okun, for example, provides evidence that in Germania Superior, Celtic woman continued to wear traditional clothing, while men did not adopt Roman-style dress, but instead created a new style of dress, indicating a continuation of a separate identity from the Romans (Okun 1991, 437).

38. Much as with the long-houses of the Iroquois organized by clan (Fagan 1995, 463). Also, it is worth noting that the Iroquois village had a council made up of a representative from each clan during the early part of the Middle Iroquoian period, but by AD 1400 larger tribal organizations had been created; these tribes were governed by a council which was an extension of the earlier village council, thereby preserving the influence of the clan (Fagan 1995, 466–467). In fact, clan ties cut across village and tribal boundaries, forming associations (phratries) that dealt with some religious matters, such as funerals (Fagan 1995, 467).

Hingley also uses this mode of argument for Roman Britain, where he believes the indigenous groups and lower classes resisted the act of interpretation and only used the traditional names for their gods, in contrast to elites and those of high military ranks who used both the traditional names and the Roman names in inscriptions and worship (Hingley 1997, 88).

39. A good example mentioned by Wells as put forward by Jones is the creation of an identity by the Tswana of southern Africa because of their exposure to European missionaries (Jones 1996, 69; Wells 2001, 29). Before the arrival of the Europeans, the Tswana did not have a conscious identity of themselves as a distinct people, but once they were in contact with a people clearly different, the Tswana were motivated to formulate their identity, which was seen as a form of opposition (Jones 1996, 69; Wells 2001, 29). 40. Ethnic groups can be defined as culturally ascribed identity groups, which are based on the expression of a real or assumed shared culture and common descent; as a process, ethnicity involves a consciousness of difference (Jones 1997, 84). 41. This position is also termed “instrumentalist” (Wells 2001, 23). In opposition is the “primordialist or essentialist” position, which “emphasizes ethnic identity as a feeling shared by people as the result of real or imagined common ancestors or common traditions and experiences” (Wells 2001, 23). Because of limited space and my own beliefs on this matter, I shall focus on the social constructionist approach. Wells makes a good point, however, that we need not think of these positions in opposition (Wells 2001, 23). 42. This approach also bridges the two positions mentioned earlier, the social constructionist and primordialist, so that we try to view identity through

102

PART II THE ARCHAEOLOGICAL PROJECT

In the Pliocene, the Morava River flowed close to the Little Carpathian Mountains, but then during the Quaternary Period jumped its bed and moved 14 km to the west (Vaškovský 1977, 104, 227). This behavior indicates a meandering river type, which would cause both deposition of river sediments by over bank flooding, also erosion by the movement of the river bed, and the creation of oxbows (Waters 1992, 129–135). This scenario probably happened several times during the Quaternary and still occurs today (Vaškovský 1977, 104 and 123). In fact, research has identified 8–10 terraces along the Morava River, of which the higher ones are not well preserved (Vaškovský 1977, 105–106 and 118). The fluvial sediments deposited by the river were then reworked by wind activity in the Pleistocene (Vaškovský 1977, 106, 227, 231). These fluvial sediments of sand and aleuritic particles (superfine grain) were transported by wind and deposited as dunes. Along with the purely sand formations, other even finer-grained particles were blown about and accumulated in the form of loess in the lowlands (Vaškovský 1977, 106 and 227). This wind activity would have smoothed out any unevenness present on the plain, with the infilling of any depressions or the deflation of loosely packed elevated features (such as dunes). In the Záhorie lowland, in some places surface drift sands were buried under loess cover and organic sediments (Vaškovský 1977, 115 and 232). Fossil soils are rarely seen in the Záhorie lowland where erosional wind activity was very strong (Vaškovský 1977, 232). In the Holocene, the Záhorie lowland formed the floodplain of the Morava River and soil formation took place on stable surfaces, while bogs formed in the dead arms of the river (Vaškovský 1977, 235). Thickness of deposition along the Morava River varies from 6 to 12 meters (Vaškovský 1977, 123).

Chapter 5 The Survey Phase of the Project at Urbárske Sedliská The chapter details the first two phases of the fieldwork portion of the project.1 There are seven sections to the chapter, with the first section, 5.1, providing geologic background for the region. Section 5.2 is an account of the magnetometer survey, along with some discussion of the use of magnetometers in archaeology. Sections 5.3– 5.4 are expressly concerned with the survey, covering methodology, soil, artifact distribution and types of artifacts recovered respectively. Sections 5.5 and 5.6 provide descriptions of work done to elucidate information uncovered during the survey. The final section, 5.7, is a summary of the chapter. 5.1. Geologic Setting of Stupava and the Surrounding Region The Morava River flows south into the Danube River through a very wide plain, called the Záhorie Lowland on the Slovak side, with the river serving as the border between Slovakia on the eastern bank and Austria on the western (Fig. 5.1). The town of Stupava sits on the edge of the plain, in the lower foothills of the Little Carpathian Mountains. The Záhorie Lowland is a part of the larger Vienna Basin. This basin is an extension of the Pannonian Basin, being separated from it by the last arm of the Carpathian Mountains where they meet the beginnings of the Alps (Ager 1980, 426) (Fig. 5.2).

The Roman-style structure at Stupava (Figs. 1.10 and 1.12) is located on a low hill of these Pleistocene loess and sand deposits, overlooked by the high and thickly forested hills of the Little Carpathian Mountains (Royden 1988, 31; Halouzka 1993, 9).2 The low hill at Stupava has an approximate elevation of 217 m.a.s.l. (Bujna 1976, 494). Areas A and B at Urbárske Sedliská (Figs. 1.10 and 1.12), excavated by Drs. Turčan and Farkaš, are located on a sand terrace above a small stream in the Záhorie Lowland (Turčan 1985, 94).

The appearance of the lowland today is a product of geologic activity starting at the beginning of the Quaternary, with fluvial, eolian, and alluvial deposits reaching depths of 50–200 m in various parts of Central Europe (Ronai 1985, 46; Wessely 1989, 34). In the Záhorie Lowland there are various depressions, which were filled in during the Quaternary Period: e.g., the Zohor-Plavec depression, filled in some areas with as much as 100 m of sediment, while further to the south, the Zohor-Marchegg depression has 85 m of deposits (Vaškovský 1977, 114, 124). The erosion of the Little Carpathians brought down into the plains various materials as sediments that were deposited and then cut through by the Morava River. This action left behind sandy terraces of well rounded gravel, which are mostly quartzite and flint, with some crystalline slate and granitoid (Halouzka 1993, 11). Where the lowland comes closest to the Little Carpathians, the sediments in depressions result from denudation and deposition by strong flowing streams; eolian sands are very rare. Once these rivers and streams moved out over the plain they lost their strong flow, reducing deposition, so that eolian sands predominate in depressions on the plains (Vaškovský 1977, 114).

5.2. The Magnetometer Survey A larger area for the magnetometer survey was laid out than for the pedestrian survey, since magnetometers could cover more ground in a shorter period of time. The magnetometer survey covered two adjoining areas, the first was 80 x 60 m and the second was 30 x 30 m, with the eastern half covering the area marked for the archaeological survey with the four-man crew (Figs. 4.3 and 4.4). The magnetometer survey, as mentioned in Chapter 1.5, was originally intended to help out with the planning of where to place excavation units. Because of the late return of results, the information could not be used as planned and, in fact, the area that yielded ancient 105

cultural features (i.e., the two Grubenhäuser) fell just outside of the area surveyed by the magnetometer crew. The information provided, however, has still proven useful in identifying soil/sediment location, as well as in confirming the course of placement for the drainage pipes from 1982, which led to the discovery of the site in the first place. The results were returned, however, in time enough to allow the placement of one trench (Unit 5) as an experiment to see if the magnetometer results alone could be used to identify cultural features.

added on top of the low hill covering Area 4 of A3/4, being 30 m x 35 m in size. Just to the west was a barbwire fence, at 45° to the survey area, coming closest to the northeast corner of the area. Because of the presence of this fence, with so much metal, we did not extend the second area further north, since it would have brought the magnetometer too close to the fence. By so doing, we fell short 0.5 m of detecting the two Grubenhäuser. The northern boundary of the second area surveyed by the magnetometer was half a meter short of meter mark 15 on transects A–F in A3/4 (Fig. 5.3 and 4.10). Upon closer inspection, Grubenhaus one (feature G1), did partially extend into the surveyed area. The hut’s main axis was oriented northwest to southeast and the southeast corner of the hut extended into the surveyed area. When figures 4.7 or 4.8 are examined, however, there is no evidence of this presence, or at least the small portion that is visible does not stand out in comparison to remainder of the map around it. Since this is such a small portion, the results of this magnetometer survey cannot be used in trying to answer the questions of a magnetometer’s ability to detect Grubenhäuser.

The problem with the use of a magnetometer to find a Grubenhaus lies with the thickness/composition of the cultural stratum. The magnetometer has been very successful in finding Neolithic/Eneolithic fortifications and rondels3 because these cultural features have deep ditches and filled-in pits that exhibit amplified magnetic properties (Dabas and Tabbagh 2000, 336).4 It has been found that the boundaries of archaeological features are more evident the more they are vertical, which is precisely why large ditches and pits would be easily detected (Dabas and Tabbagh 2000, 337). Ancient German Grubenhäuser, on the other hand, are thin features and may in fact have low magnetic properties and at the same time have very shallow boundaries or boundaries that are not sharp, but gradually change over to the soil around them. The question then was whether thin cultural features with blurred boundaries would be able to be detected by the magnetometer. As mentioned in the Introduction, I thought that they should be detected by the presence of ceramics, burnt soils, and other artifacts that would be magnetic enough to be detectable. Fired-structures, pottery, hearths and other features would have thermoremanent magnetism, which would be detectable (Dabas and Tabbagh 2000, 336). As mentioned, we were able to identify the modern drainage ditch, since that had sharp boundaries, but in trying to find Grubenhäuser we were unsuccessful since the surveyed area stopped just short of the two we did eventually find, and the anomaly we thought might be one (Unit 5) did not yield anything.

Approximately 40 meters to the north of the area to be surveyed, a second magnetometer was set up to measure changes in the earth’s magnetic field. The equipment used was a proton magnetometer, PM-2 from Geofyzika Brno; it was set to take one reading every 30 seconds, with the sensor 1.5 m above the surface. The readings from this magnetometer would be used to correct the cesium magnetometer readings based on any fluctuation in the earth’s magnetic field during the period of the survey. When readings were taken, the magnetic field ranged from 48190 to 48210 nT, with the mean value at 48203 nT. The variation locally then was between -9 to +9 nT for the total magnetic field. When the earth’s magnetic field is filtered out and only the residual magnetic field is measured, there was a variation locally of -2.3 to +2.4 nT. The residual magnetic field is the magnetic field exhibited locally by the soil and/or buried features.

The survey was done by Dr. Ján Tirpák from the Archaeological Institute located at the University of Nitra. He used a high-resolution cesium magnetometer (Scintrex SMARTMAG, model SM-4G) (Fig. 5.5) with a The resolution of ± 0.01 nano Tesla (nT) (γ).5 magnetometer had two cesium sensors rigged side by side horizontally at a distance of 0.5 m on a sled that was carried by two crewmen 0.3 m above the ground, while Dr. Tirpák walked behind it carrying the ENVI control console that recorded the readings (Fig. 5.6). The sampling speed was set at five readings every second, with a sampling rate of 0.5 m x 0.25 m. The grid was laid out with transects spaced every one meter, in a north/south direction, while the survey area was 110 m x 60 m (approx. 0.7 ha) (Fig. 5.4). Originally, a survey area of 75 m x 60 m was laid out, with the eastern boundary falling on Transect 1 in A1/2. The survey area was the slope of the low hill heading west toward the drainage channel. During the survey, because the crew was able to cover the area quickly, a second area was

The survey had anomalies caused by surface metals that created high anomalies in the readings. These areas are marked as železo (meaning iron) on the map in Fig. 5.19 and in Fig. 21 as circular areas with letters A–C. When the data were processed using Oasis Montage 5.08, the map created showed a rough alignment of magnetism from southwest to northeast (Fig. 5.7 and also visible in the Surfer 8 map, Fig. 5.8). This direction is probably the result of years of plowing on the field in exactly that general direction. The fence on the east side runs from southeast to northwest and the tractor plowed perpendicular to the fence across the field as exhibited in the magnetometer maps. This direction is further reenforced by the fact that the drainage pipes were laid down in 1982 in the same direction, draining the field toward the southwest and the drainage channel located there. When the data were processed with Surfer 8, the drainage channel that was later uncovered during excavations in Unit 3 can be clearly seen running from 106

the northeast to the southwest (Fig. 5.9, circle labeled with letter B).

By following the transects and digging STPs every five meters, we conducted a systematic field survey, ensuring even coverage (Fig. 5.11). Each STP was 50 cm square, excavated at first to an arbitrary depth of 30 cm (Fig. 5.12). After completing Area 1, it was decided to deepen the STPs to between 50 and 60 cm, with a few random STPs being dug deeper, up to 90 cm (Table 5.1). Forms were made for the crew to fill out as they worked in teams of two (Fig. 5.13). The STPs were labeled in the field notes by what transect and at what five-meter point it was located on the transects: for example, the first STP on transect one was labeled 0 and the second was 5, the third 10, and so forth. In the dissertation, when referred to, each STP is identified by area, transect, and meter mark, so for example the STP located in Area 1, on transect 1, at meter mark 0, would be referred to as 1-1-0, or the STP located in Area 3, on transect A, at meter mark 15, would be referred to as 3-A-15. In this way, the reader will know immediately where to look for a STP on the map (Fig. 5.10).

Since the two Grubenhäuser were discovered just outside the surveyed area, the magnetometer data were examined to see if a test unit could be placed based solely on the data. In examining the map, an oval shape could be seen in the second area surveyed. This circular area covers transects 1, 2, A, B and C, approximately around meter marks 20 through 40 (Fig. 5.9, circle labeled with letter E). The material recovered from the corresponding STPs, however, was small in quantity, and there was no indication of cultural features. Even though we wanted to place a unit onto an anomaly, the circular area was too large for proper investigation, and as already stated, the STPs there did not bode well for success. I decided then to examine the anomaly located at approximately meter 98 on the east/west axis and meter 18 on the north/south axis (Fig. 5.9, circle labeled with letter F). In regards to the STP transects, the anomaly fell between 3-C-25 and 3-D-25 (see map 5.10). The magnetometer picked up an anomaly that was showing high and low magnetism. We thought this might indicate an isolated pit or some kind of buried rock or artifact that had magnetism different from the surrounding soil. A 2 m x 2 m unit, designated Unit 5, was laid out over this anomaly (see figure 6.1 and 6.2 for unit placement). More will be said about Unit 5 in the next chapter when all the units will be described, but it can be said here that Unit 5 was not a success.

Table 5.1. STPs Dug Deeper Than 60 cm STP Depth in cm STP Depth in cm 1-1-10 70 4-E-45 70 1-4-20 80 4-E-55 65 3-A-20 68 4-E-60 70 3-A-25 72 5-F-20 65 3-A-30 90 5-F-40 85 4-A-35 90 5-F-45 66 4-B-35 75 5-G-20 90 3-C-15 64 5-G-25 65 3-D-5 70 5-G-35 65 4-D-55 75 5-G-45 65 3-E-15 72 5-H-30 65 4-E-35 70 5-G-60 90 4-E-40 65

5.3. The Archaeological Survey As mentioned in Chapter 1.5, the survey was switched from a pedestrian survey to one using shovel test pits (hereafter referred to as STPs) because of conditions in the field. The survey area was laid out over an area approximately 70 m x 80 m, divided by ropes into northsouth transects spaced at five-meter intervals (Figs. 5.3 and 5.10). These transects were originally placed for the pedestrian survey, but then adapted for the STPs, with flagging tape placed every five meters. The transects were arranged north-south so that they would be aligned with the magnetometer survey transects. Five areas were demarcated and labeled 1–5 (the fifth area was added toward the end of the project to identify an area that had circular patterns of grass growth, possibly marking cultural features, and also to explore the area near Hut 2) (Fig. 5.3). Area 1 was 25 m x 30 m and area 2 was 25 m x 25 m; areas 3 and 4 were each 25 m x 30 m; area 5 was a narrow strip, measuring 10 m x 90 m, with its north end at an angle because of a fence. The transects in areas 1 and 2 were labeled 1 through 7 from east to west, while in areas 3 and 4, the transects were labeled A through E from west to east, with area 5 having transects F through H (Fig. 5.10). The oddity of this procedure was because areas 1 and 2 were laid out first and labeled, and then later 3–5 were added. The labeling was chosen to minimize confusion when taking notes and bagging material.

Each crew used shovels and sorted the dirt using trowels, while material was bagged and labeled, with quantities being placed on the forms. The use of sifters is standard when STPs are 30–50 cm square and highly recommended, since they catch small artifacts and also overcome the problem of visibility when artifacts are covered in dirt (McManamon 2000, 608; Collins and Molyneaux 2003, 63–64). When STPs are small (often called shovel probes), with only two to three shovels full of dirt removed, then sorting by trowel is the usual course of action (Collins and Molyneaux 2003, 64). Sifters were tried at Urbárske Sedliská, but because of the clayey nature of the soil, they quickly clogged and slowed the teams to such a degree that they had to be abandoned in favor of sorting the dirt by trowel. There are four basic ways of laying out STPs (Fig. 5.14): A) simple random, whereby one randomly places STPs in the survey area; B) stratified random, in which the site is divided into regions and STPs are placed randomly in the regions, but in proportional amounts to the size of the regions; C) systematic; D) the more complex stratified unaligned systematic, which is a combination of the

107

various shovel test pit techniques (Renfrew and Bahn 2000, 77). The systematic method involved laying out STPs along transects that are spaced usually 10 m apart, with the shovel test intervals at 15–30 m, creating a grid. The shovel test intervals can be placed closer for a more fine-grained view of artifact distribution. It should be noted that a staggered or offset grid has been proven more successful than a standard square grid (Collins and Molyneaux 2003, 64).

possible. By digging a STP every five meters, instead of digging every 10 – 15 m, there was a greater probability of discovering cultural features. The probability was further increased by having the transects only five meters apart.6 Planning STPs every four meters or less could have increased still more the probability of success, but considering the small crew at hand, would have been inefficient and time consuming.

The systematic method has been criticized in the past for built-in bias (Nance and Ball 1986, 458; Shennan 1997, 380; Renfrew and Bahn 2000, 77; Collins and Molyneaux 2003, 63). That is, there is a risk you may hit every single example, or miss them, if features within the ancient culture were spaced at certain intervals (Renfrew and Bahn 2000, 77). Other criticisms of using STPs deal with the depth at which sites might be buried and with artifact density. STPs are useful for the discovery of sites that are near the surface and so sites buried more deeply may go undiscovered. In regards to artifact density, if the STPs are spaced too far apart, they may miss features, creating a map of the area having no cultural material, or smaller concentrations than the reality. This bias, however, is in reference to using STPs in field survey over unknown territory, where it has been shown that test-pit sampling is biased against the discovery of small, low density sites (Nance and Ball 1986, 479; Collins and Molyneaux 2003, 63). The use of STPs as a sampling technique has greater success than when STPs are used as a discovery technique (Shott 1989, 396; McManamon 2000, 609). The project here had a specific area where artifacts had already been found, so STPs were being used as a sampling technique to find areas of higher artifact density.

5.3.1. Survey Results: Soil, Stratigraphy, and Overall Artifact Distribution During the survey phase 147 STPs were dug in Areas 1– 4, with another 27 in Area 5. A total area of approximately 3,650 m2 was covered during the survey. Upon analysis of the data a clear division occurred between areas. Areas 1 and 2 were different from Areas 3–5 in stratigraphy and artifact content, even though located next to each other. Areas 1 and 2 (hereafter A1/2) were on the gentle slope going toward the drainage canal 100 m west of the site (Fig. 5.3). Areas 3 and 4 (hereafter A3/4) and 5 were at the top of the low hill, which was relatively flat. Transect 1 lay along the transition between the hill top and start of the slope, and acted as a dividing line between the two. The soil composition and stratification were the first clues that there was a difference. In A1/2, all the STPs had a uniform stratigraphy, except along Transect 1, which, as noted above, was the transition line, sharing characteristics of both areas. In A3/4 and Area 5, the STPs had a more complex stratigraphy, with three distinct strata. The stratigraphy in A1/2 had only two identifiable horizons, which were the root mat (O horizon) and the plough zone (Ap horizon) (Fig. 5.15). The O horizon was almost completely made up of roots, with some soil, trapped decaying organic material, and archaeological material. The A horizon was made up of thick, humus rich soil, with a strong clay component. In texture the soil is a sandy clay loam, with a fine granular structure. The color did not vary much being usually Munsell 7.5YR 3/2 Dark Brown just below the root mat and 7.5YR 4/4 Brown at the bottom of the STPs.

Four other factors also played a role in the original decision to use STPs. The systematic method was primarily chosen because of its simplicity, which a crew of four could do quickly (Renfrew and Bahn 2000, 77), and the use of the standard square grid instead of a staggered one was chosen for ease of set-up. A magnetometer survey was planned to provide more information to help counter any potential bias. The image created by the magnetometer was to have been compared to the artifact distribution seen in the STPs and the actual placement of excavation units would be decided based on that information. As noted in Chapter 1.5, this procedure was not followed because the magnetometer report arrived too late for that purpose; what is more, it was discovered that the STPs with the highest artifact density were, in fact, outside the area surveyed by the magnetometer. The third factor was the expectation of what the STPs would provide. The STP survey was done only to acquire information on artifact density, with the hope that this would lead to the discovery of features or houses. This method is widely used in the United States by cultural resource companies in conducting salvage archaeology with time constraints, and from personal experience I was aware that it was a successful approach. Finally, the number of STPs dug would counter the bias, by ensuring as much coverage as

The stratigraphy in A3/4 and Area 5 was more complex because of the presence of a sand layer below the plough zone, which was generally thinner in this area. STPs in A3/4 revealed a root mat (O horizon) along with the Ap horizon. The main difference was that under the Ap horizon was what can be called a C horizon made up of pure sand (Fig. 5.15). Later expanded excavations revealed a more complicated stratigraphic history (see Chapter 6.1–3 detailing the excavation phase of the project). The plough zone had in fact cut into and obliterated the original B horizon that lay immediately above the sand layer. The only remnants of the B horizon were the sediments deposited in depressions in the sand, thereby helping to show that it had existed at one time.

108

An immediate problem arose in trying to explain why there was sand in one spot and not in another. It was very interesting that the hilltop had in fact sand roughly 50 cm under the surface while only a few meters west, on the slope, there was no sand. In an effort to find out more about the stratigraphy throughout the area, Unit 1 was opened farther to the west down the slope to try to find where the sand began in that place (Fig. 5.3). The unit was a 1 m x 2 m pit and was taken down to a depth of 1.6 m (Figs. 5.4, 5.16 and 5.17). The unit was placed so far to the west in the belief that the sand would continue to the west and would be buried deeper under alluvial fill, enabling us to calculate better the slope of what we supposed to be a buried sand dune. Instead of reaching the missing sand layer, we uncovered a pure clay layer at a depth of ca. 1.4 m. The soil here was also much wetter, because the drainage channel was only 40 m from the test pit.7 It appeared then that the sand disappeared further up the slope at a sharper angle than what one sees today on the surface. A more detailed description of Unit 1 is given in section 4.5.

10% each (except modern brick fragments in A1/2), as can be seen in Charts 5.2 and 5.3. When A3/4 and Area 5 are combined, then A1/2 had only 17% of all the artifacts found and only 16% of the pottery. The artifact amounts fit well with our identification of A3/4 as being the top of a hill upon which cultural activity took place, while artifacts found in A1/2 reached there as a result of slope wash or plowing. The fields in the area are oriented so that the plow travels along the east/west axis. In the later excavation of Unit 2, we actually found old plough furrows running east/west, cutting into the sand layer (see figure 5.4). The ploughs would have either begun or ended on top of the hill, so artifacts would slowly be smeared down slope toward the western end of the field, near the drainage channel (Fig. 5.3). The plow would have actually turned and started its plowing in A3/4. The plow would only just start to dig into the soil in A3/4, and really start turning the soil in A1/2, thereby lessening the damage done to artifact concentrations in A3/4.10 Table 5.2. Artifact Totals from the Surveyed Areas Entire Artifact Survey A1/2 A3/4 Area 5 Type Area 604 96 427 81 Ceramic 55 3 52 0 Bones 19 3 15 1 Lithics 22 2 20 0 Metal 35 19 15 1 Modern Brick 27 5 22 0 Slag 23 4 18 1 Other Total 785 132 569 84

Dr. Farkaš, in the field report for his work in the area, identified the hill as a sand dune (Farkaš 1984), now buried by alluvial fill and soil formation, further modified Our fieldwork by intensive agricultural activity.8 confirms this identification with A3/4 as the top of the sand dune, while A1/2 is the side of the dune, where sand gives way to alluvial fill sitting on top of a clay stratum. The alluvial fill is not surprising since the Morava River is only five kilometers to the west and is within the floodplain. One must also keep in mind that intense agricultural activity has taken place in the region. Today the area is used to grow wheat, corn, sunflowers, and various vegetables. Along with the alluvial fill, there may have been infilling of the area with soil brought in by farmers to help create a more fertile field.

Artifacts found in Area 5 fit into the pattern seen in A1/2 and A3/4. Again, most of the artifacts were pottery, but with an even higher percentage: 97% of the total, with 81 sherds being found. Overall, only 84 artifacts were found and it would appear there is a decline in artifact density as one moved away from A3/4 toward the southeast. This decline is not surprising since there are fewer artifacts in STPs as one moves south in general, and in the southern half of the survey (encompassing all three areas) there is a slope here as well, moving toward the drainage channel located 50 m south of the survey area (see Fig. 5.3). The slope on the south side is even less pronounced than on the west side of the survey area. As mentioned earlier, Area 5 was surveyed to investigate odd circular patterns of grass, one of which lay partly within the southern end of Area 5. Another patch occurred farther to the southeast of Area 5. The general vegetation in the field is of tall grasses, bushes, and young saplings, but in these two places there was only broad-bladed grass. STPs 5H-55 and 5-H-60 were within a patch of tall grass, but did not yield any artifacts except for a stone, spherical in shape, reminiscent of shot or a sling-bullet (but too large for the latter, see the section 4.4.3). Further investigation of the patches could not be done and so their nature still remains unknown.

The magnetometer survey also supports the idea of a buried sand dune with a stronger clay subsurface component to the west. Along with the report from Tirpák detailing his magnetometer survey, a map was also given on the soil make-up of the area. On the magnetometer map (Fig. 5.18) one may see a clear transition that corresponds to presence of the sand and where the soil and clay begin (see also the description for Unit 1 in section 4.5). The fact that most cultural finds would be on top of a hill is understandable given that most people would want to live at the top, where it is level, not on a slope. 5.3.2. Types and Quantities of Artifacts from the STPs Another difference determined by the survey was that STPs from area A1/2 yielded only 135 artifacts compared to the 582 found in Area 3/4 (Table 5.2).9 When using the information from Table 5.2, then Area 3/4 had 81% of all the artifacts and 82% of the pottery. The categorization by percentage of different artifacts was similar in both areas, where pottery made up around 74% of all artifacts found, with the other types being under 109

After Bajč-Retz, there does not seem to be much activity on the site until the Roman Period. Bronze Age and Iron Age cultures do not seem to be present. Except for one Bronze Age body sherd, no other artifacts of Bronze Age or Iron Age cultures were identified in the survey. There is, however, as noted above, a large group of sherds that are ancient but could not be fully identified because of their fragmentary condition; some Bronze Age or Iron Age pieces may be among them. The survey also did not find any La Tène-period pottery, which is surprising, since there was a strong Celtic presence in Slovakia (two sherds were found, however, in Hut 1, see Chapter 9.4.2).

5.4. Discussion of the Artifacts In this section, the artifacts will be discussed by group in order to provide an overview of what was found. For a complete itemization of what was found in each STP see Appendix A, which is a chart listing all the STPs. A catalog of the more important non-Roman Period artifacts from the STPs is given in Appendix B in combination with artifacts from Units 1 and 5–8 and Strata 1 and 2 of Units 2–4. 5.4.1. Pottery11

The main focus of the survey had been to find any evidence of the Roman Period and the ancient German culture in the region. The survey did not disappoint us. Forty German sherds were found along with four Roman. These make up only 9% of the total amount of pottery found, but some are of sufficient quality for identification purposes. Another important aspect is the concentration of artifacts. Two STPs had high concentrations of German material: STP 3-C-15 had 12 pieces and STP 3E-15 had 6. The majority of STPs had either no German sherds or only one to three pieces. STPs 3-C-15 and 3-E15 also had the sherds mixed with bones in a very dark soil, which was on top of the underlying sand in area A3/4. The concentration and amounts focused our attention on the area immediately around these two STPs for the third phase of the project, when units were excavated for the goal of finding German cultural features. Based on the few diagnostic pieces found during the survey and compared to other excavations in the region, the ancient German component of Urbárske Sedliská dates to some time during the first half of the second century A.D.

One surprise of many at Urbárske Sedliská was the variety of periods represented by the ceramics. The focus of the project was to locate and excavate German cultural features, but the survey showed us that the site has a much richer history and a longer period of occupation. Since most of the ceramics came from the plough zone, however, the pottery was very fragmented, thereby making identification difficult. In the end, 57% of the pottery was identified only as being ancient and could not be assigned to a specific period or culture (Table 5.3). The earliest pottery came from the Linear Ceramic culture, which is the first identified Neolithic culture in Central Europe, also referred to as Linear Bandkeramik (LBK), spanning roughly 5600/5500–4900/4800 B.C.12 and in Slovakia from 4800–3800 BC is the Late LBK Želiezovce group (Neustupný and Neustupný 1961, 39; Farkaš 1990, 74 and 136; Nowak 2004, 11; Pavlů 2004, 83; Zvelebil 2004, 194). The sherds found during the survey are very fragmentary. A total of 60 pieces were recovered, of which there were 55 body sherds, three rims, one handle, one piece with a nodule, and one of the body-sherds had a decoration of an incised line. The fabric and color varied, ranging from grayish brown (Munsell color 10YR 5/2) to a reddish yellow (7.5YR 7/6) to a light yellowish brown (10YR 6/4). The rim pieces were too small to reconstruct body shapes, and the handle fragment could not be further identified. The piece with the nodule is a body fragment and most likely came from a closed vessel (jug or amphora), but the exact form is unknown.

After the Roman Period, there is a fall-off in ceramics until the modern period. Only ten pieces could be identified as belonging to the Medieval and Renaissance Periods, while 95 sherds could be identified as being modern. Among the pieces from the Middle Ages are a rim and a body sherd dating to the 13th century. There is great variety in pottery types from the modern period, ranging from poor-quality ceramics to glazed pieces to porcelain.

The major cultural period immediately after LBK is Lengyel (3800–2800 BC), which straddles the Late Neolithic and Early Eneolithic (Farkaš 1990, 110). We did identify three pieces from the Lengyel period; two body sherds and a rim, but as to when within the Lengyel period they fall could not be determined. Wedged between Lengyel and the later Baden culture is a brief cultural period known as Bajč-Retz (3000–2800 BC), at the break between the Early and Middle Eneolithic. Two body sherds turned up in the survey, which had the traditional form of decoration: punctured furrows (in German Furchenstich, “stab and drag”) in geometric patterns, sometimes filled in with white incrustation (Farkaš 1990, 123).

110

(1996, 154) and some of the conchoidal flakes may be from the production of blades. The source for the stone used is not known at the moment, but from the study of other sites dating to the Neolithic in the Bratislava region to the south, source material came from local places, such as the mountains nearby or the ancient river terraces of the Danube or Morava river, but some material was imported from as far away as Poland (Farkaš 1990, 119; Farkaš and Novotný 1993, 46).

Table 5.3. Pottery Totals from the Surveyed Area Entire Area Period/Type Survey A1/2 A3/4 5 Area 85 7 60 18 LBK 5 1 2 2 Lengyel 2 0 2 0 Bajč-Retz 4 0 1 3 Bronze Age 0 0 0 0 Iron Age 41 8 32 1 Anc. German 4 1 3 0 Roman 10 2 7 1 Middle Ages 1 0 1 0 Renaissance 103 34 61 8 Modern Ancient 347 42 257 48 Unidentified Total: 602 95 426 81

5.4.4. Brick and Slag These two categories have been combined because of the limited information that can be given for them at this stage of the analysis. A total of 35 modern brick fragments were found during the survey. A brick house had earlier stood near the field, but was torn down during the early 1990s (Fig. 5.3). Remains of the house are still present and the brick fragments in the field most likely came from there.

5.4.2. Metal

An interesting aspect of the survey was the discovery of slag. A total of 27 pieces were found, with 22 coming from A3/4. Two types are present, according to observations in the field: those from metal production and those from either glass production or from glass melted by high heat. The slag most likely did not come from the house that was torn down since there is no evidence of any burning that could have created slag. Another possible source are the nearby private garden plots, which locals fertilize in various ways, including the use of compost heaps, possibly containing all types of trash. Ancient metallurgical activity or glass working cannot be ruled out, however, because the nearby Grubenhaus excavated by Dr. Turčan in 1982 had evidence for jewelry making, including crucibles and slag (Turčan 1985, 114).

Twenty-two metal items were found in varied states of preservation. Identification, when possible, was done with the help of Dr. Turčan and the museum’s curator of metal artifacts, Mr. Slovak. Preliminary identification took place in the field and further examination was done after the metals were cleaned at the museum. Since Urbárske Sedliská was a field in active use, modern items included pieces of farm machinery (chain-belt, bolts, and nuts13) and farming tools (such as the blade from a hoe and a shovel). The number of metal pieces from other periods in recent history was also quite high. The Russian army during WWII had marched through this area and had actually engaged German troops stationed in Stupava (on top of the Roman-style building).14 Some pieces might date to this period, but none of the items could be identified with that amount of precision. Many pieces were so corroded it was impossible to determine their original function.

5.4.5. Bone A total of 55 pieces of bone was found, most coming from A3/4 (Table 5.4). Many pieces came from the surface or plough zone, but two STPs (3-C-15 and 3-E15) had bone mixed with pottery in a dark soil right above the sand layer. Since most of the bones (outside of those from 3-C-15 and 3-E-15) are from the plough zone, it is hard to say to what period the bones may belong. A potential source for some of the bones found in the plough zone could be the slaughterhouse located approximately 100 m from the survey area. Since the bones primarily came from upper strata, they were not sent to the lab for identification.

5.4.3. Lithics15 Eighteen lithics were found during the survey. When originally examined, the group was subdivided into ten flakes, three cores, two blades, one possible tip of a blade, one possible perforator, and one notched flake (Vrba 2004, 143–144). Upon more detailed examination and a better understanding of stone-tool technology, it is clear that the lithics were less varied. There are 13 flakes, 3 cores and 2 blades. Some flakes exhibit signs of percussion flaking, while others are flake shatters, being random fragments broken off during production. Their appearance at the site, mixed with pottery, places them in the microlith industry of the Neolithic Period, within either the LBK or Lengyel cultures (more Neolithic pottery was found in the survey than pottery of any other time period). The Linear Ceramic cultures stone tools are described by A. Whittle as the “basic flint toolkits, being based on the controlled production of blades…”

The bones from STPs 3-C-15 and 3-E-15 on inspection in the field seemed to be bovine, while the pottery was primarily of ancient German manufacture. The soil had charcoal flecking and appeared to be immediately below the plough zone. The bones that came from these two STPs, having been later identified as coming from cultural features G1 and G4 (discussed in detail in 111

Chapters 7.2.1, 7.2.5, and 10.5.2–3), were sent to the lab for identification and added to the bone counts of those features.16 Table 5.4. STPs with Bone STP Amount 1-5-5 1 2-1-35 1 2-7-50 1 3-A-20 1 3-A-25 1 3-B-15 1 3-B-20 1 3-B-25 2 3-B-30 1 3-C-0 2 3-C-15 14

STP 3-C-25 3-D-0 3-D-10 3-D-25 3-D-30 3-E-15 3-E-5 4-B-45 4-C-40 4-E-40 Total

being the transition zone between the two horizons. There was tiny mottling in the soil, with a lighter area in the middle of the unit as if indicating ash or some other unknown material, but no artifacts were found to indicate anthropogenic origin (or any strong indications that the mottling represented ash; i.e., no charcoal was present in any form or size). At the start of excavating stratum 4, because of the depth of the unit, now almost a meter deep, it was decided to step the unit inward on one side to allow easier access into the unit. A step was created in the east baulk at a depth of 90 cm (Fig. 5.17). The unit was then dug to a depth of 130 cm, where a step was created again on either side of the unit, creating a 0.8 m x 1 m area roughly at the center of the unit as a probe for deeper excavation. The drastic shrinkage in the unit at this depth was done because of the nature of the stratum, being now made up primarily of clay with little sediment and chunks of rock. The original Stratum 4 was later subdivided by texture and inclusions. Stratum 4A has the overall same color as stratum 3, but the mottling continued and became stronger with small pieces of rock and flecks of clay (Fig. 5.17). Stratum 4B can be better identified as a Bt horizon (because of the clay), with no sand being present, but clay becomes more common and in larger chunks, and at the same time there were larger rocks present in very poor condition, seeming to be decaying sandstone or limestone (Fig. 5.17). Toward the bottom of 4B, at a depth ca. 130 cm, the stratum was choked with rocks and chunks of clay. The final stratum, labeled 5, can be identified as the C horizon and was almost pure clay, in which no artifacts were found (Fig. 5.17). The clay was also wet, having a light olive grey color (5Y 6/2), with striping of brown colors. After taking the unit down another 36 cm, through pure clay with no artifacts being found, we decided to stop digging.

Amount 1 7 0 1 0 15 2 1 1 1 55

5.4.6. Other Various other items were recovered, such as sand conglomerates, bog iron, plastic, modern glass, cement, stucco, and one coin: amounts are listed in Appendix B. Modern glass (such as from soft-drink bottles) and plastic were not collected. 5.5. Unit 1 As mentioned above, Unit 1 was excavated to investigate the sand layer found in A3/4, but absent in A1/2. The unit is located 5 m west of STP 2-7-35 and was 1 m x 2 m in area with the long axis running east/west (Fig. 5.10). As mentioned above, the unit was placed at this location in the belief that we would reach sand at a deeper level than on top of the hill, thereby allowing us to reconstruct the buried slope of the sand, and enabling us to determine whether it was a continuous layer extending over a wide area with a gradual slope, or was in fact a dune or terrace with an abrupt slope. During excavation, five strata were identified, but upon examination of the baulks for drawings and photography, seven strata were recognized, with strata two and four being subdivided (Figs. 4.16 and 4.17). These divisions are based on the slightest changes of color and consistency. As we worked down, the stratigraphy changed from a primarily sandy loam to a clay loam to pure clay.

Artifacts were found only in strata 2 and 3. Since artifact recovery and recording during the excavation went by strata identified in the field, items from 2A or 2B are identified only as coming from stratum 2 in the artifact discussion below. That is, I have not tried to reconstruct which sherds belonged to 2A or 2B. Table 5.5. Artifact Totals from Unit 1, Strata 2 and 3 Artifact Stratum Stratum Type 2 3 18 31 Pottery 3 1 Metal 1 1 Bone 1 3 Misc. Total: 23 36

The upper layers were similar to those of the STPs: root mat (stratum 1, the O horizon) and plough zone (stratum 2A and 2B, the Ap horizon), made up of a dark (10YR 2/1, black) sandy clay loam (Fig. 5.15). Stratum 2B exhibited a slightly lighter color, being a 10YR 3/1, very dark gray. Strata 3, 4A, and 4B made up the B horizon (Fig. 5.17). Stratum 3 was a sandy clay loam, but with less sand as seen in strata 2A and 2B, but the color was the same as stratum 2A, 10YR 2/1, black (Fig. 5.17). In fact, stratum 3 could be identified as an AB horizon,

As can be seen in table 5.5, 18 sherds were found in stratum two, of which ten could be identified as modern (six of which were rims), but the remaining 8 could only be identified as ancient undetermined. The metal items were in a much corroded condition, but did not appear ancient and one may have been machine made. The miscellaneous piece was an iron-oxide sand conglomerate, probably bog-iron. Stratum 3 yielded the most artifacts (Table 5.5), with almost twice the number 112

of sherds being recovered. Of the sherds, three bodysherds were identified as ancient German (Fig. 5.19), on the basis of the decoration, along with 15 possible Lengyel pieces; the remainder were ancient undetermined. Only one Lengyel sherd, a rim (Inv. No. 972), was diagnostic. The miscellaneous items included one odd stone, very smooth, but pock- marked, along with two pieces of what could be wattle-and-daub, while the metal was probably a much corroded nail.

we walked the area in August, we noticed that locals had garden plots to the north of the ruined house in a cleared area (Fig. 5.20). In October, after the vegetables had been harvested, the garden plots were left alone, thereby leaving a cleared area, devoid of plants. One area was chosen for survey, roughly 10 m x 40 m in size, with the long axis running east/west, and labeled Area 6 (Fig. 5.3). One worker and I did a surface survey along this area, keeping ten meters apart, working our way down the length, collecting whatever was on the surface around us.

As stated above, we did not find the sand stratum as expected and even though the discovery of the clay was a surprise at the time of excavation, in retrospect we now see that its discovery might have been expected. The drainage channel is only another 50 m to the west and the ground is wet and probably saturated during parts of the year (Fig. 5.3). This saturation by water may have caused the leaching of clays from the upper horizons to the lower horizon, helping to form the clay layer. Because of limited time and the exhausting work of trying to dig through 100% clay, we did not dig through this layer to see what lay underneath. The magnetometer map shows a distinctive line of mid-range magnetism running north/south in the exact area uncovered by Unit 1, possibly signifying that the line represents buried clay (Fig. 5.18).17 In regards to artifacts, it is interesting that stratum 2, the plough zone, had in fact fewer artifacts than stratum 3, which is below the plough zone. Stratum 3 is not a buried paleosol, since there is no overlying sediment and no anthropogenic features were present to indicate a preserved ancient stratum. Pedoturbation and slope colluviation probably explain the artifacts as found, but it is interesting that stratum 3 had the largest concentration of Lengyel period sherds on the site (if the sherd identification is accurate). The idea that there might be a buried Lengyel feature in the area of Unit 1 seems unlikely, since the preferred habitation was on terraces, not in the midst of wet soggy ground (Farkaš and Novotný 1993, 52). Any Lengyel component to the site should be expected at the top of the low hill, in A3/4.

The hypothesis that the archaeological site continued east was confirmed but in a surprising way. A total of 57 sherds were collected along with a possible piece of wattle-and-daub. Of the 57 sherds, 47 were LBK ceramics, of which two were definitely from Early LBK. Two knobs from the bodies of two vessels were definitely from the early phase of LBK (Fig. 5.21). Area 6, in conjunction with the finds from the STPs, confirms that Urbárske Sedliská has a very strong Neolithic component dating to the LBK and, it would seem, of the earlier phase of the culture. It is hoped that in the future, further investigation of this area will be possible to explore the Neolithic component of the site. 5.7. Conclusions for the Survey Phase of the Project Urbárske Sedliská is a much more complicated site than first envisioned. The field survey did provide enough evidence of ancient German habitation for the second phase of the project to be successful in locating cultural features. With the presence of pottery from so many periods, however, it became clear that further exploration of the area and adjoining fields needs to be done to understand better the total history of the site. The survey showed that Urbárske Sedliská was occupied periodically from the Neolithic to the present. When looking at pottery percentages, the greatest amounts that could be identified came from the Neolithic/Eneolithic, Roman Period, and modern times. The pottery densities during these periods indicate a greater human presence. Even though the focus was ancient German, the number of Linear Ceramics found at the site also requires further study and it is possible that structures from that period may exist at Urbárske Sedliská. Of the total number of ceramics found during the survey, ancient German sherds make up only 8% while Linear makes up 13% (Chart 5.5). Further investigation for a LBK settlement is needed, especially if it falls within the Early LBK period. At the moment, no long-houses from the older period have been discovered, and most of the older period artifacts come from isolated pits (Farkaš 1990, 74). If Urbárske Sedliská has a preserved long-house, it would prove valuable to our understanding of the older LBK sequence in southwest Slovakia.

5.6. Area 6 During the survey and upon analysis of finds, it was clear that material was being found as far as the fence located to the east of the project. Immediately on the other side of the fence was a thicket of young trees, while just to the northeast was the ruin of the house previously mentioned (Fig. 5.3). Farther east from there was a narrow road, which serviced the slaughterhouse and processing buildings of the meat-packing installation. On the other side of the road were a grassy area and then the slaughterhouse itself, located 200 m from our site. Investigation was not possible immediately to the east of the fence because of the thicket of trees; to the northeast were the house ruins and very high grasses around it. Because artifacts were found up to the fence, it was postulated that the archaeological site most likely extended further east. We wanted to take a look to see if there was any way of confirming this hypothesis without doing any prolonged survey or study of this area. When

Two STPS, 3-C-15 and 3-E-15, had a high density of German pottery, including diagnostic pieces. These pieces were dated to the first half of the second century 113

AD, thereby providing a rough idea for the period of occupation. The presence of animal bones and a very dark organic rich soil was also an important indicator that these two STPs were important. The fill of a Grubenhaus is usually a very dark soil made up of cultural material and organic material, such as animal remains, wood and charcoal. As mentioned in Chapter 1.5, when excavation units were opened at the two STPs, two German Grubenhäuser were discovered and will be discussed in the following chapters.

pits in survey, looking at their reliability and how best to do test-pit sampling (Nance and Ball 1986). Density of artifact concentration plays a large role in the success of test-pit sampling (Nance and Ball 1986, 471–472). In this regard, I note that Urbárske Sedliská has a high artifact density. 7. In fact, within two or three days there was standing water to a depth of approximately 10 cm above the clay; then, after a few rainy days, the unit filled with water up to a depth of approximately 70 cm.

Endnotes

8. In a conversation with Dr. Farkaš about the sand dune, he postulated that the dune’s formation would have happened during the Pleistocene.

1. Portions of this chapter have been published in, “Archaeological Survey at Urbárske Sedliská, Stupava, Slovakia: Preliminary Report” (Vrba 2004, 129–156).

9. I am not including Area 5 in these figures because A1/2 and A3/4 are roughly the same size. If Area 5 was added to A3/4, the balance would be skewed: the difference in artifact counts would be determined more by size of the areas surveyed than by the fact of actual artifact density alone.

2. The Little Carpathians were uplifted in the Middle Miocene and have seen extensive erosion since, thereby creating sedimentary input into the river valley (Royden 1988, 31; Halouzka 1993, 9). Today they are more akin to high hills than mountains, and are covered by thick forests.

10. The most common plow type would be the chisel plough used in conjunction with a disc plough. These two plows usually go to a depth of 20–30 cm (How cultivation 2002). Every so often, there is the need to turn up deeper soils and so a ripper, or subsoiler, is used and this type of plough goes to a depth of 30–90 cm (How cultivation 2002). Turčan mentioned to me that the subsoiler has been used in the area where the survey was conducted.

3. Rondels first appear during the Lengyel Period and are large round compounds that have a central space encircled by a wooden palisade with four gates, which in turn is encircled by a series of ditches and embankments (Farkaš 1990, 113). Some scholars have argued against rondels being fortified structures, protesting that they have too many gates and the interior space is too small (Farkaš 1990, 113). They argue that rondels are cult sites, with four gates lined with the four cardinal directions, or oriented with the stars, much as at Stonehenge, while others argue for use as a communal gathering place in times of danger (Farkaš 1990, 113). The rondels may have also been elite residences. The problem with answering these questions is that no settlement has been excavated to show if they, too, were fortified with palisades and ditches (Farkaš 1990, 113).

11. Pottery was identified with the help of Dr. Vladimír Turčan (Roman and Late Antiquity), Dr. Zdeněk Farkaš (Mesolithic and Neolithic), Dr. Juraj Bartík (Bronze Age), and Mgr. Igor Bazovský (Roman) who are all part of the Archaeological Museum, SNM. I am grateful to them for their help, although any errors remain my own. 12. These dates are in flux and those provided are taken from I. Pavlů’s 2004 article (Pavlů 2004, 83).

4. In fact, Dr. Tirpák has been successful with the same equipment in Slovakia in the discovery of precisely the types of features mentioned, the rondels and fortifications dating to the Neolithic and Bronze Age. He has also been successful in surveying Roman marching forts in Slovakia, since once again ditches are involved.

13. My field assistants told me that the Soviet-Union-era combines were famous for breaking down quickly and also for loosing parts as they worked across the fields. The combines were copies of American models. 14. This information was told to me by Dr. Turčan and in fact is one of the reasons excavations are complicated at the Roman-style structure. The Germans had built barracks on, and dug fox-holes across, the site. The cement foundations of the barracks are still on-site and the fox-holes are constantly being rediscovered, along with ammunition from the war.

5. Technical data were taken from the Scintrex brochure for the product. In regards to the resolution, the sensitivity is high since the “intensity of the total earth magnetic field in Europe ranges from about 45,000 to 49,000 Nanotesla, the diurnal variations are in the range of 10–30 Nanotesla, and is furthermore depending upon the sun activity” (Fassbinder and Irlinger 2001, 14).

15. I thank Dr. Curtis Runnels, Boston University, for his help in identifying and better understanding the lithics recovered from the site. Any errors that remain, however, are my own.

6. It is without question that the success of the survey increases proportionally to the number of STPs dug. Nance and Ball provide a detailed analysis of using test 114

16. This action is now regretted, but it was thought at the time there was a need to keep material from the cultural features together. Even though it is necessary to keep materials together that go together, it is lamentable that the bone materials from the STPs were not recorded as they should have been, separately and part of the survey phase of the project. 17. If the map is correct in that the band is clay running roughly north-south, parallel to the modern drainage channel, then it is possibly that this clay band may be the bed of an older, abandoned drainage channel. This theory is based on the fact that the clay is in fact merely a band and does not underlie the entire area of low ground near the modern drainage channel. The amount of rocks found in stratum 4 above the clay bed may indicate that stratum 4 was the water channel. Since Unit 1 was placed by sheer accident in the middle of the ‘clay band’, we did not excavate the edges, which may or not have had the characteristics of having been the banks.

115

Charts: Chart 5.1. Percentages of Artifacts Found in the Entire Survey Area

Brick 4% Metal 3%

Slag Other 3% 3%

Lithics 2% Bones 7%

Pottery 78%

Chart 5.2. Percentages of Artifacts Found in Area 1/2

Slag Other 3% 4% Brick 14% Metal 2% Lithics 2% Bones 2%

Pottery 73%

116

Chart 5.3. Percentages of Artifacts Found in Area 3/4

Brick 3% Metal 4%

Slag Other 4% 3%

Lithics 3% Bones 9%

Pottery 74%

Chart 5.4. Percentages of Artifacts Found in Area 5

Lithics 1%

Brick 1%

Other 1%

Pottery 97%

117

Chart 5.5. Percentages of Pottery by Period/Culture in Entire Survey Area

Renaissance 0.2%

Modern 17.1%

Middle Ages 1.7% Roman 0.7% Ancient German 6.8% Iron Age 0.0% Bronze Age 0.7%

Ancient Unidentified 57.6%

Bajč-Retz 0.3% Lengyel 0.8%

LBK 14.1%

Chart 5.6. Percentages of Pottery by Period/Culture in A1/2

Modern 36% Ancient Unidentified 45%

Middle Ages 2% Roman 1% Ancient German Lengyel 8% 1%

118

LBK 7%

Chart 5.7. Percentages of Pottery by Period/Culture in A3/4 Renaissance 0.2%

Modern 14.3%

Middle Ages 1.6% Roman 0.7% Ancient German 7.5% Bronze Age 0.2% Lengyel 0.5%

Ancient Unidentified 60.3%

Bajč-Retz 0.5% LBK 14.1%

Chart 5.8. Percentages of Pottery by Period/Culture in Area 5 Middle Ages Modern 1% 10% Ancient German 1% Bronze Age 4% Lengyel 2%

Ancient Unidentified 60%

LBK 22%

119

Figures

Figure 5.1. Map of southwest Slovakia showing location of Záhorie Lowlands (based on a map from the back cover of Kováč and Michalík 1991).

120

Figure 5.2. Map of the Pannonian and Vienna Basins (based on a map in O’Nions and Ballentine 1993, 143).

Figure 5.3. Map showing location of survey area and its subdivision. The dotted line represents area surveyed by the magnetometer (Vrba 2004, 131).

121

Figure 5.4. Map showing area covered by the magnetometer, which is here shown by the solid lines, while the dotted lines represent the five areas surveyed by traditional methods (Vrba 2004, 132).

Figure 4.5. View of the cesium magnetometer.

122

Figure 5.6. Dr. Tirpák and his crew during the survey.

Figure 5.7. Map of surveyed area produced by Dr. Tirpak using Geosoft software, showing the residual magnetic anomalies; zelezo indicates metal at the surface.

123

60

8 7 6

50

5 4

40

3 0m

10m

20m

2 1

30

0 -1

20

-2 -3 -4

10

-5 -6 -7

0 0

10

20

30

40

50

60

70

80

90

100

-8

Figure 5.8. Map of Residual Magnetic Anomalies (using Surfer 8), showing magnetism between + 8 and – 8 nT).

60

8 7

50

6

A

5 4

40

3 0m

B

10m

20m

2 1

30

0

C

-1

20

F

D

E

-2 -3 -4

10

-5 -6 -7

0 0

10

20

30

40

50

60

70

80

90

100

Figure 5.9. Same map as in Figure 4.8 but with various anomalies circled and lettered. A–C indicate metal at the surface D indicates the drainage channel and pipe laid in 1982 E indicates a large circular anomaly, unknown origin F indicates the small anomaly investigated by Unit 5

124

-8

Figure 5.10. Map showing layout of transects in the five areas and the position of Unit 1 (Vrba 2004, 132).

Figure 5.11. Crew working along transects 1 and 2 (Vrba 2004, 133).

125

Figure 5.12. Shovel test pit at transect 1, meter mark 5 (Vrba 2004, 134).

126

Figure 5.13. Example of the form used during the survey (provided here in English, but a Slovak version was used for the crew) (Vrba 2004, 152).

127

Figure 5.14. Types of sampling using shovel test pits. (A) Simple Random; (B) Stratified Random; (C) Systematic; (D) Stratified Unaligned Systematic (based on an image from Renfrew and Bahn 2000, 77).

Figure 5.15. Two example STPs for the two types of stratigraphy encountered on site (Vrba 2004, 136).

128

Figure 5.16. Photograph of Unit 1 after excavation, with strata marked on north baulk.

Figure 5.17. Drawing of the strata for Unit 1.

129

Figure 5.18. Magnetometer Map of the magnetic readings of the soil. The areas marked with letters A, B and C are errors in reading because of the presence of metal objects at the surface (Vrba 2004, 138).

Figure 5.19. Image of the German sherds from Unit 1.

130

Figure 5.20. Area 6 is on the other side of the ruined house, which is in the foreground.

Figure 5.21. LBK knobs 1 and 2 from Area 6.

131

It was noted that at the bottom of stratum 2, plough furrows were seen going into the sand at the west end of the unit, oriented northwest to southeast, in alignment with the fence nearby (Figs. 6.1 and 6.4). In the northeast corner a darker staining was noticed both in the bottom of stratum 2 and in the baulk itself (Figs. 6.5 and 6.6). The staining was photographed and acknowledged in the field notes, but nothing more was thought of this staining. It was only after the discovery of the Grubenhaus (designated G1) in Unit 3, right next to Unit 2, that we realized we had excavated a corner of the hut in the northeast corner of Unit 2 and that we had actually bisected the southeast posthole (Fig. 6.7). Once this situation was realized, we came back to Unit 2 and extended the probe in the northwest corner along the entire length of the North baulk (through sectors C1–E1), thereby revealing the southeast posthole of Hut 1. The reason the corner was not identified sooner was part of the greater problem of the hut’s preservation on its south side. The south side was not as crisply preserved as the east and west sides, or as was Grubenhaus G4 (Hut 2). The poor preservation on the south side was possibly caused by the presence of two features (which might actually be one) designated G2 and G3, which disturbed the preservation of the hut at some later date. These features will be presented in Chapter 7.2.2–4.

Chapter 6 The Excavation Phase of the Project at Urbárske Sedliská Chapter 6 is organized by unit, with each one having a general description with artifact composition. In the end, a total of 16 units were excavated, of which most were expansions of an original unit (primarily units 3 and 4).1 Table 6.1 lists the units and their sizes. Units making up the Grubenhäuser (Units 3, 3A–E, 4 and 4A–C) are described as a group for each hut, but only those nonRoman Period artifacts from strata 1 and 2 are discussed in this chapter. The detailed description of the RomanPeriod artifacts, the two Grubenhäuser, other cultural features, and the artifacts uncovered in them during the excavation will be provided in Chapters 7–10. Table 6.1. Units and Their Size in Meters Unit Size Unit Size 1* 1x2 4 5x5 2 3x5 4A 1x2 3 2x2 4B 1x3 3A 3x2 4C 1x2 3B 5x2 5 2x2 3C 5x2 6 1x2 3D 5x1 7 1x2 3E .5x6 8 2x5 * Already discussed in Chapter 5.5

The artifacts uncovered were a mirror of what was found in the STPs, but on a larger scale. This time, however, there was a greater amount of ancient German pottery, which is not surprising since Unit 2 was just outside the south side of Grubenhaus G1. Charts 6.1–6.4 provide details on the amounts (in number and percentage) by period/culture of the pottery and other artifacts in strata 1 and 2. The corner of the hut was found primarily in sector D1 of stratum 2, with some in E1, and the pottery from these sectors was added to the totals of the Grubenhaus and is not reflected in the totals shown in the charts for stratum 2.

6.1. Unit 2 Once the surface survey was completed and two STPs had provided enough material to warrant opening larger units, a question arose as to where to place the first unit. STP 3-C-15 was chosen to be examined first, but it was thought best not to place the unit directly on top of the STP. Unit 2 was placed 50 cm to the south of the STP and its long axis oriented east/west (Figs. 6.1, 6.2). The unit was divided into sectors, each 1m x 1 m in size labeled A1–3, B1–3, C1–3, D1–3, and E1–3 (Fig. 6.3). This division was created for a better control of artifacts as they were found and bagged.

6.2. Units 3, 3A–3E With the apparent failure of locating a feature in Unit 2, Unit 3 was laid out as a 2 m x 2 m trench 50 cm to the north of Unit 2, next to STP 3-C-15 (Figs. 6.1 and 6.2). The original reason for the location was because of the plough furrows seen in Unit 2. It was hoped that if the plough was moving northeast to southwest, material was being dragged from a possible cultural feature northeast

The unit was at first taken down only two layers, the topsoil/root mat and then the underlying plough zone. Below stratum 2 we came upon the sand layer seen in the STPs of A3/4. It was decided that instead of excavating into probable sterile sand, we would put in only a probe in the northwest corner, 2 m x 1 m in size, and went down 20 cm; no cultural material was found. At this point, we opened Unit 3, 50 cm to the north of Unit 2 and 50 cm to the east of STP 3-C-15, thinking that Unit 2 had nothing more to offer, since only pottery was recovered but no cultural features.

of Unit 2, which the STP had revealed to be in the area. After removing the root mat and topsoil, at around 40 cm an intact cultural layer was discovered. The soil was dark black, with charcoal flecking, filled with animal bone and pottery, while in texture it had a wet/greasy feel. The methodology in Unit 2 had been to excavate arbitrary 5cm layers, within each natural stratum, in each sector, moving from one sector to another. In this case, with the discovery of the feature first in sector B2, work was stopped there and the other sectors were excavated to the same level. The dark soil covered the entire unit and it 132

was decided to expand the unit and excavate to the same stratum until the entire feature was revealed. Once the feature was entirely revealed, then it would be excavated as one piece. This approach required the addition of Units 3A, 3B, and 3C (Fig. 6.9). At a later point, the southwest corner of the hut was seen to extend into the west baulk so Unit 3D was added to reveal that corner. Upon completion of the excavation of those units, the baulk between Units 2 and 3 was removed as Unit 3E, since the southeast corner of the hut extended through the baulk into Unit 2 (Fig. 6.10). Hereafter, all of the units will be collectively referred to as just Unit 3.

earlier, Dr. Turčan had thought there was a posthole at the bottom, which would indicate we definitely had part of a hut already.3 With the removal of the first two strata, it became quickly apparent that indeed there was a second hut first revealed at STP 3-E-15 (Fig. 6.12, 6.15). Once again, the unit was too small, with portions of the hut going into the baulks and the unit was expanded to a 5 m x 5 m trench. Three adjunct units (Units 4A, 4B, and 4C) were added to reveal the entire hut, before it was excavated (Fig. 6.13, 6.14). A wide range of material was recovered once again in the first two strata. There was a mixture of pottery as seen elsewhere on the site. Once again spindle whorls were recovered, this time only in stratum 2 (see Charts 6.6, 6.8). With the discovery of so much material in the upper strata, not all of it necessarily coming from Hut 2, it is entirely possible that there are other huts in the area. Charts 6.9 and 6.10 provide the types of pottery and artifacts recovered from both strata, by amounts and percentage found.4

At first, the feature was thought to be a trash dump or pit, but once the shape was revealed it became apparent that this was a Grubenhaus, being roughly rectangular and at most 4 x 5 meters in size (Fig. 6.11). Another discovery was also made: in the north portion of Unit 3, a modern trench had cut through the Grubenhaus, destroying the northwest section of the hut (Fig. 6.11). This trench was part of the drainage project from 1982, when pipes were laid throughout the area to help drain the soils for better farming, and which led to the excavations by Dr. Turčan and Dr. Farkaš. When part of the ditch was excavated a plastic pipe wrapped in a canvas cloth was revealed at the bottom, placed there to enhance the flow of water down slope toward the drainage ditch approximately 100 m to the west.2

6.4. Unit 5 As mentioned already in section 5.2, Unit 5 was excavated to test the magnetometer data. The unit was 2 m x 2 m in size laid out over an apparent anomaly seen in the magnetometer map (Figs. 5.21, 6.1, 6.2). The unit was excavated to an overall depth of 52 cm. Being parallel on the north/south axis with Units 2 and 3, the plough zone was thinner here (in comparison to where Unit 4, 6 and 7 were located) and sand appeared at 39 cm (Fig. 6.16). Not many artifacts were found, but on the east end of the unit there appeared a darker stain running north/south in stratum 3. A probe, 1 m x 1 m, was dug along the east baulk to examine this staining (Fig. 6.16). The staining was a thin horizon and did not continue into the stratum; the probe went into sterile sand, so we stopped excavation at a depth of 74 cm.

The artifacts recovered from the first two strata were varied and continued to show a wide range of periods, but with a very strong presence of ancient German pottery. Charts 6.5–6.8 show the types of pottery and other artifacts recovered by amount and percentage. The German and Roman pottery recovered from stratum 2 most likely came from the hut itself, because a buried Grubenhaus has a convex shape, much like a loaf of bread. Ploughs over the years seem to have been sheering away the apex, digging into the remains of the hut, and working out artifacts. It cannot be proven, however, that the artifacts came from the hut and so the material is presented here as part of the overall percentages of stratum 2. Another reason for presenting the numbers here was that during excavations we penetrated into stratum 3, the hut itself, before realizing what we had uncovered and some artifacts that may have been in the hut, in sectors A2 and B2, became part of stratum 2. The movement of artifacts up and down in the soil, caused by plowing and other activities, was most clearly seen in the discovery of a coin, dating to the late nineteenth century. The coin was found next to the hut, at the bottom of stratum 2 and partially in the sand of stratum 4, which surrounds and underlies stratum 3.

Pottery sherds were found only in strata 1 and 2, and no features or artifacts were discovered that would explain the magnetometer anomaly. If there was something to clarify the anomaly in the area of Unit 5, it must be more deeply buried. The unit was interesting in that it did not yield much material other than pottery. The pottery itself was primarily LBK, with little being found of other periods. Stratum 1 yielded only pottery. Charts 6.11–6.13 provide the types of pottery and other material recovered, showing amounts and percentages.

6.3. Units 4, 4A–C

6.5. Unit 6

When work was largely complete with Grubenhaus G1 (Hut 1), a new unit was opened 50 cm north of STP 3-E15, located 10 meters east of STP 3-C-15 (Fig. 6.1, 6.2). We decided to open Unit 4, as a 3 m x 3 m trench, next to the STP, but upon inspection of the STP a few days

An examination of the map of STPs and the location of the two Grubenhäuser showed that the east area close to hut 2 (G4) was archaeologically unknown (Figs. 5.10 and 6.1); Hut 2 had even extended beyond the surveyed area. Three transects (F, G, and H) were added east of E 133

(designated Area 5), with the transects being the length of Areas of 3 and 4 combined where possible, since the fence, running at an angle, cut off the northern STPs of the three transects. As mentioned in Chapter 5.3, Area 5 was added in part to investigate circular areas of grass, but it was also added to obtain a general sense of what might be near Hut 2 and further south. The STPs yielded a variety of artifacts, but of those that could be identified, 18 were LBK sherds, making up 22% of the total. STP 5F-20 had the highest amount of LBK sherds, and the pieces appeared to come from one vessel, while another STP, 5-H-45, had the highest concentration of sherds from any period for a STP in Area 5. It was decided to put in small units at those two STPs to investigate these concentrations to see if there were any cultural features present.

6.6. Unit 7 The other STP investigated in Area 5 was 5-F-20 because of the LBK sherds discovered. The sherds were large pieces appearing to come from one vessel. Unit 7 was excavated not only to recover more of the vessel, but also to see if there was any indication of cultural features to explain the presence of the vessel (Figs. 6.1 and 6.2). The unit was excavated to a depth of 90 cm; there was no indication of any cultural features (Figs. 5.20). More pottery was recovered and two pieces in the south baulk required the excavation of an extension (82 cm x 44 cm) in the middle of the baulk (Fig. 6.2). The only items recovered, other than pottery, were two bones from stratum 1 and an odd lump of material that may be slag, also from stratum 1. Table 6.3 lists the types and amounts of pottery found. The pottery from STP 5-F-20 and the extension is incorporated into the totals of the appropriate strata.

The first STP investigated was 5-H-45, where Unit 6 was opened (Fig. 6.1). The STP had yielded 13 sherds, one of which was identified as Lengyel, but the other 12 could have also been of the same period. At the time, the sherds were thought to be of a possible shattered Bronze Age urn burial, since it was only after later analysis in the lab that the one sherd was identified as Lengyel. The STP was incorporated into the excavation of the unit. More sherds were found, but not from the same vessel and some of the sherds appeared to be LBK; there was no indication of a burial. An odd feature in the unit was a portion of a large pit (Fig. 6.17). Stratum 2 was the fill of this pit, which was made up of soil, rocks, and clay. The pit cut into stratum 3 and into the sand layer below (unexcavated stratum 4). Only the portion exposed in Unit 6 was examined, labeled as Stratum 2 in the profile drawing (Fig. 6.18), but during excavation was divided into Stratum 2A and 2B for ease of excavation and artifact control. Stratum 2A yielded two LBK sherds and three Ancient Unidentified, while 2B yielded one LBK and one Ancient Unidentified. Figure 6.19 provides a reconstruction of the probable size of the pit. Even though pottery did come from the fill, the pit itself cut through strata 3 and 4 and was immediately below a thin overburden of grass, and therefore would seem most likely modern. Because the pit was believed to be modern, Unit 6 was not expanded and the remainder of the pit was not excavated. Only one lithic item was found, all the rest of the finds being pottery. Table 6.2 shows what pottery was recovered in Unit 6 for strata 2 (the pit) and 3. The material found in STP 5-H-45 has been incorporated into the final tallies for Unit 6.

Table 6.3. Pottery Totals from Unit 7, Strata 1–3 Pottery Stratum Stratum Stratum 1 2 3 LBK 10 24 22 Anc. 3 1 0 German Modern 2 1 2 Ancient 9 0 0 Unidentified Total: 24 26 24 6.7. Unit 8 Toward the end of the field portion of the project, after backfilling had taken place, Dr. Turčan requested that the area between the two huts be excavated. The reason was that since the huts were so close, then it was possible that activities might be discerned between the huts, or other features might be revealed. Because it was close to the end of the field season, it was not possible to excavate the entire area and only a portion was marked off for investigation. Unit 8 (2 m x 5 m) was laid out in the northern portion of the area between the two huts, just south of the projected path of the modern pipe-cut (Figs. 6.1 and 6.2).5 The unit was excavated to an average depth of 51 cm, where we came upon the sand layer seen in Units 3 and 4 (Fig. 6.21). No cultural features were found, but as expected a wide assortment of pottery was recovered, along with other material. Table 6.4 lists the pottery types and artifacts recovered from stratum 1, while Charts 6.14–6.17 detail the amounts and percentages of the recovered artifacts from strata 2 and 3. Charts are not provided for stratum 1, since so little was recovered.

Table 6.2. Pottery Totals from Unit 6, Strata 2 and 3 Pottery Stratum 2 Stratum 3 LBK 3 10 Lengyel 0 1 Modern 0 2 Ancient 4 18 Unidentified Total: 7 31

134

Table 6.4. Artifact Totals from Stratum 1(Unit 8) Pottery Stratum 1 Artifact Stratum 1 Anc. 2 Pottery 14 German Modern 2 Lithics 1 Ancient 10 Bone/Teeth 1 Un. Modern 3 Brick

the majority (99%) comes from Units 2 and 3, in association with Hut 1. After pottery and bone/teeth, there is a rapid drop in totals for the other artifacts, with two next most commonly found artifacts, lithics, and metal at only 1.77% and 1.74% respectively. Wattle-anddaub numbered a total of 16 pieces and were found only in the proximity of the two huts. Overall then the following were the types of artifacts commonly found across the site in various units: pottery, metal, lithics, rocks, and slag. One aspect of the site that helped make it possible to excavate so much was the relative shallowness of the deposits. If the huts or even the general artifact smear on the landscape had been buried deeper, it is within reason to assume that STPs would not have found anything. The location on top of a buried sand dune or ridge and at its edge, whereby slope action and farming helped keep the plough zone (strata 1 and 2) thin, all helped toward the discovery of the site. What is more, since the plough zone was thin, the crew was able to do more archaeological exploration (i.e., excavate more units and of larger size) than if the plough zone had been much deeper.

6.8. Summary of the Findings The excavation portion of the project was successful in that two German huts were discovered. During the course of the project, however, it became clear based on the pottery that the occupation sequence of the site was more complex than first envisioned. Other units, primarily 6 and 7, were excavated to help elucidate the sequence, but they failed to provide any type of cultural features. The goal of the research, to find German occupation, was fulfilled, but new questions have arisen, especially in regards to the extent of the Neolithic site and why there was abandonment after the Eneolithic and then apparent reoccupation in the Roman Period. Pottery was the largest group of artifacts found at Urbárske Sedliská; this circumstance is not surprising, since pottery is usually the predominant artifact on most sites. Pottery made up 76.45% of the total artifacts recovered from the units and 45% of those came from the Roman Period (including both German and Roman pottery); the second largest group was the pottery that could be identified only as Ancient Unidentified (Chart 6.18). The third and fourth largest groups were Modern (5.7%) and LBK (5.2%) pottery respectively. Even though LBK made up only 5.2% of the total, their presence dominated units outside of the immediate vicinity of the huts themselves. This fact becomes even more obvious when we include the results from the survey, where LBK made up of 14.1% of the total in comparison to only 7.5% for Roman-Period pottery (Chart 5.5). It would seem that the Roman-Period component is a more isolated event at Urbárske Sedliská than the LBK component which appears to cover a much wider area of the site. This phenomenon can be explained in two ways: 1) the LBK period had a much larger population density at Urbárske Sedliská than the later periods, or 2) since the LBK artifacts have been at the site for 8,000 years as opposed to 2,000 for the Roman Period, the LBK artifacts have had much more time to be smeared across the landscape by natural process and human action. I am inclined to believe in the second proposition at the moment until further exploration can be done at the site. A wide range of artifacts was found mirroring that of the survey phase. Chart 6.19 details the types of artifacts found and their respective amounts and percentage of the total. After pottery, bone/teeth was the second most common item found, but this ranking is misleading since 135

Endnotes 1. I express my sincerest thank you to the crew who worked very hard not only in digging all of the shovel test pits, but then too in excavating all the units and backfilling them. The crew was P. Nagy, D. Nagyová, V. Zervan, M. Fratrič, and, from the survey portion, A. Mongiello. 2. The pipe cut was only partially excavated in order to save time for the completion of the entire excavation. I now regret making that decision, since the fill of the trench might have yielded pottery from the hut itself which may have helped reconstructing some of the vessels. 3. Dr. Turčan periodically visited the site to see how work was progressing and to ask if any assistance was needed. His visits were good opportunities to ask his opinion on various matters and his help was always greatly appreciated. 4. Because I had not made it clear to the excavators that they were to recover material from the first stratum as we had done in the other units, stratum 1 material was not collected. Once I realized this mistake had happened, material was saved for the rest of stratum 1, but too late for a large part of the unit. The amount saved is too small to be treated separately and so has been combined with material from stratum 2. 5. Later, once in the lab and having started to do research on German huts and orientation, it became clear that Unit 8 should have been placed in the southern portion of the shared area, since the huts’ entrance would have been on the south side. It is safe to assume that cultural activity outside the hut probably occurred near the entrance so other features (fire-pits, storage-pits, etc.) would be on the south side.

136

Charts Chart 6.1. Pottery by Period/Culture from Unit 2, Stratum 1, Showing Amount and Percentage

Modern, 32, 5.9% Middle Ages, 1, 0.2%

Roman, 5, 0.9%

Ancient German, 169, 31.1%

Ancient Unidentified, 331, 60.8%

Bajč-Retz, 2, 0.4% Lengyel, 2, 0.4% LBK, 2, 0.4%

Chart 6.2. Artifacts from Unit 2, Stratum 1 by, Showing Amount and Percentage

Slag, 1, 0.2% Wattle-and-Daub, 4, Lithics, 12, 2.1% 0.7% Metal, 10, 1.8%

Pottery, 544, 95.3%

137

Chart 6.3. Pottery by Period/Culture from Unit 2, Stratum 2, Showing Amount and Percentage

Modern, 11, 7% Roman, 3, 2% Ancient Unidentified, 54, 34%

Lengyel, 1, 1% Bajč-Retz, 2, 1%

Ancient German, 90, 55%

Chart 6.4. Artifacts from Unit 2, Stratum 2, Showing Amount and Percentage

Other, 1, 0.5% Bone/Teeth, 28, 13.9% Slag, 3, 1.5% Lithics, 2, 1.0% Metal, 6, 3.0%

Pottery, 161, 80.1%

138

Cart 6.5. Pottery by Period/Culture from Unit 3, Stratum 1, Showing Amount and Percentage

Modern, 49, 10.3% Middle Ages, 2, 0.4% Roman, 2, 0.4%

Ancient German, 103, 21.7%

La Tène, 3, 0.6% Ancient Unidentified, 309, 65.1%

Bronze Age, 3, 0.6% Bajč-Retz, 1, 0.2% Lengyel, 2, 0.4% LBK, 1, 0.2%

Chart 6.6. Artifacts from Unit 3, Stratum 1, Showing Amount and Percentage

Rocks, 18, 3.2%

Modern Brick, 1, 0.2% Wattle-and-Daub, 3, 0.5%

Spindle whorl, 1, 0.2% Bone/Teeth, 46, 8.2% Lithics, 6, 1.1% Metal, 10, 1.8%

Pottery, 475, 84.8%

139

Chart 6.7. Pottery by Period/Culture from Unit 3, Stratum 2, Showing Amount and Percentage

Roman, 4, 1%

Ancient Unidentified, 161, 39%

Ancient German, 241, 59%

Lengyel, 5, 1%

Chart 6.8. Artifacts from Unit 3, Stratum 2, Showing Amount and Percentage

Rocks, 1, 0.1%

Wattle-and-Daub, 9, 1.2% Coins, 2, 0.3%

Spindle whorl, 1, 0.1%

Bone/Teeth, 293, 39.9%

Pottery, 411, 55.9%

Slag, 2, 0.3%

Lithics, 8, 1.1%

Metal, 8, 1.1%

140

Chart 6.9. Pottery by Period/Culture from Unit 4, Strata 1 and 2, Showing Amount and Percentage Modern, 3, 1%

Ancient Unidentified, 7, 2%

Roman, 9, 2% LBK, 5, 1%

Ancient German, 365, 94%

Chart 6.10. Artifacts from Unit 4, Strata 1 and 2, Showing Amount and Percentage

Rocks, 1, 0.2%

Spindle whorls, 2, 0.4%

Bone/Teeth, 143, 25.3%

Slag, 2, 0.4% Lithics, 11, 1.9% Metal, 17, 3.0% Pottery, 389, 68.8%

141

Chart 6.11. Pottery by Period/Culture from Unit 5, Stratum 1, Showing Amount and Percentage

Modern, 2, 12% Bajč-Retz 6%

Ancient Unidentified 58%

LBK, 4, 24%

Chart 6.12. Pottery by Period/Culture from Unit 5, Stratum 2, Showing Amount and Percentage

Ancient Unidentified, 7, 18% Modern, 11, 28%

Ancient German, 1, 3% Bajč-Retz, 1, 3%

LBK, 19, 48%

142

Chart 6.13. Artifacts from Unit 5, Stratum 2, Showing Amount and Percentage

Rocks, 2, 4%

Charcoal, 1, 2%

Bone/Teeth, 1, 2% Slag, 1, 2% Lithics, 5, 10%

Pottery, 39, 80%

Chart 6.14. Pottery by Period/Culture from Unit 8, Stratum 2, Showing Amount and Percentage

Modern, 8, 15%

Ancient Unidentified, 21, 38%

Roman, 4, 7%

Ancient German, 12, 22%

Lengyel, 1, 2%

LBK, 9, 16%

143

Chart 6.15. Pottery by Period/Culture from Unit 8, Stratum 3, Showing Amount and Percentage

Modern, 2, 7%

Ancient Unidentified, 12, 43%

Ancient German, 9, 32%

LBK, 5, 18%

Chart 6.16. Artifacts from Unit 8, Stratum 2, Showing Amount and Percentage Rocks, 2, 3%

Modern Brick, 1, 1%

Bone/Teeth, 5, 7% Slag, 2, 3% Lithics, 4, 6% Metal, 1, 1%

Pottery, 55, 79%

144

Chart 6.17. Artifacts from Unit 8, Stratum 3, Showing Amount and Percentage

Modern Brick, 3, 8% Rocks, 1, 3% Bone/Teeth, 2, 5%

Slag, 3, 8%

Lithics, 3, 7%

Pottery, 28, 69%

Chart 6.18. Pottery by Period/Culture from all Units, Showing Total Amount and Percentage Modern, 127, 5.7% Middle Ages, 3, 0.1% Roman, 23, 1.0%

Ancient Unidentified, 956, 43%

Ancient German, 996, 44%

LBK, 116, 5.2% La Tene, 3, 0.1% Lengyel, 12, 0.5% Bronze Age, 3, 0.1%

Bajč-Retz, 7, 0.3%

145

Chart 6.19. Artifacts from all Units, Showing Total Amounts and Percentage Wattle-and-Daub, 16, 0.5% Charcoal, 1, 0.03%

Modern Brick, 8, 0.27%

Coins, 1, 0.03%

Rocks, 25, 0.85% Spindle whorls, 4, 0.1%

Other, 1, 0.03%

Bone/Teeth, 519, 17.67% Slag, 14, 0.48%

Lithics, 52, 1.77%

Metal, 51, 1.74%

Pottery, 2246, 76.45%

146

Figures

Figure 6.1. Map of the east side of the survey area and position of units. Small black squares on the transects represent shovel test pits.

147

Figure 6.2. Close-up map showing position of Units 3–5, 7 and 8. The small black dots represent the actual location of the meter point on the transects, while the small squares show the position of shovel test pits next to the meter points. The dotted lines in Units 2 and 5 show where the probes were excavated.

Figure 6.3. Drawing of Unit 2 showing the sectors.

148

Figure 6.4. Photograph of plough furrows.

Figure 6.5. Photograph of dark stains.

149

Figure 6.6. Photograph of dark stains with Grubenhaus indicated.

Figure 6.7. Photograph of north baulk after further excavation, revealing more of the hut and posthole.

150

Figure 6.8. Profile drawing of north baulk of Unit 2. Strata 1 and 2 are a fine sandy loam, while stratum 3 is sand.

Figure 6.9. Map showing all the units making up Unit 3 and their division into sectors.

Figure 6.10. Profile of Unit 3, south baulk (later removed as Unit 3E). Strata 1 and 2 are a 10YR 2/2 (very dark brown) fine sandy loam, stratum 3 is a 10YR 3/1 (very dark gray) fine wet loam and stratum 4 is a 10YR5/4 (yellowish brown) sand.

151

Figure 6.11. Image showing the dark stain of the hut, with the pipe cut indicated.

Figure 6.12. Posthole (white arrow pointing to it) in STP 3-E-15.

152

Figure 6.13. Map showing sectors of Units 4, 4A, 4B, and 4C.

Figure 6.14. Entire hut revealed as a stain before its excavation.

153

Figure 6.15. Reconstructed profile of south baulk from Unit 4, showing the southeast corner of Hut 2. The profile is based on the stratigraphy of Units 4 and 4A–C, along with measurements taken during excavation. Strata 1 and 2 are a 10YR 3/1 (very dark gray) fine sandy loam, strata 3 and 4 are a 10YR 2/2 (very dark brown) fine wet loam and stratum 5 is a 10YR 4/2 (dark grayish brown) sand, mottled with a 10YR 5/4 (yellowish brown) sand.

Figure 6.16. Baulk drawing for Unit 5: Strata 1 and 2 are a fine sandy loam, while stratum 3 is fine loamy sand.

Figure 6.17. Photo showing the excavated pit in Unit 6. 154

Figure 6.18. Profile drawing of the south baulk of Unit 6. Stratum 2 represents the pit. Strata 1and 3 are a fine sandy loam, while stratum 2 is also a fine sandy loam, but choked with clay and rocks.

Figure 6.19. Reconstruction of probable size of the pit.

155

Figure 6.20. Profile of Unit 7. Strata 1–3 are fine sandy loam, while stratum 4 is sand.

Figure 6.21. Profile drawing of south baulk for Unit 8. The south baulk is an odd length due to the presence of Unit 4C at the east end of Unit 8, see Figure 6.1. Strata 1 and 2 are fine sandy loam and stratum 3 is sand.

156

and Turčan 1990, 73). This type and its sub-types are in Bohemia and Slovakia by the first century AD and in southern Moravia by the second and third centuries; it is primarily seen among the Quadi and is considered an example of Quadi architecture (Kolník 1971, 509; 1998b, 69; Droberjar 1997, 22; 2001, 100; Varsík 1998, 4). The sub-types are given below (Kolník 1998a, 146–150; 1998b, 69).

Chapter 7 Analysis of Huts 1–2 and Features G2–G3 Chapter 7 covers four topics: 1) the Grubenhaus as a building type; 2) an analytical description of Huts 1 and 2; 3) a description of features G2 and G3, along with a discussion of how they function with Hut 1, and their impact on the hut’s south side; and 4) the dates of Huts 1 and 2.

Type IV: Pit-houses with posts in the corners, usually having a rectangular plan (Kolník 1998a, 145). This type primarily dates to the Late Roman Period and Great Migration Period, and even into the Middle Ages (Kolník 1962, 387; 1971, 509; Droberjar 1997, 20) Type V: A pit-house that is made up of two or more of the previous types joined together to make one larger house (Kolník 1998a: 145).

7.1. The Grubenhaus Germans are known to have built primarily two types of structures: the Grubenhaus and the long-house (Figs. 7.1 and 7.2).1 The long-house is an above-ground structure, rectangular in shape, using posts, and reaching lengths of up to 27 m or more. Excavations have recovered longhouses in Germany, at sites such as Ezinge and Feddersen Wierde as well as in Holland and southern Denmark (Todd 2004, 64).2 These are usually clustered into settlements, with Grubenhäuser sometimes present and identified as housing for dependants of those in the longhouses (Todd 2004, 64, 66, 67 and 70). In Slovakia (along with eastern Austria and Moravia in the Czech Republic), the Grubenhaus is the more common type and is defined as a pit-house (sunken-floored dwelling) made of a wooden frame, with walls either of wattle and daub or wood and with a roof made of thatch (Fig. 7.1) (Kolník 1971, 509). The Germans built the houses by excavating an area to a depth of 20–50 cm, creating a sunken livingfloor, usually 12 m2, but ranging from 9 m2 to 20 m2 (Kolník 1971, 509; 1998a, 145; 1998b, 69; Droberjar 1997, 25; 2001, 100). Five types of Grubenhäuser have been identified in Slovakia by Titus Kolník and have become the standard way of identifying huts by archaeologists in the country (Fig. 7.3):3 The types are as follows.

It should be emphasized that based on overwhelming evidence, the usual shape for an ancient German hut was rectangular or a rectangle with rounded corners making an oval form (Kolník 1962, 388; Kolník 1998a, 150).4 Round residential huts are not known, even though they are seen in Roman art, such as on the column of Marcus Aurelius. These are now seen by scholars as merely schematic representations to give an idea of what a German dwelling looked like and the artists were confused with structures built over supply pits (Kolník 1998a, 150). The post-holes found in huts vary in size, from 20 to 70 cm in diameter and go to depths from 15 to 70 cm below the floor of the huts (Kolník 1998a, 149). The post-holes are cylindrical with ends either tapering to a tip or flat. Sometimes, the post-holes are outside of the hut walls, but usually are found on the inside. The posts acted as support for not only the roof but also the wattle, or interlacing twigs and branches, serving as the core for the walls that would be plastered with mud, which survives in the huts as lighter colored staining in the fill or, if the hut was burned, as baked chunks (Kolník 1998a, 150). After pottery sherds, chunks of fired daub are the most commonly found item on ancient German archaeological sites (Droberjar 2001, 172).

Type I: Pit-houses without posts that are dug deeper than the other types (Kolník 1998a, 145).

The floors are usually stamped earth or clay, with evidence of repairs (Kolník 1998a, 149). The living floor sometimes was a suspended plank floor at least in part of the hut, as evidenced by smaller post-holes stains (Kolník 1998a, 149; Macphail and Linderholm 2002, 2). The presence of such a floor is also strengthened by the fact that some huts do not have evidence of a stamped earth floor and usually there is little evidence found in huts of activities that took place in them, seemingly indicating a living-floor that did not survive. There is evidence of a central post sometimes added during the life of the hut to help prop up the ridge pole holding up the roof (Droberjar 1997, 25; 2001, 100; Varsik 1998, 4). Some Grubenhäuser have several small depressions on the floor, which are either supports for a plank floor as mentioned, or depressions from where the wooden feet of

Type II: Pit-houses that have a rectangular or oval plan, with two main posts on the center axis, located opposite each other along the short walls (Kolník 1998a, 145). This type was first seen during the late La Tène period and continues into the Early Roman Period, and if found archaeologically indicates either a late Celtic or early German structure (Kolník 1962, 387; 1998a, 146; Droberjar 1997, 20; 2001, 98). Type III: A type that has a rectangular plan with rounded corners and has post-holes spread out in the shape of a hexagon, with only one post-hole along a short wall and two along the long walls (this is the most common type of Grubenhaus in Slovakia) (Kolník 1998a, 145; Bartík 157

furniture had rested, such as tables, benches, or beds (Kolník 1998a, 149; Droberjar 2001, 100).

also have a pit at the center of the hut, different from what would have been the added central post-hole or pit associated with the entrance (Fig. 7.6). These pits have vertical walls, with a flat bottom (Kolník 1998a, 149). Kolník points out that these may be the place for a stove, but there is no solid supporting evidence and he actually believes that the pits may have functioned as storage places (Kolník 1998a, 149). In fact, there is little evidence to show that stoves or fireplaces were located inside the huts; they would have been probably outside the dwelling, while braziers may have been used in the huts for heat (Kolník 1998a, 149–150).

When the huts were built, a large amount of evidence (such as the presence of an entrance niche) shows that the entrance was usually on the south side, varying 45° to either side of south (Droberjar 1997, 23; Kolnik 1998a, 154–155). The reason for this orientation may have been to provide light for the interior. If the huts had no windows and interior lighting during the day depended solely on the door, then it would be beneficial to point south. At the latitude in Europe where Slovakia is located, the sun does not come up very high during winter, riding low in the southern sky. At the summer solstice, the sun is at 63° 54’ in the southern sky (above the horizon, which is 0° 0’), while at the winter solstice, the sun is only at 17° 47’ in the sky.5 By having the door face south, the inhabitants would maximize the amount of light that entered the hut. Kolnik postulates that the positioning of the entrance was also caused by wind direction, especially during the winter months (Kolnik 1998a, 155).

The average lifetime of a hut is thought to be around twenty years (Droberjar 1997, 26; 2001, 100).6 It is not surprising that many huts exhibit evidence of burning, if we consider the use of so much wood and loam. The burning of huts should not be immediately associated with the Marcomannic Wars, since various activities in huts (smelting, cooking, etc.) could have also caused fires, along with intentional burning by the occupants or nefarious activities between locals (Droberjar 1997, 26). Dr. Turčan has mentioned a theory that when a hut reached its end, either because of structural weaknesses or vermin infestation, the inhabitants would strip it of whatever wood was useful, and then the remainder was torn down on the spot and set on fire. The fire served not only to consume what was left of the hut, but also to destroy the vermin. The burned-out hut, with its depression, became a refuse dump for the inhabitants who lived on in a new hut nearby and for the community as a whole. The fact that the living floors often do not survive and old huts possibly served as trash dumps for other huts makes it highly unlikely that anything found in a hut is from primary deposition. It is safe to assume that most items found in the fill of a Grubenhaus are from secondary or even tertiary deposition. This idea has been validated by recent studies of Grubenhäuser fills in England and by micromorpholgy (Macphail and Linderholm 2004). The Grubenhaus still provides valuable information on how these huts were built and also on activities in the community as a whole, even if we cannot know how the specific family that built the hut lived during its lifetime.

Because of the variability of design for Type III huts, Kolník further subdivides the type into four subcategories labeled III/1, III/2, III/3, and III/4 (Kolník 1998a, 146– 149)(Fig. 7.4). III/1) The standard design with six postholes and no entrance niche (as describe above). III/2) The design has the standard six postholes along with an entrance niche on the south side. Droberjar sees these huts (especially those with the entrance niche) as uniquely Quadi in design (Droberjar 1997, 22). III/3) The hut has seven or more postholes without an entrance niche. III/4) The final design has seven or more postholes with an entrance niche. In regards to the Type III huts without the entrance niche (types III/1 and III/3, but also sometimes with the other two sub-types as well), usually there is a circular or oval depression along the inside of the south wall at the middle (Droberjar 1997, 25; 2001, 100) (Fig. 7.5). The reason for the depression has something to do with the entrance, but has not yet been determined. Since they are directly in the path of the entrance, were they left open, filled in, or covered in some way? Their fill usually corresponds to what was found in the rest of the hut or is similar to the fill of the post-holes. One theory has the depression acting as a rainwater-catch to keep the rest of the hut dry. This seems unnecessary when a shallow ditch outside, around the hut could have been dug to lead rainwater away from the hut.

In the description given so far, the huts have been viewed as dwelling places, but in fact the huts had more varied uses. In Northern Germany, where the long-house dominates from the first century BC to the second century AD, the Grubenhaus is seen as an ancillary structure for extended family or servants or where work took place, such as metal production, pottery making, weaving, etc. (Todd 2004, 69–70). In southwest Slovakia as well, the Grubenhaus could also function as a workplace. Huts have been found with kilns and furnaces in them, indicating production activity, along with the requisite evidence of slag and other refuse. Such huts frequently were located at the edge of settlements (Kolník 1998a, 152), both to lessen fire dangers and to keep smells and fumes away from habitation areas.

If the entrance niche is present (Type III/2 and III/4), its floor is usually slightly higher than the floor of the hut, and there is evidence that the niches at some huts were paved with river pebbles (Kolník 1998a, 149). Some huts 158

to west posthole, was 5.153 m, while the average width of the hut was 4 m, thereby creating a space of 20.6 m2. Hut 1 is thus at the large end of hut sizes as given by Kolník above. The hut had been excavated to depth of 15 cm into

7.2. Huts 1 and 2 (Features G1 and G4) Both huts can be classified as a type III/3 hut, with no evidence of a south-side entrance niche (Fig. 7.7). The huts probably had their entrance on the south side, based on archaeological experience, but only Hut 1 had clear evidence to support this. Hut 1 had a pit (feature G2) on the inside of the south wall at the middle, where one would expect such a feature playing a role for the entrance (Fig. 7.7). The interesting aspect of feature G2 was its possible connection to a shallow ditch leading away from the hut to the southwest, labeled as feature G3 (Fig. 7.7). Both huts were probably dwelling places since there is not enough evidence to show any type of industrial activity.

Table 7.1. Distances Between Postholes of Hut 1 (in meters) Postholes Distance East to West 5.153 NE to SE 3.900 NW to SW 4.100 NW to NE 3.003 SW to SE 3.101 NW to SE 4.980 NE to SW 5.020

Huts 1 and 2 had the standard arrangement of six postholes for a Type III hut as given above (Fig. 7.7). Both huts also exhibited the addition of a middle post to help support the ridge-pole supporting the roof. In the fill of the huts, there were fired pieces of wattle-and-daub providing evidence of building technique. In the floor of the huts were various stains, with Hut 1 having a possible storage pit next to the central posthole and Hut 2 exhibiting small pits filled with dark soil that may be evidence of furniture legs or supports for a wooden plank floor.

the sand (stratum 4), but if we include the ancient top soil that would have been overlying the sand, the true depth would have been greater. The depth of the fill above the sand was around 5 cm, so it is possible that the true depth of the excavated hut was 20 cm, but since there are questions of potential plough damage, it cannot be said for sure what that depth might have been (Fig. 7.8). The postholes varied in size, both in diameter and depth. Along with shape/size variability, the profiles revealed two postholes having non-cylindrical forms. The SW posthole (Fig. 7.10) was the best preserved and had two tapered ends, indicating that there may have been two smaller posts tied together instead of one larger posthole (or one post with a broken end). The SW posthole also revealed the width of the posthole and the actual posts by the varied coloring in the sand. The posthole has a diameter of 32 cm, but it appears that the post or posts only had a diameter of 26 cm. The shallower tip ends at a depth of 41 cm, while the other goes on to a depth of 64 cm. The W posthole also had some evidence of two posts together, but it was not as clearly seen as in the SW posthole (Fig. 7.10). The W posthole is cylindrical, having a width of 24 cm, but there are two tips, one at a depth of 36 cm, and the other at 50 cm. Table 7.2 provides the diameter and depth of each posthole belonging to Hut 1. In general, the postholes fit within the parameters set forth by Kolník above for posthole sizes. Four postholes had artifacts in them. The NE, Center, and E postholes had each a German sherd, while the SW posthole had a fragment of bone.

The fill of the huts was a soil rich in organic material, especially charcoal flecking, pieces of burned wood, and bones, along with pottery and other cultural material. The soil at times had a greasy feel, probably because of the large amount of bones and whatever fleshy animal parts had decayed in the fill. It was also noticed that the soil held moisture longer than the surrounding soil or sand. 7.2.1. Hut 1 (Feature G1) This hut is not as well preserved as Hut 2. Hut 1’s preservation suffered from three factors (Fig. 7.8): 1) the digging of a ditch through the northwest corner in 1982, 2) the disappearance of the upper portions of the fill as a result of farming activity, and 3) the disruption of the south side of the hut by feature G3 (and possibly G2). All seven postholes did survive, with only the west posthole cut in half by the modern ditch. The hut fill (which is stratum 3) provided a wide assortment of artifacts, but probably would have been greater if not for the disruptions. As mentioned in Chapter 6, when the fill of the hut was first discovered, the unit was expanded until the entire hut was revealed. The shape of the hut could be clearly seen as a rectangle, oriented at a 35o angle from the north/south axis. The entrance opened toward the southwest.

The indication of average diameter or depth is a reflection of the posthole shapes. The cylindrical standard from, with no tapering and a flat bottom, is seen in the NE posthole (Fig. 7.10). Postholes E, NW, and SE taper more or less, either coming to a point or flat bottom (Fig. 7.10). The E posthole at the surface is 38 cm in diameter, but then tapers to 14 cm at the bottom. The NW posthole starts at 24 cm, but tapers to 10 cm at the bottom. The SE is cylindrical for the first 14 cm, but then tapers on one side so that the bottom of the posthole is 18 cm in diameter. The central posthole has a very wide upper portion (50 cm diameter), only about 4 cm in depth,

The dimensions of the hut are given below in Table 7.1, with the postholes labeled by their position according to the cardinal directions, i.e. east posthole, west posthole, northeast posthole, and so forth (see Fig. 7.9 for the layout). The length of the central axis, from east posthole 159

which then drastically decreases in diameter to 28 cm, then tapers to a 2-cm end (Fig. 7.10).

survives because the structure was burned creating ceramic-like chunks. This event is referred to as sintering, and the product is identified as “sintered daub” (Shaffer 1993, 60). Experiments have shown that fires of short duration or of low intensity do not create a large amount of sintered daub, but only 1–3% of the total amount available (Schaffer 1993, 60–61).7 In regards to Hut 1, a total of 49 chunks of sintered daub were found in the fill and if we include the pieces found in Unit 2 and stratum 2 of Unit 3, then a total of 69 was recovered. The number may seem large, but many pieces were small, some weighing only one gram (the total weight of the 49 pieces from the hut fill was only 119 grams). This amount is small in comparison to the total amount of daub that would have been present in the hut, thereby supporting the hypothesis of a low intensity fire or one of short duration.8

Table 7.2. Diameter and Depth of Hut 1 Postholes (in centimeters) Posthole Diameter Depth 22 36 NW 27 47 NE 24 Avg. is 40 W Avg. is 20 51 E 32 Avg. is 48 SW 28 36 SE 45 Center Avg. is 22 In regards to staining, there was no evidence in the fill itself, which was uniformly black in color (Fig. 7.11). Underneath, the bottom of the hut cuts into the surface of the buried sand dune. The sand had various stains, ranging in color from deep orange to white, along with the dark grey staining of the postholes, and next to the center post, on the east side, a rectangular gray/black stain (ca. 70 cm x 90 cm in size) (Fig. 7.12). This stain was mentioned above as possible evidence of a storage pit, but this is probably not the case since it had no real depth (only going approximately 3 cm into the sand). If the comments by Droberjar and Kolník are correct that small circular stains could be the evidence of furniture feet, then the rectangular stain could be evidence of a larger item, such as a chest, crate, or some other item that could account for the shape. There were no other dark stains present in the hut that would accord with furniture feet or plank supports. On the south side of the hut, at the middle of the long wall, on the inside, was the odd white sand pit, feature G2 already mentioned above (Figs. 7.8 and 7.13). Feature G2 is still something of a mystery: possibilities range from a feature having a function related to the entrance to just being a rodent hole. The deep orange-colored sand stains on the hut’s floor may have been created when the hut was burned, as we know it was from the abundant charcoal flecking in the fill, the charcoal pieces, and sintered daub, as discussed below.

The daub is of interest because some pieces preserve the impression of the wood lattice that had been present in the hut walls. A chunk from stratum 2 of Unit 3 (Cat. no. 437) has preserved the impression of two branches that ran parallel to each other (Fig. 7.15). Another piece (Cat. no. 433), this time from the hut fill, has preserved the impression of four branches, two on either side (Figs. 7.15 and 7.16). Upon close inspection, one can also see the impression of the wood grains or bark as well. It would appear that the smaller pieces of wood were probably unworked branches, interlaced and connected to the posts. Small branches were preferred for the lattice since they could be easily bent and threaded, compared to larger, stiffer branches. The fill also yielded a wide assortment of bone, pottery, and other artifacts (see Charts 7.1 and 7.2 and Table 7.3). A total of 514 pieces of pottery was recovered, of which 95% was German, 4% Roman, and 1% La Tène. A total of 896 artifacts were recovered other than pottery. There was a total of 759 animal bones (including shells of mussels) and 71 teeth/fragments recovered, coming from cows, pigs, sheep/goats, horse, domestic bird, wild ox, deer and freshwater mussels, with cow and pig dominating the assemblage (a detailed description is given in Chapter 10). Three bronze items were also found along with four stones and three lithics. Seven pieces of slag were also recovered.

The major stain or feature that cut through the hut was the 1982 ditch, which had a width of 80 cm (Figs. 6.11 and 7.8). Only a small portion was excavated as mentioned in Chapter 6.2 (see also Chapter 6, footnote 2) (Fig. 7.14). At a depth of 56 cm, the bottom was reached, where there was a plastic pipe wrapped in a heavy canvas cloth. The pipe-cut went through the entire unit, at an angle in the northern half, from the west baulk to the east baulk, and was dug with crisp edges. The magnetometer survey shows the length of the ditch (see Figs. 5.20 and 5.21). The ditch cut off the northwest posthole from the rest of the hut.

During the excavation of Hut 1, various samples were also taken for an assortment of other archaeological analyses.9 The fill had a large amount of charcoal, but 99% was just flecking or minute pieces mixed in the fill: two pieces were of significant size to be collected for radiocarbon dating and one of these was submitted for testing.

The fill of the hut had pieces of baked daub from the walls. Daub is usually a clayey soil applied as a thick plaster to the wooden lattice making up the wall. Because of the presence of clay, as the walls dry out and are exposed to the sun, the daub hardens and can survive in the archaeological record, but more often daub 160

Table: 7.3. Pottery from Grubenhaus 1 (Feature G1) Pottery Count Weight (in grams) 493 6434 German 18 126 Roman 3 50 Other 514 6610 Total:

corroded metal (iron), three lithics, one slag, 18 bone fragments, and five teeth (Table 7.5). During the excavation of G2, a field mouse nest was found in the wall on the south side of the feature, around the middle (Fig. 7.18). The nest was small, only 15 cm across and eight cm high, where six baby mice rested: their eyes had not yet opened and they had no fur.11 It was noticed that at the bottom of G2, in the east wall of the feature, at the northeast corner was a modern tunnel that the mother had been using. During excavation, the tunnel had apparently collapsed (the walls being sand) and escaped our notice until the mouse and her nest were discovered. There is no indication that Feature G2 was connected with the field mouse found in the unit; her tunnel bisected G2 only at the east end, and since the entire pit was filled with sand and artifacts, which were all ancient, it does not seem likely that the field mouse (or generations of mice) had dug the feature. It would appear then that G2 is not a modern rodent hole. As to whether G2 is an ancient rodent-hole or even rabbit den is possible, but its orientation toward the southwest and proximity to feature G3 make it seem less likely.

A rough flotation system was devised and a sample of the hut fill was sieved and the material collected for examination by a paleoethnobotanist. It is hoped that some pollen or phytoliths are present to determine diet and define the ancient landscape. A small amount of fill was also taken for phosphate analysis. Three samples of fill were also taken for geoarchaeological analysis, such as micromorphology, in order to better understand the fill itself.10 7.2.2. Feature G2 As mentioned above, a pit was uncovered in the hut at the middle of the south wall, in the area where the entrance would be expected (Fig. 7.8). The fill of the pit, which was sand of a different color (10YR 6/2, light grayish brown) and included artifacts, was designated Feature G2 to differentiate it from the hut itself (stratum 3 of Unit 3). The pit is oblong, curving toward the southwest, and is dug into stratum 4 (Figs. 7.8, 7.12 and 7.13). The width is around 30 cm and is 185 cm long. The pit is deepest (48 cm) in the northeast and steadily decreases to the southwest, while maintaining a constant width. In the southwest, the pit slopes upward, then there is a shallow shelf, which then slopes up to the level of the hut’s bottom. This southwest end is at the edge of the south side of the hut itself, and appears to have extended beyond the hut by 10 cm. The middle of the pit is around 25 cm deep, while the southwest shelf is 10 cm below the bottom of the hut. The northeast part (the deep end) exhibits an interesting microstratigraphy (Fig. 7.17). The east wall of the pit (and seen in the feature itself at the east end during excavation) shows multiple micro-layers of alternating brown and white sand. These layers are only 1–2 mm in depth and were too numerous to count. These layers indicate that there was an opening (most likely at the southwest end of the feature) to the surface and most likely water would come in depositing a microlayer of brown topsoil in the deepest part of the feature, then sand would overlay this until water brought in a new layer of topsoil. This deposition did not continue until G2 was filled with these micro-layers because at a depth of 24 cm the micro-layering stopped and the rest of the fill was white sand, a circumstance possibly indicating that the opening was plugged at some point.

Table 7.4. Pottery Types and Amounts from Feature G2 Pottery LBK Lengyel German Roman Anc. Undet. Total:

Amount 2 2 94 1 19 118

Table 7.5. Artifacts from Feature G2 Artifact Amount Pottery 118 Metal 6 Lithics 3 Slag 1 Bone/Teeth 23 151 Total: 7.2.3. Feature G3 During the excavation of the southwest area of Unit 3 (in the south end of Unit 3D), a darker stained feature could be seen just outside of Hut 1 in the sand and continued into the south baulk, but did not appear in Unit 2. This stain became especially pronounced when the baulk between Units 2 and 3 (labeled Unit 3E) was removed (Fig. 7.20). Another pit could be seen, labeled feature G3, extending from stratum 2 into stratum 4, which is the sand. The portion that was visible was excavated, but it appeared to continue into the west wall of the Unit (Fig. 7.19, A). After Hut 2 was excavated, we returned to feature G3 and created an extension of Unit 3 to investigate the feature. The second investigation was labeled as the G3 Extension, but both parts will be taken

Feature G2 contained a great quantity of pottery, considering its size. There was a total of 118 pieces of pottery, of which 94 could be identified as ancient German, one Roman piece, and the rest as belonging to earlier cultures (Table 7.4). There were also six pieces of

161

as a whole in this text as merely Feature G3. There are, however, two parts to the feature that roughly correspond to the two names. In Unit 3 itself, there is the pit (the original feature labeled G3) that then at its western end connects to a shallow ditch heading toward the southwest. So, in effect, the G3 Extension actually corresponds to the shallow ditch (Fig. 7.20). In order to avoid confusion, the two elements combined will be referred to as Feature G3, but in describing the feature, the pit and ditch will be discussed separately.

fragments, three rocks, two lithics and one badly worn fragment of what appeared to be modern brick (Table 7.7). The pottery could be subdivided into one LBK, 24 German, two Roman, two Modern and 23 Ancient Unidentified (Table 7.6). Table 7.6. Pottery Types and Totals from Feature G3 Pottery Pit Ditch Ditch Stratum 2 Stratum 3 LBK 2 2 1 Lengyel 0 2 0 German 19 30 24 Roman 0 3 2 Modern 0 6 2 Anc. Unident. 4 20 23 25 63 52 Total:

As mentioned, the pit was discovered first. The pit is 1.37 m long with a width that varies (Fig. 7.19, B). The north side of the pit is straight and is aligned with the north side of the shallow ditch that continues toward the southwest. The south side of the pit is oddly shaped. The eastern edge of the feature is 35 cm from the western edge of Feature G2, which is in alignment with Feature G3. At its widest, Feature G3 is 60 cm in the eastern end, but to the southwest as it approaches the shallow ditch, it aligns itself to equal the width of the ditch at 75 cm in the south baulk of Unit 3. The pit also varies in depth. The east end is shallow at 6 cm, but drops to 17 cm at the center and then rises again in the west to equal the depth of the shallow ditch at 10 cm.

Table 7.7. Artifact Types and Amounts from Feature G3 Artifact Pit Ditch Ditch Stratum 2 Stratum 3 Pottery 25 63 52 Metal 0 0 2 Daub 0 1 3 Bone/Teeth 4 5 12 Lithics 1 5 2 Rocks 0 4 3 Charcoal 0 1 0 Mod. Brick 0 0 1 30 79 74 Total:

We only followed the ditch for 2.6 meters (Figs. 7.7 and 7.21), because of time restraints. The ditch remains constant in width at 75 cm and is only 10 cm deep at the center, having a concave appearance much like a very shallow trough. It appears to continue into the west baulk of the extension. There is no definite indication of this shallow ditch in the magnetometer map (Fig. 5.20), but there is a general indication of disturbance just outside the southwest corner of the white box containing the north arrow and scale. This area of disturbance corresponds to the location of feature G3. We dug a probe (50 cm x 50 cm) at the east end of feature G3 to a depth of 50 cm to provide a better profile of the ditch (Fig. 7.22) and to see what may be underneath, but only more sand was discovered and there were no artifacts.

7.2.4. Features G2–G3 and Their Relation to Hut 1 The critical aspects of Feature G3 are its date and its relation to Hut 1 or even Feature G2. Connected to those two aspects are also the questions of Feature G2’s date and function. In regards to dating Feature G3, the artifacts from stratum 3 (the fill in the ditch itself) are somewhat inconclusive. Two small sherds with glaze on them could be Roman, medieval, or modern, and there was also a small, badly worn piece of what appeared to be modern brick. These three items would seem to point to a modern date for the ditch (or at least post-Roman Period), but these could be intrusions, since the overlying stratum had modern sherds. The ditch fill also had a large amount of ancient artifacts, with 26 pottery sherds being from the Roman Period. There is also the question of G3’s apparent alignment with feature G2, both of which point toward the southwest (Fig. 7.8). Is this mere coincidence or did they somehow function together? From all the evidence, Feature G2 does appear to be ancient; it probably dates to the same time frame as Hut 1, since it has no modern artifacts but has a majority of Roman Period items. If this conclusion is accepted and we accept that G2 and G3 are connected in function, then G3 too is of ancient origin, not modern. A total of 15 cm separate the two features, both of which are pointed in the same direction (Fig. 7.8). Another interesting aspect to the two features is that G3 appears to be higher than feature G2.

Since the two parts of G3 have been discussed separately, their material will be discussed separately as well. In regards to the pit that was excavated first, there were a total of 25 pieces of pottery recovered, of which two were LBK, 19 German, two Roman, and four Ancient Unidentified (Table 7.6). Along with the pottery were four fragments of bone and one lithic (Table 7.7). In regards to the extension, full excavation had to be carried out to reach the shallow ditch. Stratum 1 was the root mat and first 5 cm of top soil, of which nothing was kept. Stratum 2 yielded an assortment of items listed in Table 7.7. There were 63 pottery sherds recovered, of which two were LBK, two Lengyel, 30 German, three Roman, six Modern, and 20 were Ancient Unidentified (Table 7.6). Along with the pottery, bone, daub, rocks, charcoal, and lithics were recovered as recorded in Table 7.7. Stratum 3 was identified as the fill of the trough. This stratum yielded 52 pieces of pottery, along with 11 bones and one tooth, three pieces of daub, two metal 162

Above, when discussing the entrance niches for Type 3 huts, it was mentioned that these niches are usually elevated higher than the base level of the huts, but in no way can we assume that the pit portion of Feature G3 is inside an entrance niche. The shape and size of the pit, its position (not along the center of the wall), and its smooth transition to the ditch portion all preclude it from being within an entrance niche. This conclusion, however, does not preclude the two features being connected to an entrance in general.

discount the argument just given above, dating both features to the same time period as Hut 1, and instead believe that G2 and G3 are later, possibly medieval or modern (based on the two small sherds and a heavily eroded brick fragment), then we could argue that the construction of the pits and ditch disrupted the south side of Hut 1. What is more, the direction of the ditch toward the southwest also corresponds to the alignment with the 1982 ditch and to the direction of plowing in the field. This line of reasoning could neatly explain the frayed nature of Hut 1, but I believe that the evidence points to a Roman Period date for features G2 and G3 (especially G2). If this date is correct, then the explanation for the frayed nature of the hut’s south side cannot rely on modern intrusion. Also, just because two ditches (the 1982 ditch and Feature G3) have the same alignment does not mean they are contemporary, since two minds in two different time periods may realize that the ground naturally slopes toward the southwest, so that was the best direction for a drainage ditch. Plowing, of course, could be mentioned and has been used to explain the position of pottery and daub fragments in Unit 2. That explanation, however, was used only to account for why so many Roman Period artifacts were found in Unit 2, possibly coming from the top of the Hut fill; the theory was not intended to explain disruptions located deeper in the earth. The simplest and most likely explanation is that, because of the hut’s position, nearer to the dividing line between Areas A1/2 and A3/4, where sand disappears and soil dominates and where the slope begins, 2,000 years of natural actions would have caused the poor preservation in Hut 1 on the south side. Modern action, of course, affected preservation on the northwest side, with the creation of the 1982 ditch.

The question of the function of both Features G2 and G3 is somewhat answered by the evidence of the microstratification at the east end of G2 (Fig. 7.17). It would appear that G2 was open to the outside and water did come in periodically. At the same time, as mentioned in Chapter 5.3.1, the ground slopes toward the west and southwest. I conclude, therefore, that both G2 and G3 functioned to remove water from the hut entrance during periods of rain, as mentioned earlier as a possible explanation for these mysterious pits. The features’ functions were connected because all the rainwater would not be caught by G3, which would drain water down the ditch sloping to the south/southwest, but some would still get into the hut, where G2 would function as a secondary rainwater-catch, absorbing the water into the sand layer under the hut. When the idea of the pits acting as a water-catch was mentioned above in section 6.1 discussing Grubenhäuser, an objection was raised that it would have been easier just to dig a ditch around the hut to move water away and there would be no need for a pit. This objection has merit, but does not really discount the pit acting as a rainwatercatch, since there was no uniform law for how to deal with rain water and not everyone thinks alike in dealing with a problem. What has to be kept in mind is that the huts had a hard daub exterior whose lower part was flush with the ground. When water would pour down onto the hut, the roof would shunt the water down all around the hut, and also the rainwater would be forced to flow around the hut because of the hard wall. The walls would act as a natural slew, and on the south side the water would come to the open entrance and naturally want to enter the hut, the floor of which was below ground level; so the interior pit was necessary to catch the water and drain it under the hut,12 while the outside pit and ditch acted as a slew for the remainder that came around the hut. At the same time, there was the need to dispose of the water coming off the roof right over the entrance. Both pits would deal with this water as well.

7.2.5. Hut 2 (Feature G4) The second hut was found 10 m to the east of Hut 1, along the same east/west axis (Fig. 7.7). This hut was also oriented toward the southwest. Hut 2 is well preserved, being buried deeper than Hut 1 and having no modern intrusions (Fig. 7.24). Once again, when a portion of the hut was discovered, the unit was expanded until the entire hut was revealed, which can be easily seen in Figure 7.25. There is no evidence for a south-side entrance niche. When excavated, the fill of the hut was convex in form, so that the center of the fill was thicker than the edges. At the center, the fill was 23 cm deep, while at the edges (east and west sides) the fill was 10 cm deep, which is the depth excavated into the sand stratum below the hut. The hut’s depth below ground surface when constructed was probably greater than 10 cm, but less than 23 cm. One can see in the fill’s profile (Fig. 7.26) that as the fill approaches the edges, there is an abrupt drop. This drop may be indicative of where the wattle-and-daub walls once stood and does in fact correspond to where the postholes are. At the transition point, where the drop begins (possibly the inside portion of the hut walls), the fill depth is practically the same at both ends: 17.5 cm for the east side and 18 cm for the

During the excavation of Hut 1, and especially after the exposure of Hut 2, it became clear that there were issues with preservation in Hut 1. When looking at previously excavated huts and also Hut 2, the fill of the hut usually created a crisp rectangular shape, but Hut 1 on the south side had a form that evokes the image of a frayed flag, with the dark fill being very broken and “fuzzy” (Fig. 7.23). When this poor preservation was mentioned in Chapter 6.2, the possibility was raised that Features G2 and G3 might have been partly responsible. If we 163

west side. These depths likely represent the true depth to which the hut had been originally excavated.

In regards to staining in the hut fill, nothing was seen out of the ordinary. The fill was a uniform black in color, with charcoal flecking. In the sand, below the hut, there was no evidence of pits either at the center or in the area where the entrance would have been. There were, however, numerous small stains scattered on the east side of the hut. Fifteen small stains were documented as can be seen in Fig. 7.28. These small stains seem to conform to those mentioned by Droberjar and Kolník as either indicating where furniture once stood or the supports for a suspended floor. The stains seem to form a vague halfcircle and may indicate a suspended floor, but this is speculation. They were quite thin and much of their form was lost as a result of our repeated cleanings in the hut for photography. Stains 1–5 and 15 were bisected and recorded (Fig. 7.29). Because of the shape of stain one, it may in fact be a rodent hole of some type. Stains two and three were both 10 cm in diameter and around two cm thick. Stain four was five cm in diameter and three cm deep. Stain five had an odd shape, vaguely like an ‘L’, the drawing of which (Fig. 7.29, 5) bisects the stain at an angle providing a cross-section of both parts. Stain six was documented during the initial excavation, but by the time the drawings and bisections were to be done, the stain was gone, having been scraped away during cleaning. Stains seven and eight, upon closer inspection, seemed to be rodent holes. Stains nine and 13 were scraped away, while stains 10–12 upon further inspection appeared to be rodent holes. Stain 14 may be a result of ant action. Stain 15, however, is well preserved, being seven cm in diameter and four cm deep. Of the 15 stains, eight (1–5, 9, 13, and 15) are probably anthropogenic, while seven appear to have resulted from bioturbation (6– 8, 10–12 and 14).

Two measurements were taken for the size of the hut, since it was undisturbed. The first measurements were of the actual fill, while the second set was taken from the postholes. The fill measurements were 4.9 m for the central axis (east to west) and 3.80 m for the short axis (north to south). The central axis, based on the east/west postholes, is 4.87 m, while the short axis (north to south) has an average of 3.54 m. If we go by the fill measurements, then the hut was 18.62 m2, while being 17.24 m2 by the postholes. Since there is no evidence of disturbance after the destruction of the hut, I believe the fill itself is a more accurate measurement of how large the hut had been. In either case, the hut is smaller than Hut 1, but still at the large end of the spectrum created by Kolník. The dimensions based on the postholes are given below in Table 7.8; as with Hut 1, the postholes are labeled according to their position in relation to the cardinal directions (Fig. 7.9). As with Hut 1, the postholes for Hut 2 varied in depth and diameter. The difference with the Hut 2 postholes is their uniformity of form, all being cylindrical and flat bottomed, with no indication of two posts being used. The only posthole with something of a taper is the center one, but since it is sometimes not present in huts and usually added later, it can be considered a non-standard posthole (Fig. 7.27). Four of the postholes, W, NW, E and NE (Fig. 7.27), were well enough preserved that the actual posts could be discerned from the postholes. Table 7.9 provides posthole diameters and depths, while the diameters and depths for the four posts that can be discerned are in parentheses after the posthole diameters and depths. Three postholes had German pottery sherds in them. The SE and E postholes each had two sherds, while the NW had one sherd.

Pieces of sintered daub were found in the fill of Hut 2. A total of 105 pieces was recovered, weighing a total of 1,003 grams, an amount that again indicates a lowintensity or short-duration burning of the hut. Most of the pieces were very small and lightweight, with only a handful making up most of the weight. Hut 2 provided some very large pieces of daub in comparison to Hut 1, but none of them had preserved impressions of wood as did the few cited from Hut 1. One large piece (cat. no. 899) appears to preserve the outside or inside surface of the hut wall, having one side that is flat (Fig. 7.30, A). There are other pieces as well that may preserve the outer surface, having what appeared to be a compacted, flat top layer (cat. nos. 910 and 915) (Fig. 7.15: 915-4-4 and Fig. 7.30, B).

Table 7.8. Distance between Postholes for Hut 2 (in meters) Posthole Distance East to West 4.87 NE to SE 3.38 NW to SW 3.70 NW to NE 2.35 SW to SE 2.91 NW to SE 4.27 NE to SW 4.47 Table 7.9. Diameter and depth of Hut 2 Postholes (in centimeters) Posthole Diameter Depth NW 26 (14) 36 (26) NE 36 (24) 32 (22) W 32 (22) 38 (32) E 42 (22) 34 (30) SW 24 48 SE 28 30 Center 26 38

Hut 2, being better preserved than Hut 1, yielded a greater amount of artifacts (see Table 7.10 and Chart 7.3). A total of 695 pottery sherds were recovered, of which 661 were German, 28 Roman, and six Other.13 The pieces recovered in Hut 2 were more varied in terms of types and preservation was better, with many large pieces, but no whole pots were recovered and no whole pots could be reconstructed. It would seem that either pieces disappeared from the hut or whole pots were never present in the fill. Perhaps only broken pots were deposited in the fill, thereby supporting the earlier 164

suggestion that the hut remains served as a trash dump for the community.

Table 7.11. Radiocarbon Dates from the Two Charcoal Samplesd S1-Hut 1a S2-Hut 2b Project No. c Beta-209061 Beta-209062 Lab No. 3 4 Unit 3 (Hut Fill) 3/4 (Hut Fill) Layer -26 -24.4 Δ 13C (%) 14 1870 ± 40 1910 ± 60 C Date B.P. AD 40–140 Calibrated ± 1σ AD 90–220 40 BC–AD 240 Calibrated ± 2σ AD 60–240 (95%) (95%) a Sample was dated using AMS-Standard delivery b Sample was dated using Radiometric-Standard delivery c Analyses were carried out by Beta Analytic Inc. of Miami, Florida. d Analyses were funded by an award from the Department of Archaeology and Center for Archaeological Studies at Boston University and a donation by Mr. and Mrs. Valentin Vrba.

Table 7.10. Pottery Numbers and Weights (in grams) from Hut 2 (Feature G4) Pottery Amount Weight German 661 12306 Roman 28 352 Other 6 30 Total: 695 12688 Along with the pottery, a wide variety of artifacts was recovered, mirroring Hut 1 in some respects in that there were again bones, bronze pieces, and rocks, in addition to the already mentioned daub (see Chart 7.4). There were 723 bone fragments recovered, along with 48 teeth, one shell, and nine antler fragments. The antler fragments appeared burnt but also worked. The bones and teeth showed a similar variety of animals as seen in Hut 1, with cows, pigs, sheep/goats, birds, deer, and even bear appearing. Once again, however, cows and pigs dominate the assemblage (see Chapter 10.4.2). In the metal category, there was a bronze hairpin, sewing needle, two belt ornaments, and a melted, unidentifiable bronze piece, along with one unidentifiable iron piece. There was a large bead made of green glass paste, along with a nodule of what appears to be graphite, along with a broken bone hairpin. Among the rocks were some smooth ones, one of which could be a whetstone, along with a sphere of red rock that may be a sling bullet or gaming piece (discussed in greater detail in Chapter 10.2.5).

7.3.1. The Date of Hut 1 The German pottery from Hut 1 provided the foundation for the date of occupation, further corroborated by a bronze fibula and the radiocarbon sample. From the German pottery, seven diagnostic sherds provided a range of AD 50 to 150, but another three sherds dated to the second century (see Chapter 8 for a detailed description of the German pottery). The Roman pottery was not well preserved and no specific well-dated examples were found; the sherds could only be identified as most likely being of the second century based on the Brick-colored wares (see Chapter 9). Because there are German diagnostic sherds that only date to the second century and the majority of dated pieces have either forms or decorations that are known to fall out of use by AD 150, it is unlikely the hut was occupied in the last half of the first century AD. The period of occupation for Hut 1 is most likely AD 100–150 based on the pottery alone.

As with Hut 1, samples were also taken from Hut 2 for other archaeological tests. Two samples were taken for radiocarbon dating along with four soil samples for micromorph analysis. One of the radiocarbon samples was sent for testing. A flotation sample was also taken from the fill of the hut for future paleobotanical research to recover pollen and any seed remains. 7.3. Period of Occupation for Huts 1 and 2

The bronze fibula was identified as an A 68, dating to AD 50–150 (see Chapter 10.2.1 for a detailed description of this fibula). Because of preservation problems, the fibula could not be better placed within the sequence of A 68 fibulae, but it does fit in with the dates provided by the pottery.

Both huts provided a wealth of datable material. The pottery from the huts indicates an occupation period sometime within the first half of the second century AD. This occupation date was corroborated by Roman bronze import items from both huts and further strengthened by radiocarbon dates (Table 7.11). In regards to the Romanstyle structure in Stupava, the huts date to just before or during the very early part of the first phase (see Chapter 3.8).

The radiocarbon sample (S1-HUT1) was AMS dated to 1870 ± 40 BP (Cal AD 60 to 240) (Table 7.11). Based on the pottery and fibula, the 1 Sigma calibrated result, even though at 68% probability, is acceptable at Cal AD 90 to 220. What is more, the intercept of the radiocarbon age with the calibration curve provides a date of Cal AD 130, which falls within the date range of AD 100–150 provided by the pottery.

In regards to the radiocarbon samples, the reporting of radiocarbon age estimates follows the suggested format of Stuiver et al. (1998). Calibrations of radiocarbon results were performed using the INTCAL98 database and the Pretoria Calibration Procedure (Stuiver and van der Plicht 1998; Stuiver et al. 1998; Talma and Vogel 1993).

It is highly probably then that the dates of AD 100–150 are accurate for Hut 1 and based on the radiocarbon data, Cal AD 130 is not out of the question for a specific year.

165

7.3.2. The Date of Hut 2

7.4. Summary

As with Hut 1, the German pottery provides the foundation for the date of occupation, with corroboration from the Roman pottery, a bronze hairpin, and the radiocarbon date. The majority of German diagnostic sherds based on forms provided a general date of the second century AD (with some forms continuing in usage into the third century); one sherd possibly dates to the middle third of the first century (see Chapter 7 for a detailed discussion of the German pottery). The decorations seen on some of the German sherds provided a date range from the first to third centuries, but the presence of comb-lines (in use only in the first and second centuries) would seem to point to the second century as the period for the German pottery in Hut 2. This date is re-enforced by the presence of two sherds with barbotine decoration, which appears in the second century and lasts only until around AD 180 (Varsik 2002, 136). In regards to Roman Pottery, the Brick-colored Wares indicate the second century as well, while one Pannonian-Striped Ware possibly dates to sometime between the mid-first and first third of the third century and another one dates to sometime within the second century and first half of the third century. Based on the pottery evidence, Hut 2 was most likely occupied sometime in the second century.

Overall, both huts accord well with the evidence discussed by Kolník and Droberjar for previously excavated Quadi huts in southwest Slovakia and Moravia. Both huts face toward the south/southwest, having a sixposthole arrangement with a seventh added later (Fig. 7.31). The huts were built of wattle-and-daub, on wooden supports. Only Hut 1 had evidence of a pit at the entrance, while Hut 2 provided evidence either of furniture legs or supports for a suspended floor. The material found in the two huts seems to follow the theory that the fill is of secondary or tertiary deposition. If the huts had primary deposition or artifacts left in situ, then we could have expected whole vessels or at least sherds that could reconstruct whole (or nearly whole) vessels positioned in the hut in such a way that would indicate they had been placed there for a purpose. Both huts, however, had none of this evidence. All the pottery found was fragmentary. Only two vessels could be partly reconstructed: a bowl (85% complete) from Hut 1 and a large cooking pot (70% complete) from Hut 2. The bronze pieces from both huts also exhibit damage, which appears to have occurred before deposition in the hut. The bronze fibula is missing the lower end of the bow, while the needles are bent or broken, the bronze hairpin is bent out of shape (but it should be noted that some hairpins were purposefully bent while in use), and the bone hairpin is missing its lower portion. The belt decoration pieces probably came from discarded broken leather belts. If the hut had been destroyed quickly by attack or accident, then we should see complete pottery vessels (even if shattered) and the few bronze items found should show heat damage (only one possibly does), but in fact there is no such solid evidence. We do know the huts were burned because of the presence of charcoal flecking and burnt wood pieces, along with sintered daub, in a soil that is blackened. The idea of secondary deposition is also supported by the lack of a large amount of sintered daub. If the hut had been purposefully burnt or destroyed, with all the wood and interior items present, then one would expect a more intense fire and so more sintered daub should have survived. The hypothesis that the lack of sintered daub could be explained away as a result of poor preservation conditions does not work, since pieces do survive along with more fragile burnt wood. Secondary deposition of artifacts is also supported by the radiocarbon date of Hut 2 in relation to the date of the pottery as already discussed above.

The bronze hairpin found in the hut could not be matched with other known hairpins. Only a general category could be identified, Beckmann’s Group IVa, which dates to the Early Roman Period, within Period B (AD 10–150). Even though the date range is large, the hairpin does fall within the range provided by the pottery. The radiocarbon sample (S2-Hut2) was dated to 1910 ± 60 BP (Cal BC 40 to AD 240). Based on the pottery and hairpin, the 1 Sigma calibrated result is acceptable at Cal AD 40 to 140. What is more, the intercept of the radiocarbon age with the calibration curve provides a date of Cal AD 90, which just falls outside of the second century date provided by the pottery, but a difference of 10 years is not that great. Hut 2 most likely dates to the second century and with the presence of the radiocarbon date, the first quarter of the second century is more likely. These date ranges match that of Hut 1 and it can be said the two huts were roughly contemporary. It is clear, however, that much of the pottery from Hut 2 are of a later date than that of Hut 1, but this is tempered by the radiocarbon date from Hut 2. There is a distinct possibility that Hut 2 was burned and abandoned at the Cal AD 90 date, but then used as a trash dump by other huts (even Hut 1) into the second century. Such a proposal would fit in well with the ideas presented earlier in this chapter about hut abandonment and reuse as trash dumps.

It would seem, therefore, that the huts were burned, but after wood had been removed. The fire did not last long, based on the lack of a large quantity of sintered daub. The presence of so much broken pottery and other artifacts, along with a large amount of animal bones, point to the use of the huts’ remains as refuse dumps. The metal pieces also show no sign of having been exposed to heat, so they too seem to have been deposited in the huts after the huts had burned. If this is all true, then there should be more huts in the near vicinity to account for the origins of the trash.

166

Endnotes complex, with those pit-houses with no post-holes first. Droberjar’s types follow a chronological sequence of use, with Type A first because these appear in the Late La Tène, while Type D huts appear in Moravia during the Late Roman and Great Migration Periods (Droberjar 2001, 100). Kolnik’s Type V huts may appear in the Czech Republic, but Droberjar apparently did not deem it necessary to make a type category for them.

1. Droberjar provides a detailed discussion and overview of the research that has been done on German settlements in various parts of Central Europe in chapter two of his Studien zu den Germanischen Siedlungen (Droberjar 1997, 9–14) 2. Only one site in Slovakia, Cífer-Pác, has evidence of long-houses and dates to the fourth century, thereby possibly making the site an isolated occurrence (Kolník 2000, 41). More may exist in Slovakia, but have yet to be discovered. Titus Kolník has commented that the characteristics of the soil make the post-holes of the longhouse hard to detect (Kolník 1998b, 69). Kolník also postulates in another article that long-houses are usually found on multi-component sites, with long occupation histories, and other kinds of structures (Kolnik 1998a, 144). These sites have many features and especially postholes, pits, and rodent tunneling across the site, and so the challenge then is to try to identify what is a post-hole and which ones belong together, since post-holes could exist from fences and other construction (Kolník 1998a, 144). Other explanations for the absence of long-houses include the possible lack of adequate woodland to supply the material for a long-house or even inadequate excavation techniques by the archaeologists (Kolník 1998a, 145). There is also the possibility that the absence of the long-house may be social. By the time Germans moved into southwest Slovakia in the early first century AD, their social organization may have changed from extended families living in long-houses (a living arrangement which would explain why long-houses were so large) to nuclear families on isolated farmsteads or small clusters of farms, using single-family huts (see Chapter/section 3.3.1). This idea would explain the dominance of huts.

4. This is in fact an old argument going back 80 years, first discussed by R. Mielcke (1915) and F. Drexel (1918), but worth noting still today (Kolník 1998a, 150). 5. These positions are based on a calculation provided by Jurgen Giesen (2005) for finding the altitude of the sun where you are. The calculation is: summer solstice: 90° – latitude of the location + 23.44° (declination of sun) = altitude of sun winter solstice: 90° – latitude of the location – 23.44° = altitude of sun For the purpose of providing the altitude of the sun in southwest Slovakia, I used the latitude of Bratislava (Lat. 49° 09’, Long. 17° 07’), which is 16 km south of Stupava. In comparison, Boston has the sun at an altitude of 71° 25’ for the summer solstice, while 24° 37’ for the winter solstice. A more southerly location, such as Panama City, Panama would have the sun for the summer solstice at 104° 48’ and for the winter solstice at 57° 48’, which is almost equal to the height of the summer sun in Slovakia. These calculations are based on the present day declination of the sun. In fact, the declination has changed over the centuries, steadily decreasing. According to a chart (see below) used by Ing. Zdeněk Ministr (1999, 250), the declination of the sun around AD 100 was actually 23.7°. If we recalculate using this declination, then the sun at winter solstice in the Bratislava area reached an altitude of 17° 21’, which is actually not much different than today’s altitude of 17° 47’, but if we pushed the clock back to 5000 BC, to the period of the LBK settlement, then the sun reached an altitude of only 16° 73’ during the winter solstice (the sun’s declination was around 24.18°).

It should be mentioned that small (smaller than longhouses) above-ground buildings do exist and have been uncovered in Slovakia and to a lesser extent in Moravia (Droberjar 1997, 28). These are usually square, with posts in the corners and possibly more along the walls. The function of these buildings is debated; they could have functioned as homes or workplaces. 3. A similar typology has been developed in Moravia for Grubenhäuser by Droberjar (2002, 99). In comparison to Droberjar: Kolník Type I II III IV V

= = = = =

Droberjar Type D A B C -

6. The average span of 20 years is an estimate based on the lifespan of wattle-and-daub constructions. A similar span is seen in other parts of the world for wattle-anddaub huts, such as among the Mikea in Madagascar (Kelly, Poyer, and Tucker 2005, 404).

Droberjar’s types do not differ from Kolnik’s in form, but only in their sequence within the typology, as indicated with Kolnik’s Type I and Droberjar’s Type D. Kolnik has his huts organized by construction, going from simple to

167

7. The Shaffer article (1993) investigates Neolithic huts in southern Italy because of the large amount of sintered daub found in them. A theory was developed to explain why so much sintered daub survived in these huts. Old huts were intentionally burned to create a large amount of sintered daub which could then be used in the construction of new huts (Shaffer 1993, 62). This process was economical because it took less time and effort than going out and creating fresh daub. Samples were subjected to archaeomagnetic study to try and prove the theory with some success (Shaffer 1993, 72). In regards to ancient German huts of the Roman Period, the theory is intriguing and worth keeping in mind, but at least at Urbárske sedliská the huts had such a small amount of sintered daub it does not seem likely that the huts were burned to create a supply for the construction of a new hut.

13. The six “Other” sherds are intrusions from around the hut. Two are LBK handles, one is Lengyel knob, and the rest are Ancient Unidentified. Because the number was small and these are clearly intrusions, they are listed under the category of “Other” instead of receiving their own categories in the discussion.

8. Unless of course Shaffer’s theory also applies to the ancient Germans and so the absence of sintered daub would be a result of its reuse in a new hut nearby. 9. The results of these tests will be published separately. 10. A micromorph analysis may lead to a better understanding of how the fill was created. Macphail and Linderholm, who have already been mentioned, have been doing tests on Grubenhaus fills in England and Germany to better understand their formation. They have identified two types of Grubenhaus fills, created by different processes and are in fact secondary or tertiary deposition of materials (Machphail and Linderholm, In Press). The two types are: “1) Heterogeneous fills that probably relate to secondary use and dumping, and which provide evidence of site activities, and 2) Moderately homogeneous fills that could relate more dominantly to in situ formation, construction/use and disuse history.” (Machphail and Linderholm, In Press) It is hoped that the samples taken from both huts may be able to show what kind of filling took place here at Urbárske sedliská. The study of these samples, however, falls outside of the dissertation. 11. The mother was quite distraught at our having come upon the nest unexpectedly. She had not been in the nest at the time of discovery and was seen to be running back and forth at the east side of Unit 3. We stopped excavations and stood outside the unit, at a distance of two meters and watched her. She ran into the unit and the nest, and removed her children one at a time to another nest located somewhere near STP 3-D-10. After she left, we examined the nest, saw dried field grasses, no artifacts, and then filled it in with excavated sand. 12. The sand under the hut would also absorb the water, helping to keep the hut dry.

168

Charts Chart 7.1. Pottery from Grubenhaus 1, Showing Amount and Percentage

Other, 3, 1% Roman, 18, 4%

German , 493, 95%

Chart 7.2. Artifacts from Grubenhaus 1, Showing Amount and Percentage

Pottery, 514, 36.5%

Bone/Teeth, 830, 58.9% Daub, 49, 3.5% Bronze, 3, 0.2% Stones, 4, 0.3% Lithics, 3, 0.2% Slag, 7, 0.5%

169

Chart 7.3. Pottery Amounts and Percentages from Hut 2 (Feature G4)

Other, 6, 1% Roman, 28, 4%

German, 661, 95%

Chart 7.4. Artifacts from Hut 2, Showing Amount and Percentage Graphite, 1, 0.1%

Seeds, 1, 0.1% Bone Pin, 1, 0.1%

Glass Bead, 1, 0.1%

Pottery, 695, 43.2% Bone/Teeth, 781, 48.5%

Daub, 105, 6.5%

Lithics, 10, 0.6%

Metal, 6, 0.4%

Stones, 9, 0.6%

170

Figures

Figure 7.1. Reconstruction of what a Grubenhaus may have looked like (Droberjar 1997, 28, Abb. 15).

Figure 7.2. Reconstruction of German long-houses from Feddersen Wierde (Franke 1980, figure four).

171

Figure 7.3. The five types of Grubenhäuser as proposed by Titus Kolník (modified image from Kolník 1998a, 145).

Figure 7.4. The four subcategories of Type III huts (modified image from Kolník 1998a, 145).

172

Figure 7.5. Ground plans of two Grubenhäuser with a pit present near the entrance niche. The top is Hut 16/41 from Branč, Slovakia (Kolník 1998a, 148) and the bottom hut is from Rajhrad, Moravia (Droberjar 1997, 24; 2001, 100).

Figure 7.6. Hut 19/61 from Branč with a storage pit at the center (Kolník 1998a, 147).

173

Figure 7.7. Ground plans of Huts 1 and 2 with pipe cut and features G2 and G3 shown, as positioned within Units 3, 4, and 8.

Figure 7.8. Ground plan of Hut 1, showing the depth of postholes and other features. The dotted line circle indicates a round stain that may be a post added later, but no actual posthole was detected.

174

Figure 7.9. Layout of line measurements and labels for the postholes.

Figure 7.10. Profiles of the seven postholes found in Hut 1. The numbers refer to Munsell colors: 1. 7.5YR 4/4, brown 9. 10YR 5/2, grayish brown 2. 10YR7/2, light gray 10. 10YR 4/4, dark yellowish brown 3. 10YR 6/4, light yellowish brown 11. 10YR 4/2, dark grayish brown 4. 10YR 6/3, pale brown 12. 10YR 4/1, dark gray 5. 10YR 6/2, light brownish gray 13. 10YR 3/6, dark yellowish brown 6. 10YR 6/1, gray 14. 10YR 3/2, very dark grayish brown 7. 10YR 5/4, yellowish brown 15. 10YR 2/2, very dark brown 8. 10YR 5/3, brown

175

Figure 7.11. Outline of stains indicating the modern trench for a pipe and the southeast side of Hut 1 is indicated by the black lines.

Figure 7.12. Photograph of Unit 3 with hut fill removed and stratum 4 revealed. Various features are indicated, including the faint square stain at the center of the photograph next to the central posthole.

176

Figure 7.13. Feature G2 is clearly seen here as the lighter colored stain at the center of the photograph.

Figure 7.14. The west end of the modern pipe-cut excavated to reveal the plastic pipe wrapped in a heavy canvas cloth. 177

Figure 7.15. Daub 437 was recovered in stratum 2 of Unit 3, but most likely comes from Hut 1. To either side of the central dividing line, running lengthwise, are the impressions of where twigs once were. Pieces 432 and 433 come from the fill of Hut 1, while 915 was recovered from the fill of Hut 2. (Drawings by Peter Šimčík)

Figure 7.16. Close-up photograph of one side of daub no. 433. The impression in the daub preserves the surface texture of the wood.

178

Figure 7.17. The east end of Feature G2 showing the microstratigraphy.

Figure 7.18. Field-mouse nest at the center of the photograph.

179

A B Figure 7.19. Image A shows the baulk between Units 2 and 3 with the pit portion of Feature G3 indicated by the arrow. The pit extended north into Unit 3, but this portion had been excavated already, with some staining still visible in the photograph. Image B shows the pit excavated.

Figure 7.20. The ditch portion of Feature G3 with fill still in place.

180

Figure 7.21. Feature G3 with pit and ditch indicated.

Figure 7.22. Profile of West baulk of Unit 3 extension. Stratum 3 is the fill of the shallow ditch that had a different color, but the dotted line and the area marked as 3a should be added to the actual size of the ditch. Strata 1 and 2 were fine sandy loam, while strata 4 and 5 are sand.The colors are as follows: Stratum 1 – 10YR 4/2 (dark grayish brown), Stratum 2 – 10YR 3/2 (very dark grayish brown), Stratum 3 – 10YR6/3 (pale brown), Stratum 4 – 10YR 4/2 and Stratum 5 – 10YR 7/2 (light gray). 181

Figure 7.23. View of Hut 1 before the fill (stratum 3) is excavated. The southwest side of the hut is in a tattered condition, while the northeast side is disrupted by the pipe-cut.

Figure 7.24. Ground plan of Hut 2, showing the depth of postholes and other features.

182

Figure 7.25. Hut 2 (Feature G4) in Unit 4 before the fill was removed. STP 3-E-15 is in the foreground.

Figure 7.26. Profile of the Hut 2 fill along the long axis from the southeast to northwest.

183

Figure 7.27. Profiles of the seven postholes found in Hut 2. The numbers refer to Munsell colors: 1. 10YR 6/4, light yellowish brown 7. 10YR 4/2, dark grayish brown 2. 10YR 6/3, pale brown 8. 10YR 3/4, dark yellowish brown 3. 10YR 5/4, yellowish brown 9. 10YR 3/3, dark brown 4. 10YR 5/3, brown 10. 10YR 3/2, very dark grayish brown 5. 10YR 5/2, grayish brown 11. 10YR 3/1, very dark gray 6. 10YR 4/3, brown 12. 10YR 2/1, black

184

Figure 7.28. Plan of Hut 2 with the fifteen identified stains indicated by a number. There were other stains, as shown on the plan unnumbered, but after repeated cleanings they could not be relocated.

185

Figure 7.29. Profiles of small stains 1–5 and 15 from Hut 2 (Feature G4). The dark stains, being either color number 7 or 8, are a sandy loam, while the surrounding matrix is sand. The Munsell colors are as follows: 1. 10YR 6/4, light yellowish brown 5. 10YR 4/3, brown 2. 10YR 6/3, pale brown 6. 10YR 4/2, dark grayish brown 3. 10YR 5/4, yellowish brown 7. 10YR 3/2, very dark grayish brown 4. 10YR 5/3, brown 8. 10YR 2/1, black

186

A B Figure 7.30. Image A shows a chunk of daub with a flat outer surface, possibly preserving the outer or inner facing of the hut wall. Image B shows another piece of daub with the flat surface indicated.

Figure 7.31. Overall map of the site with the two huts in situ (not to scale).

187

corresponding to Vannius’ kingdom, which started around the year AD 20 in southwest Slovakia, there is a marked increase in the number of German sites (Kolník 1971, 514 and 520; 1981, 116). These two phases, B1B and B1C, are seen as a period of intensified German growth in the region because of the “regnum Vannianum.” The pottery types present at German sites during this period also show an increased variety of forms and decorations. The number of settlements, cemeteries, and finds continues to increase in the B2 and later phases.

Chapter 8 The German Pottery from Features G1–G4 Chapter 8 is solely concerned with the ancient German pottery recovered from the various features dating to the Roman Period. The chapter is divided into four sections: 1) a brief introduction defining the pottery, 2) the catalog of vessel forms, 3) the catalog of decorative elements found on the pottery, and 4) a discussion of the German pottery as a group from Urbárske Sedliská with reference to other sites in southwest Slovakia and southeast Moravia.

German pottery can be divided into two groups: fine ware and coarse ware (Kolník et. al. 1993, 255–256). German fine ware is primarily dark gray to black in color with burnishing as the most common surface treatment, producing a highly lustrous finish, reminiscent of Etruscan bucherro and some Celtic vessels. Graphite can also be applied to the surface and then buffed to create a highly lustrous surface. There are also examples of fineware vessels (found at Urbárske Sedliská) with clay that is light brown in color which is then covered over with graphite, copying the appearance of the more common black-fabric vessels. Fine-ware fabric is made of well levigated, high-quality clay and is the primary determinate of this vessel type. The inclusions are small to minute (less than 1 mm), usually sand and/or mica (Droberjar 1997, 30). The black color of the fabric is uniform throughout, showing controlled firing with an intentional reducing atmosphere. The forms of these pots are usually crisp in their lines, with well defined rims and divisions between the various parts of the body. The walls of fine-ware vessels are thinner (usually around 4 mm at Urbárske Sedliská, but thicker examples do exist) than those seen in coarse ware, and help to create more delicate shapes, which are most often dishes, bowls, and cups/flagons. Some fine-ware forms are known to be imitations of Roman metal and glass vessels (see section 7.2.1 below). Fine-ware decorations are varied, ranging from patterned lines, combed lines, stamped decorations, to small wheel decorations (detailed discussion of the decorations on German pottery is below in section 7.3). The decoration is usually well controlled and applied more methodically than the decorative work seen on the coarse-ware vessels.

8.1. Introduction to German Pottery of the Early Roman Period German pottery can be divided into two periods for Southwest Slovakia (see Table 2.1 for periods and subperiods and Appendix A for cultural chronology charts). The first roughly corresponds to the Early Roman Period (AD 1–180: Phases B1–B2/C1) and can by broadly defined as German pottery being all handmade. The second period corresponds to the Late Roman Period (AD 180–375: Phases C1–C3), and can by defined as German pottery makers adopting the wheel and slowly abandoning handmade forms, the process of which continues into the Great Migration Period (AD 375–568: Phases D1–D3) (Todd 1995, 135; Droberjar 2002, 120; Kolník 1998b, 73). The following discussion of German pottery is only about handmade forms seen in the Early Roman Period, which is the period of focus for this dissertation. With the appearance of Germans in southwest Slovakia during the Early Roman Period there is a marked decline in the quality of pottery found in the region. The late La Tène period Celts produced both hand-made vessels, but also high-quality wheel-made pottery, having a wide variety of forms and decorations (Pieta and Zachar 1993, 157) (see Chapter 9.4.1 for a discussion of late La Tène pottery). These forms are replaced by German ones, corresponding to the withdrawal of Celts from southwest Slovakia and replacement by the Germans starting in the first part of the first century AD (Kolník 1981, 116). During this period, however, while some Celtic forms disappeared, others influenced German pottery types (see Chapter 9.4.1).

Coarse-ware vessels are more common than fine-ware ones and also have a greater variety of forms, decorations, and quality of fabric. Coarse-ware fabrics can have a wide range in color, but the most common are dark colors approaching black, shades of gray, some brown, and even reddish brown. The firing conditions are less controlled and the fabrics show evidence of fluctuating temperatures, usually with the core showing a reducing atmosphere (black or gray core) and the process ending with an oxidizing atmosphere (surface lighter color, varying from gray to brown to red); rarely some fabrics have multiple striations showing a greater frequency in temperature fluxuation. Wall thickness also varies greatly, usually never thinner than 5 mm and sometimes reaching up to 13 mm, as seen on vessels from Urbárske Sedliská. The clay itself also varies in quality, with some pots having a large amount of inclusions

The amount of German pottery found in southwest Slovakia is small during the earliest phase of German encroachment (the B1A phase), mostly coming from isolated graves or from a few cemeteries (Kolník 1971, 511; 1981, 116). German pottery of this period varies in quality both in the clay used and in manufacture (Kolník 1998b, 73). During the B1B and B1C phases, roughly 188

(Droberjar 1997, 30). These inclusions could have been part of the clay originally or added in by the potter. Inclusions can vary in size (Droberjar 1997, 30), including small pebbles 6 mm in diameter, seen at Urbárske Sedliská, to sand 0.03 mm in size. The most common inclusions are sand and mica, but the fabric, being black in color, may also indicate the presence of organic materials that have not been completely burned away; there is also evidence of organic temper sometimes being used (seen as elongated empty spaces in the postfired fabric). Coarse-ware forms include stew-pots, dishes, cups, bowls, and various containers, as well as colanders and other forms serving specific functions. There is a wide variety in surface decoration, including impressions, notches, incised lines, comb lines, stamp decorations, and plastic ornamentation (see section 7.3). The application of these decorations varies from welldefined patterns to random, almost chaotic, treatments.

8.2. Typology for German Pottery at Urbárske Sedliská In the following catalog, I shall be using Varsik’s typology as the foundation for the typology used at Urbárske Sedliská, because his is the most recent and based on a Quadi site roughly belonging to the same time period in Slovakia. Droberjar’s typology will also be referenced since the sites he uses are Quadi and directly north of Urbárske Sedliská. In fact, when looking at a map of the region, Urbárske Sedliská is on the western side of the Little Carpathians along the Morava River in easy contact with sites further north in Moravia (such as Blučina, Křepice, Horní Věstonice, Mikulov, Mikulčice, or Hrušky) (Fig. 8.1), while other Quadi sites to the east, such as the one excavated by V. Varsik (Varsik 2002), are on the other side of the mountain range, but closer in terms of actual distance as can be seen in Figure 8.1. There are, however, broad similarities for Quadi pottery and so the typologies by Varsik and Droberjar are not mutually exclusive. Special attention will also be given to the finds from Dúbravka (Elschek 1995 and 1997), which sits 10 km to the south and has a clear view of Stupava (Fig. 8.1). Other sites from southwest Slovakia will also be used for comparative purposes, including the first hut excavated at Urbárske Sedliská, Hut US82 (Turčan 1985), and huts at Kuchyňa (Turčan 1988), Senec and Kráľová pri Senci (Turčan 1984), Báhoň (Bartík and Turčan 1990), Trnava/Bohdanovce (Turčan 1996), and Komjatice (Točík 1980) (Fig. 8.1). The cemeteries at Sládkovičovo, Kostolná pri Dunaji, and Abrahám (Kolník 1980) and the burials at Bratislava-Devínska Nová Ves (Kraskovská 1977) will also be used for comparative purposes (Fig. 8.1).

There is no standard typology for German pottery in southwest Slovakia. Various typologies have been created for specific sites, but no overall “supra-regional” one has been adopted. One of the problems for the creation of a typology in Slovakia is the absence of published catalogs for major German settlements, such as Branč or Pobedim that were excavated up to forty years ago. There have been some attempts in the past to create typologies, but they have not been broadly adopted (M. LamiováSchmiedlová 1969 or J. Hečková 1991). A commonly used publication for reference is T. Kolník’s catalog of material recovered from three Roman-Period cemeteries from southwest Slovakia: Abrahám, Kostolná pri Dunaji, and Sládkovičovo (Kolník 1980). While there is a wealth of material from these graves, the variety of German pottery is small, limited primarily to specific fine-ware forms. In Moravia (eastern part of the Czech Republic), E. Droberjar (1997) has created a detailed typology of German pottery both in form and decoration, which is based on data from 10 excavated settlements, and serves as a useful comparative work. Droberjar’s typology is based on the type of vessel, whether a dish, tureen, cup, bottles, storage containers, etc., with sub-forms based on how the vessels differ in style (Droberjar 1997, 29). K. Elschek, working with material excavated at Dúbravka (southwest Slovakia), created a simple seriation based on the different periods at the site, but did not broaden it beyond the site itself (Elschek 1995, 52). A more recent attempt at a typology in Slovakia was done by V. Varsik (2002) based on his excavation of a settlement, called Zadné, located in the eastern suburbs of Bratislava (Bratislava-Trnávka) (Fig. 8.1), further incorporating some information from the unpublished material at Branč. Varsik’s typology is, however, another isolated one and he openly states that his typology is usable only for Bratislava-Trnávka (Varsik 2002, 152). It is hoped that once the catalogs are published from more settlements a comprehensive typology can be created for southwest Slovakia. 1

The pottery from the huts and features at Urbárske Sedliská are being grouped together based on form and decoration. The first part lists all the different vessels found with a focus on form; the fine ware is listed first followed by coarse-ware vessels. The format for each artifact entry will have a number in the catalog, followed by the inventory number, what feature it came from, if applicable, and figure number. The description will follow a format explained below and in Tables 8.1–4. The second part of the catalog will list all the decorations found on the various vessels, with references to specific vessels listed in the first part of the catalog where applicable. It should be noted that the catalog is a sampling of the diagnostic sherds recovered from the features. Those sherds preserving a large portion of the original vessel were preferred along with those pieces providing other important information.2 A total of 1,079 German sherds were collected from the two huts. Only a small percentage was diagnostic. Out of the total amount of pottery recovered from the huts, 915 were coarse ware and 168 were fine ware. Hut 1 yielded 399 coarse-ware sherds and 91 fine-ware, while Hut 2 yielded 516 coarse-ware and 77 fine-ware. These amounts are even more pronounced when total weights 189

are considered, with coarse-ware sherds having a combined weight of 15.97 kg and the fine-ware only 1.37 kg.

Table 8.2. Form II IIA – Cylindrical Neck IIA1 – body has a sharp conical form IIA2 – body with conical form, but slight bulge at top IIA3 – body with spherical shape IIB – Neck slanted inward IIB1 – body has a sharp conical form IIB2 – body with conical form, but slight bulge at top IIB3 – body with spherical shape IIC – Neck slanted outward IIC1 – body has a sharp conical form IIC2 – body with conical form, but slight bulge at top IIC3 – body with spherical shape

Even though V. Varsik states that his typology is usable only at Bratislava-Trnávka, the forms he uses and the criteria to define them can be applied to the pottery excavated at Urbárske Sedliská. Below are the definitions of the form typology I shall be using for the pottery from Urbárske Sedliská, incorporating much from V. Varsik’s typology.3 Each form is based on how the vessels were created and the shape they have (Varsik 2002, 152). The major difference between Varsik’s forms and the ones given here are the interchangeability for the forms to define both fine-ware and coarse-ware vessels. Varsik’s Forms III and IV are specific for fine ware and coarse ware respectively, which in my opinion deviates from the concept of a typology based on vessel shapes alone, and so I have combined Varsik’s III and IV into my Type III forms (Varsik 2002, 133–134).4

The lower portion can also have different treatments, either a sharp conical form (IIA1, IIB1, or IIC1); or a conical form with a slight bulge at the top (IIA2, IIB2, or IIC2) pushing the vessel toward having a shoulder and thus being an intermediate form between II and III (but lacking the clear S-shape seen in Form III vessels); or a spherical shape for the body (IIA3, IIB3, or IIC3).

1. Form I (Table 8.1 and Fig. 8.3) includes vessels that have one-piece shapes with some forms having rounded inverted rims (Form IB1, and IC1) or with straight ends, rounded or tapered (Form IA, IB2, IC2, and ID) (see Varsik 2002, 132) (Fig. 8.2 for Varsik’s typology). The vessel shape varies from cylindrical beakers (Form IA) to shallow bowls or dishes (Form IB) as well as narrow bowls/stew pots (Form IC) and cups/mugs (Form ID). According to Varsik, one-piece vessels are the most commonly found at Quadi habitation sites (Varsik 2002, 132).

3. Form III (Table 8.3 and Fig. 8.8) includes vessels consisting of three parts: the neck, shoulder, and body (see Varsik 2002, 133–134) (Figs. 8.4 and 8.5 for Varsik’s typology). Many of these vessels have an S shape to their body when viewed in profile. The S shape appears very often in German ceramics, in all sizes and includes narrow-mouthed containers (IIIA), widemouthed containers (IIIB), shallow bowls/stew pots (IIIC), and drinking vessels (IIID) (Varsik 2002, 133 and 134).

Table 8.1. Form I IA – Cylindrical Beakers IB – Shallow Bowls/Dishes IB1 – rims rounded inward IB2 – rims with straight ends, rounded/tapered IC – Deep Bowls/Stew Pots IC1 – rims rounded inward IC2 – rims with straight ends, rounded/tapered ID – Cups/Mugs

Table 8.3. Form III IIIA – Narrow-Mouthed Containers IIIB – Wide-Mouthed Containers IIIC – Shallow Bowls/Stew Pots IIID – Drinking Vessels (Cups) When indicating what form each vessel has based on the above typology the letter ‘F’ will be placed as a subscript before the type for fine-ware vessels (i.e., FIIA2), while a ‘C’ will be placed as a subscript before the type for coarse-ware vessels (i.e., CIIA2). This is done to alleviate confusion later when various vessels may be referred to and what type they are. The designations will also allow the reader to know immediately where they are in the catalog itself instead of having to look for the header. In describing the various sherds, basic terminology will be used. When describing the shape of the vessel, I shall use definitions based on Droberjar’s forms. His typology has the basic vessel shapes described as Töpfe (pots), Schüsseln (bowls), Terrinen (tureens), and Becher (cups) (Droberjar 1997, 29). Narrow-mouthed containers (Droberjar’s Töpfe) are vessels which have height (h) greater than diameter (d) (herein h>d); wide-mouthed

2. Form II (Table 8.2 and Fig. 8.5) encompasses twopiece bowls and small bowls and cups with sharp carination (see Varsik 2002, 132–133) (Fig. 8.3 for Varisk’s typology). This form has a cylindrical upper portion (neck) with rounded rim sometimes everted, or with an added lip. At the point where the neck meets the lower portion (body) there is a sharp carination; the body is conical in form, tapering inward to a flat bottom or foot. The neck is usually smaller in proportion to the lower part. There are variations to Form II in the treatments of the cylindrical upper portion, which can be either straight (IIA), slanted inward (IIB) or slanted outward (IIC).

190

containers (Droberjar’s Schüsseln) have d>h; shallow bowls/stew pots (Droberjar’s Terrinen) are usually Sshaped, having three parts, and have d>h or h=d; while cups (Droberjar’s Becher) are straight forward in shape and function (Droberjar 1997, 29). There are intermediate forms, which have h=d (Droberjar 1997, 29), and will indicated where applicable in the catalog.

decoration preserved, a line pressed into the clay as a channel (Type IXc), instead of being incised, running parallel to the axis of the vessel. The surface appears to have been burnished. The outer surface is smooth and the texture is fine. There is a moderate amount of mica inclusions. Not enough survives of the vessel to obtain a clear picture of what it had looked like. The profile of the sherd seems to point to the vessel being a beaker much like 1 above, imitating a Roman glass vessel of some type. The channeled line also seems to have been inspired by glass decoration as well.

In regards to the color of the sherds, Munsell colors will be given for the outer surface (designated O. in the catalog), interior surface (I. in the catalog), the fabric/core (C. in the catalog), the outer margin of the core (O.M. in the catalog) and inner fabric of the core (I.M. in the catalog) (see Orton et al 1993, 231). In regards to the feel of the sherd, this will be for the outer surface and either harsh (feels abrasive to the finger), rough (irregularities can be felt) or smooth (no irregularities can be felt) will be used as a descriptive term (see Orton et al. 1993, 235). In regards to the abundance of inclusions, abundant, moderate or sparse will be used (see Orton et al. 1993, 235). In examining the fabric for inclusions only a visual examination was done, so that only larger inclusions were cataloged. The texture description (or how the sherd breaks) uses the terms: subconchoidal (breaks somewhat like glass or flint), smooth (flat or slightly curved, no visible irregularities), fine (small, closely spaced irregularities), irregular (larger, more widely spaced irregularities), hackly (large and generally angular irregularities), and laminated (“stepped” effect) (see Orton et al. 1993, 235). Table 8.4 provides a summary of the terms used in the catalog.

8.2.2. Fine-Ware Vessels: Form II 3. Bowl or Cup. Inv. No. 1-3-3. Ft. G1. Fig. 8.10. Fr. FIIA2. D. 11.4 cm. B.D. 6 cm. H. 4.5 cm. Th. 4.1mm. W. 120 g. O., I., and C. are black (10YR 2/1). An almost complete small bowl (three sherds joined). The lip is rounded and everted, attached to a cylindrical neck. Where the neck joins the body there is a raised border circling the vessel, creating an illusionary shoulder. The body has a curved conical form ending at a bottom which has a raised center. The profile has a slight S-shape, thereby pushing toward a IIIC Form. There is no decoration, but the surface has been burnished. The surface is smooth and the texture is fine. There is a moderate amount of sand and mica inclusions. The basic form is common in German fine-ware ceramics—everted lip and a cylindrical neck attached to a conical body. Differences occur in the size of the vessels, the amount the lips are turned out, and in the treatment of the body. Some of these vessels can have an extremely elongated body, creating a high-footed flagon or cup (see Varsik 2002, 132, 140, Taf. I:7, 9 and 10, and 141, Taf. II:2,g).5 In Droberjar’s typology the vessel fits into form 2300 (sharp-profiled dishes) (Droberjar 1997, 51, Abb. 35:2305 or 2309). A similar vessel was found in Hut 9 at Křepice, dating to the B2 period, but the bottom is not preserved and the lip has a recessed border below it (Drobejar 1997, 309, Taf. 83:7). A vessel from a hut at Křižanovice, dating to AD 50–150, also has a very similar form, but the base appears different, being much thicker and there is an absence of the raised border at the join between neck and body (Droberjar 1997, 330, Taf. 113:5).

8.2.1. Fine-Ware Vessels: Form I 1. Rim. Inv. No. 27-3-3. Ft. G1. Figs. 8.9 and 8.31:1. Fr. FIA. D. 14 cm. P.L. 9.2 cm. Th. 8 mm. W. 65 g. O. and I. very dark gray (10YR 3/1), C. black (10YR 2/1). The surface has a smooth feel and the texture is fine. The fabric of the clay has moderate inclusions and includes both fine sand and mica. The outside has a decoration of round depressions (Type V) in what appear to be a random pattern, while both sides may have been polished, with the interior portion still bearing some luster. The shape of the vessel appears to match similar ones found in Moravia and E. Droberjar identifies them as German imitations of a Roman glass vessel, Type E 187, or a beaker, Type Gose 255 (Droberjar 1997, 112). Droberjar’s typology identifies it as a conical beaker, Type 4301 (Droberjar 1997, 60). From the finds in Moravia, these vessels date to the B2 period (AD 50–150) (Droberjar 1997, 129), which in Slovakia would fall between late B1b and the end of B2C.

4. Partial bowl. Inv. No. 535-4-3/4. Ft. G4. Fig. 8.10. Fr. FIIA2 (or FIIIB). D. 20 cm. P.L. 10.7 cm. Th. 7 mm. W. 90g. O. brown (7.5YR 5/2) for the fabric and black for the graphite, I. light brown (7.5YR 6/4) and C. brown (7.5YR 4/4).

2. Rim. Inv. No. 701-4-3/4. Ft. G4. Fig. 8.9. Fr. FIB2. D. 10 cm. P.L. 3.1 cm. Th. 4 mm. W. 2 g. O., I. and C. are black (10YR 2/1).

One quarter of a bowl that has a cylindrical neck and everted lip. At the join between neck and body there is a raised border, while the body itself is conical, with a slight curve. The outer surface does not have any decoration but graphite had been applied to it. The

A small sherd from a possible small bowl or dish. The sherd has a tapered end and there is one surface 191

surface is smooth and the texture is fine. There is a moderate amount of mica inclusions.

cemeteries. As an example, Grave A at Kostolná pri Dunaji used a biconical vessel, with two circulating raised borders on the neck (Kolník 1980, 94 and 237, Taf. LXXIV:A,a).

Even though the vessel does not have a black fabric it is included as fine ware because of the fine quality of the clay and the use of graphite on the outside. The form itself resembles 3 in basic shape, but being larger with a more pronounced conical body. A vessel of similar form, but with a straight conical body, decorated with comb lines and with a more rounded border between neck and body, was found at Dúbravka in Hut 13/88, dating to the middle third of the first century AD, B1B/B1C (Elschek 1995, 40 and 47, Abb. 3:6). This may also be the top portion of a high-footed cup, much like one from Bratislava-Vajnory, where the body curves in to an elongated form (Varsik 2002, 132 and 140, Taf. I:7). In Droberjar’s typology this vessel is a 2300 type: sharply carinated bowls (Droberjar 1997, 51–53).

7. Rim and neck. Inv. No. 704-4-3/4. Ft. G4. Fig. 8.10. Fr. FIIB (?). D. 21 cm. P.L. 5 cm. Th. 4 mm. W. 10g. O. and C. are black (7.5YR 5/6) and I. is mottled with shades of black. Rim and neck sherd of a biconical vessel. The rim is everted and the neck is turned inward. It appears that the vessel was graphitized. The outer surface is smooth and the texture is fine. There is a moderate amount of mica inclusions. 8. Rim and neck. Inv. No. 207-3. Ft. G2. Fig. 8.10. Fr. FIIC2. D. 18 cm. P.L. 5.8 cm. Th. 5 mm. O. and I. are very dark gray (10YR 3/1) and C. black.

5. Rim. Inv. No. 696-4-3/4. Ft. G4. Fig. 8.10. Fr. FIIA2. D. 12 cm. P.L. 3.6 cm. Th. 4.7 mm. W. 10 g. O., I., and C. are black (10YR 2/1).

Sherd preserving rim, neck, and portion of the shoulder to a bowl. The rim has a lip and is turned outward, while the neck is funnel-shaped. There is a shallow shoulder to the vessel helping to create a weak S-shape body. The outer surface is smooth and the texture is fine. There is a sparse amount of sand and mica inclusions.

The sherd is from a small bowl. The sherd has a cylindrical neck with a very large lip. The join between neck and body has a raised border, while the body appears to have been conical, but not enough is preserved to be certain. The outer surface is smooth and the texture is fine. There is a moderate amount of sand and mica inclusions. The surface appears to have been burnished.

8.2.3. Fine-Ware Vessels: Form III 9. Rim and neck. Inv. No. 703-4-3/4. Ft. G4. Fig. 8.11. Fr. FIIIA. D. 10 cm. P.L. 6.3 cm. Th. 4 mm. W. 10 g. O. and I. light brownish gray (10YR 6/2) and C. grayish brown (10YR 5/2).

6. Rim. Inv. No. 26-3-3. Ft. G1. Fig. 8.10. Fr. FIIB1 (or FIIB2). D. 9.5 cm. P.L. 6.2 cm. Th. 4 mm. W. 20 g. O., I., and C. are very dark gray (10YR 3/1). A rim sherd to a narrow-mouthed container with a conical neck. There is a raised border partway down the neck, going round the vessel. The shoulder was most likely carinated, with the conical body matching the angle of the neck, creating a biconical vessel. The outer surface is smooth, appearing to have been burnished, and the texture is fine. There is a moderate amount of mica inclusions.

The sherd preserves the lip and neck to a narrow-mouthed container. The lip is out-turned and the neck is high cylindrical, which starts to curve out as it approaches the body. The vessel would have had a weak S-shape. The outer and interior surfaces appear to have had graphite applied. The outer surface is smooth and the texture is fine. There is a moderate amount of mica inclusions. With the absence of the body, it is hard to determine parallels. The high cylindrical neck, however, seems to point to a possible identification along the lines of Droberjar’s type 3300: tureens with cylindrical necks (Droberjar 1997, 58, Abb. 42: 3300). If in fact the vessel’s neck should be slanted more outward, creating a funnel shape, then the vessel would easily fit into Droberjar’s type 3400: tureens with funnel-formed necks, and more specifically type 3404 (Droberjar 1997, 59, Abb. 43:3400 and 3404). The 3400 type dates to the B2 period in Moravia (Droberjar 1997, 95). I am more inclined to believe 9 is similar to Droberjar’s type 3404, but the lack of the body prevents closer identification.

According to V. Varsik it is a rare form and parallels have been found only at Bohdanovce, Zohor and at Bratislava-Trnávka (Varsik 2002, 133 and 147, Taf. VIII:22). The vessel from Bratislava-Trnávka is larger, with a higher neck. The date of the form seems to be in the first half of the second century AD (Varsik 2002, 133). The vessel from Bohdanovce is slightly different in that its mouth is wider and the border is a grooved line instead of the raised border. V. Turčan classifies it as a form identified by E. Droberjar as late La Tène form that was used by the Przeworsk Culture and from them infiltrates into German usage; the date of the form falls sometime at the end of the first century and beginning of the second century AD (Turčan 1996, 109–110, Obr. 4:2). It must be said, however, that with so little preserved, one cannot say with any certainty as to what type of vessel is 6. The biconical form itself is found in larger fine ware examples, such as those used in

10. Half of small bowl or cup. Inv. No. 9-3-3. Ft. G1. Fig. 8.11. Fr. FIIIB. D. 13 cm. B.D. 5.5 cm. H. 11 cm. Th. 6 mm. O., I., and C. are black (10YR 2/1).

192

Twelve sherds were joined for the vessel. The form is a weak S-shape, with no lip. The neck is cylindrical, the rim tapered, the shoulder is shallow, and the body is conical. There is no decoration on the outer surface, but it was burnished. The surface is smooth and the texture is fine. There are moderate amounts of sand and mica.

second century AD (Varsik 2002, 132–133, Taf. II:2c). The main difference is the protruding border area, which is more rounded on the vessel from Branč. In Droberjar’s typology 12 fits into 2300, sharp profiled dishes, but the sub-type could either be 2304 or 2321, dating to the B2 and B2/C1 phases (Droberjar 1997, 51–52, 95 and 96).

The form can be considered an intermediate form between II and III, since the S-shape is not pronounced lacking an out-turned lip/rim. In Droberjar’s typology the vessel falls in the 2100 series (Intermediate Form I: Sshaped formed pot-dishes), but there is no exact match with any of the sub-types (Droberjar 1997, 49, Abb. 33:2100).

13. Half of dish or cup. Inv. No. 948-4-3/4. Ft. G4. Fig. 8.12. Fr. FIIIB. D. 12 cm. H. 10.5 cm. Th. 5 mm. O., I., and C. black (10YR 2/1) Two sherds refitted to half of a dish or cup with the rim turned slightly outward, coming to a tapered end. The neck is cylindrical attached to a body that has a slight bulge at the top and the remainder conical. There is a recessed border at the join between neck and body, while the bottom of the vessel has a foot. The form is a weak Sshaped and could be considered an intermediate form. There is no decoration on the surface, but the outer surface does appear to have been burnished. The outer surface is smooth and the texture is fine. There is a sparse amount of mica and sand inclusions.

11. Rim, neck, body. Inv. No. 28-3-3. Ft. G1. Fig. 8.11. Fr. FIIIB (or FIIIC). D. 8 cm. P.L. 6 cm. Th. 4 mm. W. 17 g. O., I., and C. are black (10YR 2/1). This is a sherd preserving the neck, shoulder, and part of the body to a wide-mouthed container or bowl with a cylindrical neck, shallow shoulder, and spherical body. The outer surface is smooth and the texture is fine. There is a moderate amount of mica inclusions. The surface appears to have been burnished.

The form is basically the same as 10, except for the recessed border between neck and body and the raised border on the foot.

A similar vessel (but with a rounded rim) was found by Varsik at Bratislava-Trnávka from Grubenhaus 3 and he places this form in the B1 and B2 periods (Varsik 2002, 133 and 144, Taf. IV:7). V. Turčan found a similar vessel (the lip is larger and more pronounced) at Bohdanovce (Turčan 1996, 109, Obr. 3:5). Droberjar has a similar type, 3404, a tureen with funnel-form neck, coming from Hut 6 at Křepice and dating to the B2 period (Droberjar 1997, 292, Taf. 66:14). Hut 9 at Křepice also yielded a similar vessel dating to the B2 period (Droberjar 1997, 306, Taf. 80:2). A similar form was found by T. Kolník at Abrahám in grave 5, but it is a wheel-made import from the empire, possibly Raetia (Kolník 1980, 23 and 175, Taf. XII: Grab 5:g1). There are differences, with the imported vessel having a taller neck and decoration on the body, but the basic shape is the same. It is possible that 11 is part of another German form imitating Roman imports, perhaps like 1 above.

14. Rim, neck, body. Inv. No. 686-4-3/4. Ft. G4. Fig. 8.12. Fr. FIIIC. D. 23 cm. P.L. 7.1 cm. Th. 6 mm. W. 35 g. O. and I. are black (10YR 2/1), C. very dark gray (10YR 3/1). Sherd preserving the neck and part of the body to a widemouthed container or bowl. The neck is slanted inward joining an out-turned lip, while the body appears to have been conical, but could have been spherical. There is a weak S-shape to the form of the vessel. The outer surface is smooth and the texture is irregular. There is a moderate amount of sand and mica. 8.2.4. Fine-Ware Vessels: Bases and Unidentifiable Rims

12. Rim. Inv. No. 33-3-3. Ft. G1. Fig. 8.12. Fr. FIIIB. D. 18 cm. P.L. 9 cm. Th. 5 mm. W. 35 g. O., I., and C. are black (10YR 3/1).

15. Base. Inv. No. 22-3-3. Ft. G1. Fig. 8.13. Fr. Un. B.D. 10 cm. P.H. 3 cm. Th. 5 mm. W. 10 g. O. and I. black, C. very dark gray (10YR 3/1).

Originally discovered in STP 3-C-15, later identified as coming from G1. The rim is from a bowl. There is a well defined out-turned lip. The joining shoulder between the neck and body has a thickened body, flared outward as a well defined border. On the lower portion of this border is a low protrusion creating a softly defined line. The outer surface is very smooth and was probably burnished, while the texture is fine. The only discernable inclusion is mica in abundant quantity.

Sherd preserving the base and part of the body to a probable narrow-mouthed container. The outer surface was burnished. The outer surface is smooth and the texture is fine. There is a moderate amount of sand and mica inclusions. 16. Base. Inv. No. 23-3-3. Ft. G1. Fig. 8.13. Fr. IIIB or IIID. B.D. 4 cm. P.H. 2 cm. Th. 6 mm W. 35 g. for body and 3.5 mm for base. O., I., and C. are black.

Varsik has a similar vessel, from Branč, labeled as an intermediate Form II/III, which he dates to the first century AD, disappearing in the very early part of the

The footed base to what was probably a high-footed drinking vessel or bowl. The outer surface is smooth and 193

the texture is fine. There are sand and mica inclusions in moderate amounts. The outer surface may have been burnished.

22. Base. Inv. No. 706-4-3/4. Ft. G4. Fig. 8.13. Fr. Un. B.D. 5 cm. P.H. 2.1 cm. Th. 6 mm at body and 8 mm at base. W. 20 g. O., I., O.M., and I.M. black, C. very dark gray (10YR 3/1).

17. Base. Inv. No. 24-3-3. Ft. G1. Fig. 8.13. Fr. FIIA2 or FIIIB. B.D. 4 cm. Th. 4 mm. W. 5 g. O. and I. very dark gray (10YR 3/10), C. black (10YR 2/1).

Sherd preserving the base and a portion of the conical body to a dish or pot. The outer surface is smooth and the texture is fine. There is a moderate amount of mica inclusions and a sparse amount of quartz pebbles up to 2 mm in diameter.

The bottom to a small fine-ware vessel, most likely a bowl. The surface is smooth and the texture is fine. The outer surface may have been burnished.

23. Rim. Inv. No. 31-3-3. Ft. G1. Fig. 8.14. Fr. Un. (FIII ?). D. 9 cm. P.L. 3 cm. Th. 4 mm. W. 2 g. O., I., and C. very dark gray (10YR 3/1).

The form may have been like 3. 18. Base. Inv. No. 25-3-3. Ft. G1. Fig. 8.13. Fr. Un. B.D. 4 cm. Th. 6 mm at body and 10 mm at base. W. 50 g. O., I., and C. black.

Rim fragment too small to postulate vessel form. The outer surface is smooth and the texture is fine. There is a moderate amount of mica inclusions. The surface appears to have been burnished.

Sherd preserving the base and a small portion of the body. There outer surface is smooth and the texture is fine. There are moderate amounts of sand and mica inclusions.

24. Rim. Inv. No. 32-3-3. Ft. G1. Fig. 8.14. Fr. Un. (FIII ?). D. 16.5 cm. P.L. 4 cm. Th. 4 mm. W. 2 g. O. and I. very dark gray (10YR 3/1), C. dark gray (10YR 4/1).

19. Base. Inv. No. 65-3-3. Ft. G1. Fig. 8.13. Fr. Un. B.D. 5 cm. P.H. 5 cm. Th. 8 mm at the body and 10 mm at the base. W. 100 g. O. and I. dark gray (10YR 4/1), C. very dark gray (10YR 3/1).

Rim fragment too small to postulate vessel form. The outer surface is smooth and the texture is fine. There is a moderate amount of sand and mica inclusions. The surface appears to have been burnished.

Base with a portion of the body preserved. The outer surface is smooth and the texture is fine. There is a moderate amount of mica and sand inclusions with a sparse amount of quartz pebbles up to 2 mm in diameter.

25. Rim. Inv. No. 38-3-3. Ft. G1. Fig. 8.14. Fr. Un. (FIII ?). D. 23 cm. P.L. 3.5 cm. Th. 5 mm. W. 7 g. O. and I. dark gray (10YR 4/1), C. very dark gray (10YR 3/1).

20. Base. Inv. No. 574-4-3/4. Ft. G4. Fig. 8.13. Fr. FIIID. B.D. 5 cm. P.H. 4 cm. Th. 5 mm at body and 9 mm at base. W. 100 g. O., I., and C. black (10YR 2/1).

Two sherds joined of a rim too small to postulate vessel form. The outer surface is smooth and the texture is fine. The outer surface may have been burnished. There are moderates amount of sand and mica inclusions.

A base with foot preserving a portion of the conical body. The base is flat and solid. The outer surface is smooth and the interior is rough, with fine texture. There is a moderate amount of mica and sand inclusions.

26. Rim. Inv. No. 677-4-3/4. Ft. G4. Fig. 8.14. Fr. Un. D. 20 cm. P.L. 3.6 cm. Th. 5 mm. W. 5 g. O., I., and C. black (10YR 2/1).

Not enough survives of the vessel for precise identification, but with the base’s foot and the body having a high angled conical form, it is possible the vessel was either a footed dish or drinking vessel. A similar base was found in hut US82 by V. Turčan (Turčan 1985, 102, Obr. 7:17 and 104).

Rim sherd too small for vessel identification. The lip is out-turned and the neck appears to have been funnel shaped. The outer surface is smooth and the texture is fine. There is a moderate amount of mica inclusions and sparse amounts of sand.

21. Base. Inv. No. 705-4-3/4. Ft. G4. Fig. 8.13. Fr. FIIIB. D. 5.5. P.H. 4.5 cm. Th. 6 mm. W. 50 g. O., I., O.M., I.M. black, C. is very dark gray (10YR 3/1).

27. Rim. Inv. No. 685-4-3/4. Ft. G4. Fig. 8.14. Fr. Un. (FIIA ?). D. 24 cm. P.L. 6.7 cm. Th. 5 mm. W. 20 g. O., I., and C. black (10YR 2/1).

Sherd preserving the base and conical body to a container. The bottom of the base may have been rounded, but not enough survives to absolutely certain. The outer surface of the vessel appears to have been burnished. The outer surface is smooth and the texture is fine. There is a moderate amount of mica inclusions.

Rim piece to either a wide-mouthed container or bowl. The neck is cylindrical with out-turned lip, which is irregular. The surface is smooth and the texture is fine. There is a moderate amount of mica and sand inclusions, some reaching up to 1 mm in size.

194

28. Rim. Inv. No. 692-4-3/4. Ft. G4. Fig. 8.14. Fr. Un. D. 11 cm. P.L. 2.5 cm. Th. 4 mm. W. 5 g. O., I., and C. black (10YR 2/1).

amounts of sand and mica, with sparse pebbles, some reaching 2 mm in diameter. 32 may date to the late B1B to B2 periods, based on similar vessels with this pattern from Moravia (sites of Blučina and Křepice; Droberjar lists B2, according to the Czech/Moravian chronology, i.e., AD 50–150) (Droberjar 1997, 249, Taf. 23:11; 280, Taf. 54:20; 283, Taf. 57:13). See 33 for a description of similar vessel shapes.

Rim sherd that is too small to identify, but possibly belonging to a tureen. The sherd has an out-turned lip and the neck appears to have been cylindrical. The outer surface is smooth and the texture is fine. There is a moderate amount of mica inclusions.

33. Rim. Inv. No. 90-3-3. Ft. G1. Figs. 8.15, 8.44:33. Fr. CIB1. D. 24 cm. P.L. 9 cm. Th. 7.5 mm. W. 40 g. O. mottled grayish brown (10YR 5/2) and black (10YR 2/1), I. and C. black (10YR 2/1).

29. Rim. Inv. No. 694-4-3/4. Ft. G4. Fig. 8.14. Fr. Un. D. 16 cm. P.L. 2.8 cm. Th. 5 mm. W. 5 g. O., I., and C. black (10YR 2/1). Rim sherd too small to allow proper vessel identification. The lip is out-turned and the neck appears to have been funnel-shaped. The surface is smooth and the texture is fine. The outer surface was burnished. There is a moderate amount of mica inclusions and a sparse amount of sand.

A rim sherd belonging to a deep bowl or cooking pot with an in-curved rim. On the outer surface is a band of fingernail impressions (Type XIIa) or the top portions of long notched lines running down the body of the vessel, while the top of the rim itself also has a series of notches (Type XVI). The mottling between grayish brown and black on the outer surface appears to be caused by burning, either from the pot’s use as a cooking vessel or some other fire. The surface of the sherd is harsh and the texture is irregular. The only visible inclusions are sand and mica in moderate quantity, but ill sorted with some up to 3 mm in size.

30. Rim. Inv. No. 698-4-3/4. Ft. G4. Fig. 8.14. Fr. Un. D. 15 cm. P.L. 5 cm. Th. 6.5 mm. W. 10 g. O., I., and C. black (10YR 2/1). Rim piece that is too small for proper vessel identification. The sherd preserves a funnel neck and the rim is turned inward. The outer surface is smooth and the texture is fine. There is a moderate amount of mica and sand inclusions, with some mica reaching up to 2 mm in size.

With so little of the body preserved, it is hard to identify the probable vessel type. 33 has some resemblance to vessels found by Turčan at hut US82 (1985, 99, Obr. 5:2 and 101, Obr. 6:8) and Droberjar’s form 2450, which dates to either B2 or B2/C1 (1997, 54 and 96). An intriguing vessel from Dúbravka has a similar form and decorative pattern on the rim, while the body has notched lines running down the vessel. These lines may be on 35, but only the very upper portions survive (Elschek 1995, 49, Abb. 5:14). The vessel came from feature 10/88 which was dated to B1C (Elschek 1995, 40).

8.2.5. Coarse-Ware Vessels: Form I 31. Rim. Inv. No. 488-4-3/4. Ft. G4. Fig. 8.15. Fr. CIBI. D. 14 cm. P.L. 4.5 cm. Th. 7 mm. W. 15 g. O. and I. black (7.5YR 2.5/1), C. very dark gray (7.5YR 3/1).

34. Rim and body. Inv. No. 226-3. Ft. G2. Fig. 8.15. Fr. CIB1. D. 24.2 cm. P.L. 6.9 cm. Th. 7 mm. O. mottled light brownish gray (10YR 6/2) and grayish brown (10YR 5/2), I. grayish brown (10YR 5/2), C. very dark gray (10YR 3/2).

Rim sherd to a bowl with the rim rounded and bent inward. There is a comb-line decoration (unknown type, too little survives) on the outer body. The outer surface feels smooth and the texture is irregular. There is a moderate amount of mica and sand inclusions, with sparse quartz pebbles reaching up to 4 mm in size.

Sherd preserving the rim and body to a dish/bowl. The body is rounded with the rim turned inward. The surface is rough and the texture is irregular. There is a moderate amount of mica and sand inclusions with some reaching 2 mm diameter.

32. Rim and body. Inv. No. 7-3-3. Ft. G1. Fig. 8.15 and 8.42:143. Fr. CIB1. D. 28 cm. P.L. 27 cm. Th. 6 mm. W. 300 g. O. mottled brown (7.5YR 5/2) and black (7.5YR 2.5/1), I. black, O.M. yellowish red (5YR 5/6), I.M. and C. black.

35. Rim. Inv. No. 496-4-3/4. Ft. G4. Fig. 8.16. Fr. CIB1. D. 18.5 cm. P.L. 6.5 cm. Th. 7 mm. O. and I. very dark gray (10YR 3/1), C. black (10YR 2/1).

Six sherds joined, forming part of a rim and body of a deep dish or cooking pot, outer surface decorated by incised lines placed in a chaotic criss-cross pattern (Type XIVd). The body turns inward approaching the rim, but then the rim is bent back, to stand straight. The outer surface is mottled with blackened sections suggesting having been burned or used in cooking. The outer surface feels rough and the texture is hackly. There are moderate

Rim sherd of a dish/bowl with the rim rounded. The outer surface is rough and the texture is irregular. There is a moderate amount of mica and sand inclusions. Similar vessel shapes can be seen at the site between Trnava and Bohdanovce, where two dishes were found (Turčan 1996, 115, Obr. 8:1 and 15). In Droberjar’s 195

typology the vessel is a type 2000, but in regards to subtype, it is hard to determine since the bottom does not survive. The vessel could either be a 2400, deep dish, or 2700, dish with rounded bottom.

Four sherds joined from rim and body belonging to a deep bowl or cooking pot of some sort but with the break occurring at the point where the neck joins the body so that a precise identification cannot be given. The inner curve of the sherds, however, suggests that the vessel’s lower portion curves inward, but whether conical or spherical is unknown. The surface has a rough feel and the texture is fine, with sand and mica inclusions in abundance.

36. Rim and body. Inv. No. 206-3. Ft. G2. Fig. 8.16. Fr. CIB2. D. 16 cm. P.L. 6.6 cm. Th. 9 mm. W. 40 g. O. brown (7.5YR 5/2), I. dark gray (7.5YR 4/1), C. black. Sherd preserving rim and body of a bowl. The rim is rounded and bent inward, while the body is conical in shape. The outer surface is rough and the texture is irregular. There is a moderate amount of mica and sand inclusions, some reaching 3 mm in diameter.

A similar fragment with combed decoration was found by V. Turčan at Urbárske Sedliská in Grubenhaus US82 (Turčan 1985, 105, Obr. 9:3). 41. Rim. Inv. No. 950-4-3/4. Ft. G4. Fig. 8.18. Fr. CIIA2 (CIIIC). D. 19 cm. P.L. 11.5 cm. Th. 6.5 mm. W. 35 g. O. very pale brown (10YR 7/3), I. and C. gray (10YR 5/1).

37. Rim and body. Inv. No. 544-4-3/4. Ft. G4. Fig. 8.16. Fr. CIB2. D. 16 cm. P.L. 7.1 cm. Th. 7 mm. W. 35 g. O. mottled pale brown (10YR 6/3) and gray (10YR 6/1), I. and I.M. reddish yellow (7.5YR 6/6), O.M. and C. gray (10YR 6/1).

Two rim sherds joined, belong to a bowl. The lip is thick and out-turned, attached to a cylindrical neck. Where the neck joins the body, there is no shoulder or sharp border, but instead the join is smooth with the body of the vessel appearing to have been conical. There is a moderate amount of mica and sand inclusions.

Sherd preserving the rim and body of a bowl. The rim is tapered. The outer surface is smooth and the texture is irregular. There is a moderate amount of mica and sand and a sparse amount of quartz pebbles. A similar vessel in form was found at Kuchyňa by V. Turčan from feature 4 (Turčan 1988, 114 and 116, Obr. 7:7). 37 also fits into the 2000 type of Droberjar’s system and also may fit into either type 2400 or 2700.

The shape is similar to fine-ware vessel 3; 41 may have had a graphitized outer surface, like that of 4, but has been worn away. If this is so, then the vessel is fine ware, but one using coarse-ware clay. It is not as strongly carinated as a fine-ware vessel. In Droberjar’s typology it may fall under type 2300, sharply formed dishes, more according to the basic shape then actually being sharply profiled.

38. Rim and body. Inv. No. 70-3-3. Ft. G1. Fig. 8.17. Fr. CIC1. D. 25 cm. P.L. 4.5 cm. Th. 11.5 mm. W. 70 g. O. light brown (7.5YR 6/4), I. reddish yellow (7.5YR 6/6), C. strong brown (7.5YR 4/6).

42. Rim and body. Inv. No. 493-4-3/4. Ft. G4. Fig. 8.18. Fr. CIIA3. D. 22 cm. P.L. 14.6 cm. Th. 7 mm. W. 140 g. O. mottled light brown (7.5YR 6/4) and strong brown (7.5YR 4/6), I. black (7.5YR 2/1), O.M. yellowish red (5YR 5/8), I.M. and C. black (7.5YR 2.5/1).

Two sherds joined from the rim and body of a deep bowl/cooking vessel. The walls of the vessel curve inward with a rounded rim, and it is difficult to say if there is a lip, or it is a potter’s finger line. The outer surface feels rough and the texture is irregular. There is an abundant amount of sand and mica up to 1 mm in size.

Sherd preserving the neck and body to a deep bowl or cooking pot. The lip is out-turned, attached to a cylindrical neck, which joins a spherical body. The outer surface is rough and the texture is irregular. There is a moderate amount of mica with sparse amounts of small pebbles up to 1 mm in diameter.

39. Rim and body. Inv. No. 955-4-3/4. Ft. G4. Fig. 8.17. Fr. CIC1. D. 34 cm. P.L. 14.8 cm. Th. 11 mm. W. 245 g. O. and I. mottled black and strong brown (7.5YR 5/6), C. black. Two sherds joined, preserving a portion of the rim and body to a large storage vessel or cooking pot. The rim is curved inward and has a flat end, but undulates probably because of the maker’s unconcern for a well formed rim. There is a moderate amount of sand and mica inclusions.

A similar vessel form is seen in Hut 13/88 from Bratislava-Dúbravka (Elschek 1995, 47, Abb. 3:10). 42 fits into Droberjar’s typology as a type 3300, tureen with cylindrical neck (Droberjar 1997, 58, Abb. 42:3300). 43. Body. Inv. No. 8-3-3. Ft. G1. Fig. 8.18. Fr. CIIB1. Est. D. at body is 27 cm. P.L. 10 cm. Th. 5 mm at neck, 13 mm at body. W. 100g. O. dark gray (10YR 4/1), I. very dark gray (10YR 3/1), C. black (10YR 2/1).

8.2.6. Coarse-Ware Vessels: Form II 40. Rim. Inv. No. 2-3-3. Ft. G1. Fig. 8.18. Fr. CIIA2. D. 20 cm. P.L. 10 cm. Th. 4.8 mm. W. 45 g. O., I., O.M., and I.M. reddish gray (10YR 5/2), C. very dark bluish gray (Gley 2 3/1-5PB).

Five sherds joined, preserving the biconical shape of the vessel with a sharp carination between the neck and body. The outer surface is smooth and the texture is irregular. There is a moderate amount of sand and mica inclusions 196

with a sparse amount of pebbles, reaching up to 2 mm in diameter.

rounded with a slight bulge. On the outer surface there is no decoration present. What appears to be an incised line is in fact surface damage. The outer surface is rough and the texture is hackly. There is an abundance of coarse sand inclusions up to 3 mm in size and a moderate amount of mica.

44. Rim and body. Inv. No. 4-3-3. Ft. G1. Figs. 8.19 and 8.37:117. Fr. CIIB2. D. 20 cm. P.L. 13.5 cm. Th. 6.8 mm. W. 105 g. O. mottled reddish brown (5YR 5/3) and Dark Gray (5YR 4/1), appears burned, I. grayish brown (10YR 5/4), C. dark bluish gray (Gley 2 4/1-5PB).

47. Rim and neck. Inv. No. 95-3-3. Ft. G1. Fig. 8.20. Fr. CIIB2. D. 25 cm. P.L. 8.5 cm. Th. 8 mm. W. 45 g. O. and I. dark gray (7.5YR 4/1), C. black (7.5YR 2.5/1).

Three sherds joined, forming upper portions of a container or cooking pot with cylindrical neck slanted inward, having a rounded rim, while there is a slight bulge at the transition to the body, creating a shallow shoulder. The body is conical, curving inward toward the bottom of the vessel which does not survive. This vessel can also be seen as an intermediary from between II and III. Decorating the outer surface are randomly placed curved comb-line ornament (Type XIIId). The outer surface is smooth and the fabric texture is irregular. There is a moderate amount of sand and mica, with sparse amounts of large sand/pebbles up to 2 mm in size.

Two sherds refitted, part of a deep bowl or tureen, with a conical neck and a very small lip. The outer surface feels smooth and the texture is fine. Inclusions include a moderate amount of mica and sand, and a sparse amount of small rocks up to 3 mm in size. For a discussion of the form please see 45. 48. Rim, neck, and body. Inv. No. 492-4-3/4. Ft. G4. Fig. 8.20. Fr. CIIB2. D. 22 cm. P.L. 13.3 cm. Th. 8 mm. W. 140 g. O. mottled black (7.5YR 2.1/1) and brown (7.5YR 4/3), I. very dark gray (7.5YR 3/1), O.M. very dark gray (7.5YR 3/1), I.M. and C. black (7.5YR 2.1/1).

A vessel from hut US82 at Urbárske Sedliská has both a similar shape and has comb-line decoration, but elongated (Turčan 1985, 99, Obr. 5: 4). Another example of similar form (but with a slightly pronounced lip) with combed-line decoration comes from a hut (Hut 1) near the town of Bohdanovce; the date of the vessel, however, is not certain and either dates at the first half of the second century AD or the break between the second and third centuries AD (Turčan 1996, 110, 111, Obr.5:17, and 119).

Sherd preserving the rim, neck, and part of the body of a container/cooking pot. The neck is slanted inward, but then turned out for the rim. The rim itself is very undulating and clearly hand made, without concern for a well formed rim. There is a slight shoulder where the neck joins the body, creating a very weak S-shape, while the body itself is conical. The outer surface is rough and the texture is irregular. There is a moderate amount of sand and a sparse amount of mica inclusions.

45. Rim. Inv. 84-3-3. Ft. G1. Fig. 8.19. Fr. CIIB2. D. 18 cm. P.L. 4.5 cm. Th. 4 mm. W. 10 g. O. mottled light yellowish brown (10YR 6/4) and dark gray (10YR 4/1), I. dark gray (10YR 4/1), C. black (10YR 2/1).

A similar example was found in Hut US82 at Urbárske Sedliská, but with comb line decoration (Turčan 1985, 96, 97, Obr. 3:1). A similar vessel was also found at Kuchyňa, gray-black in color and approximately the same size (Turčan 1988, 114 and 116, Obr. 7:6). Kuchyňa dates to the Early Roman Period, sometime in the second half of the second century AD, but this date is a rough estimate because of the lack of precisely datable finds and poor preservation of the features (Turčan 1988, 108–109 and 124). In Droberjar’s typology, 48 falls into the 2100 series, intermediate form I (S-shaped pot-dishes) (Droberjar 1997, 50, Abb. 34:2100).

Rim piece from a possible container or cooking pot. The rim is out-turned and rounded. The sherd is part of the neck of the vessel, slanted inward, with only a small portion of the body still attached, showing that the lower portion was probably spherical. There is a moderate amount of mica inclusions in the fabric, mostly less then 1/8 mm in size. With the lower portion not preserved, it is difficult to accurately identify the vessel and find similar ones. 45 may be placed in Droberjar’s 3200 group or 3600 (Droberjar 1997, 58 and 59).

49. Rim, neck, body. Inv. No. 540-4-/3/4. Ft. G4. Fig. 8.20. Fr. CIIB2. D. 20 cm. P.L. 4.2 cm. Th. 9 mm. W. 10 g. O. and I. mottled brown (7.5YR 5/3) and black (7.5YR 2.5/1), C. brown (7.5YR 5/2).

46. Rim and neck. Inv. No. 88-3-3. Ft. G1. Fig. 8.19. Fr. CIIB2. D. 36 cm. P.L. 8 cm. Th. 12 mm. W. 50 g. O. yellowish brown (10YR 5/4), I. mottled dark gray (10YR 4/1) and light brown (7.5YR 6/4), C. dark gray.

Rim sherd preserving the neck and a portion of the body. The rim has an everted lip, while the neck and body create a biconical vessel, probably a dish. The outer surface is smooth and the texture is irregular. There is a moderate amount of mica and sand inclusions.

Two sherds joined, making up the rim and neck of a large storage vessel or cooking pot. The neck is conical in form, with the rim bent back outward, but only reaching a parallel position to the axis of the vessel. The rim is 197

The form of the vessel fits into Droberjar’s 2300 type, sharply carinated dishes, and more specifically with subtype 2338, which dates to the B2/C1 period and was found at the sites of Blučina and Hrušky (Droberjar 1997, 53, Abb. 37:2300 and 2338; 95 and 96).

8.2.7. Coarse-ware Vessels: Form III 53. Shoulder and body. Inv. No. 6-3-3. Ft. G1. Fig. 8.22. Fr. CIIIA. D. at shoulder is 15.5 cm. P.H. 16 cm. Th. 8 mm at shoulder and 5 mm at the body. W. 310 g. O. mottled dark grayish brown (10YR 4/2), brown (10YR 5/3) and black (10YR 2/1), I. very dark gray (10YR 4/2), O.M. yellowish red (5YR 4/6), I.M. and C. black (10YR 2/1).

50. Neck and body. Inv. No. 537-4-3/4. Ft. G4. Fig. 8.21. Fr. CIIB3. D. 30 cm. P.L. 12.5 cm. Th. 10 mm. W. 160 g. O. and O.M. reddish yellow (7.5YR 6/6), I. mottled light brown (7.5YR 6/4) and brown (7.5YR 5/3). I.M. and C. black.

Six sherds joined from the lower portion of a pot. Part of the shoulder is preserved above a conical body that tapers down to a flat bottom. The bottom had material accretions on the outside that will be analyzed. There is no decoration. The outer surface feels smooth and the texture of the fabric is hackly. There is a moderate amount of sand and mica inclusions, with sparse pebbles up to 3 mm in size.

The sherd is from a large dish/cooking pot with an inward-slanted neck and spherical body. There is an outturned lip. The junction of the neck and body has a raised border. The outer surface is rough and the texture is hackly. There is a moderate amount of sand and mica and a sparse amount of pebbles up to 2 mm in diameter. There are also holes and pot-marks suggesting some kind of organic temper was used that burned away.

A similar vessel in form, but twice the size and in better condition, was found by V. Varsik at Trnávka and dates to the second century AD (Varsik 2002, 130, Abb. 2; 135 and 148, Taf. IX:18).

One vessel from hut US82 at Urbárske Sedliská has a similar form (Turčan 1985, 105, Obr. 9:3). In Droberjar’s typology, 50 falls into the 2000 type: dishes.

54. Rim and neck. Inv. No. 495-4-3/4. Ft. G4 Fig. 8.22. Fr. CIIIA. D. 17 cm. P.L. 8.5 cm. Th. 8 mm. W. 35 g. O. dark grayish brown (10YR 4/2), I. brown (7.5YR 5/2), O.M. black (10YR 2/1), I.M. brown (7.5YR 5/2), C. dark yellowish brown (10YR 4/6).

51. Rim and neck. Inv. No. 73-3-3. Ft. G1. Fig. 8.21. Fr. CIIC (CIIIC). D. 17.5 cm. P.L. 4.5 cm. Th. 6 mm. W. 25 g. O. grayish brown (10YR 5/2), I. mottled grayish brown and light brownish gray (10YR 6/2), C. dark grayish brown (10YR 4/2).

Rim and neck sherd to a narrow-mouthed container. The rim is out-turned, the neck appears to have been cylindrical and based on the amount the rim is out-turned it is probable the vessel itself has a strong S-shape. The outer surface is rough and the texture is hackly. There is a moderate amount of sand and mica inclusions.

The upper portion of a tureen, with inward-slanting conical neck (straight, not curving) and a pronounced outturned lip. The outer surface feels rough and the texture is irregular. The inclusions include moderate amounts of mica and abundant sand, with sizes ranging up to 1 mm. The rim and form of the upper portion possibly match Droberjar’s broad category of 3600 (Tureens with conical necks) and the specific shape of 3602 (1997, 59). The match with 3602 is more likely because 51 exhibits a similar thinning of the neck wall as it reaches the junction with the body.

55. Rim and neck. Inv. No. 528-4-3/4. Ft. G4 Fig. 8.22. Fr. CIIIA. D. 20 cm. P.L. 12.5 cm. Th. 14 mm at the thickest and 8 mm remainder. W. 210 g. O. reddish yellow (7.5YR 6/6), I. brown (7.5YR 5/3), O.M. yellowish red (5YR 5/6), I.M. and C. red (2.5YR 4/8). Rim and neck sherd belong to a narrow-mouthed container. The rim is turned out, while the neck is conical in shape, with a very thick wall at the point where the neck turns out toward the rim. The preserved portion indicates the vessel had a strong S-shape. The outer surface is rough and the texture is irregular. There is a moderate amount of mica and sand, with a sparse amount of quartz pebbles up to 2 mm in diameter, while the mica can also be up to 1 mm in diameter. There are pits and hollows on the surface, possibly indicating where pebbles were once attached and have since fallen off.

52. Rim and neck. Inv. No. 87-3-3. Ft. G1. Fig. 8.21. Fr. CIIC (CIIIC). D. 14 cm. P.L. 7.5 cm. Th. 7.7 mm. W. 40 g. O. light brownish gray (10YR 6/2), I. dark reddish gray (2.5YR 4/1), C. very dark gray (10YR 3/1). The upper portion of a container, with inward-slanting conical neck (straight, not curving) and a pronounced outturned lip, similar to 51. The outer surface is rough while the texture of the fabric is hackly. Mica and sand inclusions are moderate with some sparse pebbles, reaching sizes of up to 6 mm in diameter.

In Droberjar’s typology the vessel fits into the 1200 series: S-shaped, keg form pots (Droberjar 1997, 44, Abb. 28: 1200).

For possible similar vessels see 51.

198

56. Rim, neck, and shoulder. Inv. No. 536-4-3/4. Ft. G4. Fig. 8.23. Fr. CIIIA. D. 13 cm. P.L. 18 cm. Th. 8 mm. W. 160 g. O. and I. pinkish gray (7.5YR 6/2), O.M. and I.M. reddish yellow (7.5YR 6/6), and C. gray (7.5YR 6/1).

is gently conical. The only differences are that the Křepice vessel is a fine-ware one and no evidence of having had handles. 58. Rim. Inv. No. 71-3-3. Ft. G1. Fig. 8.23. Fr. CIIIB. D. 23 cm. P.L. 5 cm. Th. 6 mm. W. 10 g. O. and I. light brown (7.5YR 6/4), C. red (2.5YR 5/6).

A large portion of a narrow-mouthed container/storage vessel, preserving the rim, neck and shoulder. The rim is out-turned with a lip, while the neck is cylindrical. At the join of neck and shoulder there is a depressed border. The shoulder is wide, while the body would have most likely been conical, with a flat bottom. The outer surface is rough and the texture is irregular. There is a moderate amount of sand and mica inclusions.

Rim piece to a bowl or cooking pot with a rounded lip turned inward. The neck of the vessel is not tall and joins to a shallow shoulder, which in turn joins most likely to a conical lower body. The outer surface has a harsh feel, the texture is irregular, and the fabric has abundant sand and mica, up to 1 mm in size. 59. Base and body. Inv. No. 951-4-3/4. Ft. G4. Figs. 8.24 and 7.41:59. Fr. CIIIB. B.D. 12.5 cm. P.H. 14.6 cm. Th. 8 mm at upper body, 10 mm at lower body and 13 mm at base. O. yellowish red (5YR 5/6), I. and C. gray (5YR 5/1).

The form of 56 fits in with Droberjar’s type 5200: Flaschengefässe, with a close parallel to form 5203, which differs, however, by having two raised lines forming the boundary to a depressed band along the shoulder of the vessel (Droberjar 1997, 62, Taf. 46:5203). The actual vessel on which type 5203 is based, however, is a fine-ware example from Hut Z-11 at Komořany, dating to the B2 period (Droberjar 1997, 265, Taf. 39:8). Vessel 56 may be similar to 57 in that its form is based on outside influence, such as from the Przeworsk culture.

Thirteen sherds joined, preserving the base and a large portion of the body of a container/cooking pot. The form of 59 was probably a strong S-shape based on the decoration on the outside and parallels with 60 below and vessels from other sites. The shoulder of the vessel has two rows of thumb impressions (Type XIIb) and the body has incised lines in a crisscross pattern (Type XIVc). The outer surface is rough and the texture is irregular. There is a moderate amount of mica and sand inclusions, some reaching a size of 3 mm.

57. Rim, neck and part of body. Inv. No. 953-4-3/4. Ft. G4. Fig. 8.23. Fr. CIIIA. D. 18 cm. P.L. 13 cm. Th. 8 mm. W. 160 g. O., I., O.M., I.M. (7.5YR 6/6), C. gray. A large portion of a wide-mouthed container/cooking pot, preserving the rim, neck, part of the body and one handle. The rim is out-turned and rounded. At the join of neck and body, near the bottom of the handle, is a depressed border (grooved line) along the vessel. There is also a very slight bulge at the join of neck and body. The body was gently conical. The handle is bent downward and thick, having a diameter of c. 2 cm. The outer surface of the vessel is rough and the texture is hackly. There is an abundant amount of sand and mica inclusions.

For a detailed discussion of the vessel type see 60 below. From the information below, see specifically vessels 11 and 12 from Rajhrad, found in Hut 1 and vessel 8 from Hut 3 at Mušov (Droberjar 1997, 354, Taf. 128:11 and 12; 369, Taf. 143:8). 60. Rim, neck and body. Inv. No. 954-4-3/4. Ft. G4. Figs. 8.25, 8.26, 8.33:60, 8.43:60, and 8.44:60. Fr. CIIIB. D. 27.2 cm. P.H. 21.5 cm. Th. 9 mm. O. and O.M. mottled light reddish brown (5YR 6/3) and gray (5YR 5/1), I. gray as well, I.M. and C. black.

57 is the only non-Roman pottery sherd with a handle found from the two huts. Vessel 57 may be an import from further north, such as Quadi settlements in Moravia or from the Przeworsk culture in northern Moravia and southern Poland6 based on its shape, handle, and presence of incised line; or at least vessel 57’s form may have been influenced by Przeworsk design. There are parallels, however, in handle design and vessel form found in German huts from Moravia, where there were Przeworsk settlements during the La Tène and B1 phase (Droberjar 1997, 135; 2002, 270). A vessel (04457-12/77) from Feature 1 at Horní Věstonice, dating to the B2 period, has similar handle design, along with a vessel (442-231/55) from Hut Z-2 at Komořany, dating to the B2/C1 period (Droberjar 1997, 216, 217, Taf 25:23, Taf. 33:13). The actual form of vessel 57 is similar to a vessel (483-27/61) from Hut 16 at Křepice, dating to the B2 period (Droberjar 1997, 222, Taf. 104:7). The Křepice vessel form also has an out-turned rim, straight neck, incised line at the border between neck and body, and a body that

Six sherds joined, preserving a large portion of a highly decorated container or cooking pot. The form is a strong S-shape with a very pronounced out-turned rim, conical neck, slight shoulder and a conical body. The base is not preserved, but most likely existed near to where the body is broken off. The outer surface is decorated with five rows of decorations. Immediately below the rim is a zigzagging line decoration (Type XV), followed by a row of thumb impressions (Type XIIb). Near the shoulder is a herring-bone pattern (Type XIVf); while on the shoulder are three rows of thumb impressions (Type XIIb). The conical body is decorated with incised lines in a crisscross pattern (Type IVc). It should be noted that some of the thumb impressions on the shoulder are so deep they cause bumps and protrusions on the interior of the vessel, almost punching through the wall itself. These bumps and protrusions may have weakened the integrity of the vessel itself around the shoulder. The wall 199

thickness is a relatively uniform 9 mm along the entire vessel, but within the thumb impressions on the shoulder it is estimated that wall thickness could be as little as 5 mm or even less. The outer surface is rough and the texture is irregular. There is a moderate amount of mica and sand inclusions, some reaching 1 mm in size.

similar design is also seen at Dúbravka, from hut 12/88 (Elschek 1995, 47, Abb. 3:7).

The S-shape of the vessel is a common shape both in southwest Slovakia and Moravia, but the amount of decorations present on the exterior is almost unique. Other vessels of this form with multiple decorative patterns have been found, but usually no more than two types of decoration present (usually thumb impressions and the incised line in crisscross pattern), with rare examples of three types. A similar vessel was found by V. Turčan at Báhoň, dating to the second half of the second century, having incised lines on the rim, incised lines in a crisscross pattern on the body, while the shoulder had a row of thumb impressions, a zigzag line, then two rows of thumb impressions (Bartík and Turčan 1990, 78, Obr. 5:16; 79 and 82–83). In regards to Droberjar’s typology 60 fits into the 2100 type, S-shaped formed pot-dishes, and more specifically type 2111 (Droberjar 1997, 48, Abb. 32:2111). Hut 1 at Blučina, dating to B2/C1 (AD 150/160–180/200), had an S-shaped vessel with four rows of thumb impressions on the shoulder and one row on the neck, while having crisscrossed incised lines on the body and the rim has incised lines (Droberjar 1997, 228, Taf. 2:1). Hut VI at Blučina, also dating to the same period, contained a sherd of another S-shaped vessel with the herring-bone pattern on the shoulder similar to 60 (Droberjar 1997, 242, Taf. 16:3). Two vessels from Rajhrad, found in Hut 1 (dating to the B2 period), both have S-shape forms with three rows of thumb impressions on the shoulders and incised line patterns on the body, one definitely being crisscrossed (Droberjar 1997, 354, Taf. 128:11 and 12). Another example is a vessel from Hut 3 at Mušov dating to the C1 period (Droberjar 1997, 369, Taf. 143:8). The vessel has a strong S-shape and once again has three rows of thumb impressions on the shoulder and the crisscross pattern on the body. It is difficult to date 60, since there is no exact copy of the vessel, but similar ones date from the B1C to C1 periods, AD 50/70– 250/260.

62. Base. Inv. No. 61-3-3. Ft. G1. Fig. 8.27. Fr. Un. B.D. 13.8 cm. P.H. 2.1 cm. Th. 9 mm at body and 8 mm at base. W. 18 g. O. brownish yellow (10YR 6/6), I. and C. gray (10YR 6/1).

8.2.8. Coarse-Ware Vessels: Bases and Unidentifiable Rims

Sherd preserving a portion of the base and body. The outer surface is rough and the texture is hackly. There is a moderate amount of mica and sand inclusions and a sparse amount of quartz pebbles up to 1 mm in diameter. 63. Base and body. Inv. No. 68-3-3. Ft. G1. Fig. 8.27 Fr. Un. B.D. 15 cm. P.H. 4.7 cm. Th. 12 mm at body and 11 mm at base. W. 65 g. O. dark gray (10YR 4/1), I. very dark grayish brown (10YR 3/2), C. grayish brown (10YR 5/2). Sherd with portion of base and body preserved. The outer surface is decorated with thumb impressions (Type XIIc) and is rough, while the texture is irregular. There is a moderate amount of sand and mica inclusions with a sparse amount of quartz pebbles up to 1 mm in diameter. 64. Base and body. Inv. No. 582-4-3/4. Ft. G4. Figs. 8.27 and Fig. 8.38:123. Fr. Un. B.D. 9 cm. P.H. 8.3 cm. Th. 9 mm. W. 130 g. O. mottled brown (7.5YR 5/3) and light red (2.5YR 6/6), I. light brown (7.5YR 6/3), C. gray (7.5YR 5/1). Sherd preserving a portion of the conical body and base. The outer surface is rough and the texture is irregular. There is a moderate amount of sand and mica inclusions. Body has comb-line decorations creating arcs randomly placed (Type XIIIi). 65. Base and body. Inv. No. 665-4-3/4. Ft. G4. Fig. 8.28. Fr. Un. (CIIIA ?). B.D. 12 cm. P.H. 9.5 cm. Th. 6 mm upper body, 10 mm at lower body and base. W. 130 g. O. mottled light reddish brown (7.5YR 6/4), gray (7.5YR 6/1), and reddish brown (7.5YR 5/4), I. mottled light reddish brown, gray, and black, O.M. reddish yellow (5YR 6/6), I.M. light brown (7.5YR 6/4), C. brown (7.5YR 5/2).

61. Rim and neck. Inv. No. 86-3-3. Ft. G1. Fig. 8.27. Fr. CIIIC. D. 16 cm. P.L. 8.5 cm. Th. 7 mm. W. 50 g. O. and I. dark gray (10YR 4/1), C. very dark gray (10YR 3/1). The upper portion of what was probably a narrowmouthed vessel. The out-turned lip is attached to a conical neck that slopes inward. The lower portion of the neck is broken off before reaching the body, which probably had a shallow shoulder leading to a conical lower body. The outer surface feels rough and texture is hackly. Inclusions include an abundance of mica, sand, and small rocks, up to ½ mm in size.

Sherd preserving the conical body and a small portion of the base of a container or cooking pot. There is a moderate amount of sand and mica inclusions with a sparse amount of quartz pebbles. Some of the mica can be up to 3 mm in diameter while the quartz pebbles are up to 5 mm in diameter. The vessel was probably an S-shaped pot, similar in design to 58 and 59. Please see the discussions of those two vessels for a detailed description of the form and examples from other sites.

The surviving neck portion seems to point to the vessel matching Droberjar’s type 3600, sub-type 3601, a tureen with a conical neck (Droberjar 1997, 59). A vessel of 200

66. Base and body. Inv. No. 949-4-3/4. Ft. G4. Fig. 8.28. Fr. Un. B.D. 9.5 cm. P.H. 4.2 cm. Th. 6 mm at upper body, 5 mm at lower body, and base 8 mm. W. 260 g. O. mottled black, reddish yellow (7.5YR 6/6), and light brown (7.5YR 6/4), I. and C. black.

grayish brown and brownish yellow (10YR 6/6), C. very dark gray (10YR 3/1). Rim sherd with a small portion of the neck. Too little survives to be sure of the original vessel shape. The rim is out-turned with a flattened edge. The outer surface is smooth and the texture is fine. There is a sparse amount of mica and sand inclusions.

Two sherds joined, preserving the base and portion of the body of what was most likely a cooking vessel. The base has a concave center on the bottom with a corresponding slight rise of the interior portion of the base. There is no outer decoration but there are black areas resulting from its use as a cooking vessel or during the firing of the vessel, with other vessels pressed up against it in the kiln. The outer surface is rough and the texture is irregular. There is a moderate amount of sand and mica inclusions.

71. Rim. Inv. No. 205-3. Ft. G2. Fig. 8.29. Fr. Un. D. 15 cm. P.L. 4.5 cm. Th. 4.5 mm. W. 5 g. O. and I. brown (7.5YR 5/2), C. black. Rim sherd preserving a portion of the body of a possible dish or cylindrical cup. The rim has a lip and the body is funnel shaped. The outer surface is rough and the texture is irregular. There is a moderate amount of mica and sand inclusions.

67. Base and body. Inv. No. 952-4-3/4. Ft. G4. Figs. 8.28 and 8.42:67. Fr. Un. (CIIIA ?). B.D. 10 cm. Th. 8 mm at body and 12 mm at base. W. 270 g. O. and I. grayish brown (7.5YR 5/2), C. gray (7.5YR 5/1).

72. Rim. Inv. No. 225-3. Ft. G2. Fig. 8.29 and 8.44:72. Fr. Un. D. 24.2 cm. P.L. 3 cm. Th. 10 mm. O. and I. very dark gray (10YR 3/1), C. black.

Five sherds joined, base with part of the conical body. Body is decorated with incised lines in a criss-cross pattern (Type XIVd). The outer surface is rough and the texture is hackly. There are mica and sand inclusions in moderate amounts and sparse amounts of quartz pebbles up to 1 mm in diameter.

Rim sherd preserving little of the neck. The rim has incised lines repeating across it as decoration. The outer surface is rough and the texture is irregular. There is a moderate amount of mica and sand inclusions with some reaching 4 mm in diameter.

The vessel form was probably a S-shaped pot, similar to 58 and 59 based on what vessels the criss-crossed incised line pattern appears.

73. Rim. Inv. No. 494-3-3/4. Ft. G4. Fig. 8.29. Fr. Un. (CIIB ?). D. 17 cm. P.L. 7 cm. Th. 10 mm. W. 60 g. O. brown (7.5YR 5/2), I. and C. black (7.5YR 2.5/1).

68. Rim. Inv. No. 85-3-3. Ft. G1. Fig. 8.29. Fr. Un. D. 25 cm. P.L. 8 cm. Th. 8 mm. W. 55 g. O. gray (10YR 6/1), I. dark gray (10YR 4/1), C. dark gray (Gley 4/N).

Sherd preserving the rim and neck of a storage vessel/cooking pot, but not enough remains for precise identification. The neck is slanted inward, but is then turned out at the rim. The outer surface is smooth and the texture is irregular. There a moderate amount of mica inclusions and a sparse amount of quartz pebbles, some up to 3 mm.

Rim piece of a probable cooking pot. The rim is turned outward with no lip. On the lower portion of the sherd’s outer surface are comb decorations applied in a chaotic manner (Type XIIIb). The outer surface feels rough the texture of the fabric is irregular. There is a moderate amount of mica and sand inclusions, with a sparse amount of small pebbles, no larger than 2 mm in diameter.

74. Rim. Inv. No. 506-4-3/4. Ft. G4. Fig. 8.30. Fr. Un. D. 19 cm. P.L. 6.6 cm. Th. varies between 8 and 10 mm. W. 40 g. O. brown (7.5YR 5/4), I. black, C. very dark gray (7.5YR 3/1). Rim sherd preserving part of the body with a row of lightly pressed thumb impressions (Type XIIh) on the outer surface. The rim is slanted inward and rounded. The outer surface is rough and the texture is irregular. There are mica and sand inclusions in moderate amounts.

Because of the complete absence of the body, a proper identification cannot be made. The rim, however, probably comes from either a Form II or III vessel. 69. Rim. Inv. 203-3. Ft. G2. Fig. 8.29. Fr. Un. D. 22.2 cm. P.L. 5.4 cm. Th. 7 mm. O. and I. grayish brown (10YR 5/2), C. very dark gray (10YR 3/1).

75. Rim. Inv. No. 546-4-3/4. Ft. G4. Fig. 8.30. Fr. Un. (CIIB or CIIIA ?). D. 15 cm. P.L. 7 cm. Th. 6 mm. W. 20 g. O., I., O.M., and I.M. pale brown (10YR 6/3), C. dark grayish brown (10YR 4/2).

Rim sherd preserving a portion of the body. The rim is turned out and has a lip, while the body is funnel shaped. There is a moderate amount of mica and sand inclusions.

Sherd preserving the rim and neck of a storage vessel/cooking pot. The neck is conical with the rim bent outward at a 45 degree angle. There is a rounding to the rim on the downside portion, creating something of a lip.

70. Rim. Inv. No. 204-3. Ft. G2. Fig. 8.29. Fr. Un. D. 15 cm. P.L. 3.1 cm. Th. 4 mm at body and 6 mm at rim. O. grayish brown (10YR 5/2), I. mottled 201

76. Rim. Inv. No. 547-4-3/4. Ft. G4. Fig. 8.30. Fr. Un. (CIIA or CIIIA ?). D. 15 cm. P.L. 8.7 cm. Th. 9 mm. W. 40 g. O. and I. mottled pale brown (10YR 6/3) and very pale brown (10YR 7/3), C. gray (10YR 5/1).

8.3. Catalog of the Types of Ornamentation V. Varsik provides a detailed description of the various types of ornamentation seen on German pottery at Bartislava-Trnávka (Varsik 2002, 134–138). He has 14 main types with variations as sub-types. These types by V. Varsik can also be applied to the ornamentation seen on pottery at Urbárske Sedliská. There are differences, however, with new types and sub-types appearing at Urbárske Sedliská, while some of Varsik’s types do not appear at all. I have added three new types and many subtypes where needed (Figs. 8.31–44). Even though all of Varsik’s 14 types are not seen at Urbárske Sedliská, a description will be given below for each so as to provide a comprehensive list of what can be found in southwest Slovakia. E. Droberjar has also created an ornament typology for Moravia, with nine categories, each subdivided by variants (Droberjar 1997, 29). Droberjar’s typology will also be mentioned to provide a broad view of the types of ornaments found on German pottery. Table 8.5 provides a quick reference guide to the main decoration types described below.

Sherd with out-turned rim and added lip on a high neck. The vessel was probably a cooking pot or storage vessel. The outer surface is rough and the texture is irregular. There is a moderate amount of mica, sparse amount of quartz pebbles and limestone (up to 8 mm in size). 77. Rim. Inv. No. 553-4-3/4. Ft. G4. Fig. 8.30. Fr. Un. (CIA ?). D. 12 cm. P.L. 9 cm. Th. 8 mm. W. 50 g. O., I., and O.M. reddish brown (5YR 5/3), I.M. and C. black. Rim sherd of a possible beaker/cup, preserving part of the neck/body with a row of thumb-impressions (Type XIIf). The rim is rounded and the outer surface below the rim is poorly preserved, having flaked off some time in the past. There is also what appears to be a notch in the sherd on the rim. The outer surface is rough and the texture is hackly. There is a moderate amount of sand, mica and quartz pebbles, some reaching up to 4 mm in diameter.

8.3.1. Type I. Rolled-Wheel Patterns (Droberjar Type 600) This type of decoration uses a small spiked wheel rolled across the surface of the vessel to create decorative patterns (Fig. 8.31:I,a–d). Various patterns can be found, such as zigzag, repeating lines, and more complex patterns, such as a panel with lines repeating in threes that create triangles framed by lines above and below the panel. Type I decorations are found in very small numbers in southwest Slovakia, apparently not having the popularity of other types (Varsik 2002, 134). The dates of use for rolled-wheel patterns are based on one example dating to the first century AD at Dúbravka, while three examples were found dating to the second century at Trnávka, and still others dating to the third century (Elshcek 1995, 40, Abb 4:5; Varsik 2002, 134). The small wheel is also used in Moravia over a similar period of time to create various motifs (Droberjar 1997, 66–68). From the examples seen in Moravia, it would appear that this is primarily a fine-ware decoration. No Type I decorative patterns were found at Urbárske Sedliská.

Table 8.5. The 17 Decoration Types Type Description Droberjar’s Types Rolled-Wheel 600 I Patterns Dotting By 150 II Small Wheel Filled Triangles 240 III Incised Lines IV with Dots Dimples 570 V Fine Punctures 130,140, 160, VI 190 Barbotine 520 VII Grooved 200 VIII Lines/Furrows Fluted/Impressed 550 and 560 IX Lines Roughening X Wedges 170 XI Nail/Finger 110, 160 and XII Impressions 170 Comb Lines 330, 350–370 XIII and 390 Incised Lines 200–220 XIV Finger Fluting XV Rim Incising XVI Combination 920–980 XVII Patterns

8.3.2. Type II. Dotting by Small Wheel (Droberjar Type 150) The decoration is basically made up of rows of small dots, much like those seen with the Type I decoration. It is thought, therefore, that these rows are also made with small wheels (Varsik 2002, 135) (Fig. 8.31:II,a–c). Examples have been found both at Dúbravka and Bratislava-Trnávka (Elschek 1995, 40, Abb. 4:10; Varsik 2002, 135). Droberjar has four sub-types, with Type 153 being the closest in design to Varsik’s Type II (Droberjar 1997, 32, Abb. 17:153a and 153b). These two sub-types, 153a and b, date to the B2 and B2/C1 periods, being found in Hut I and IV at Křepice and Hut VI at Blučina 202

(Droberjar 1997, 71). As with Type I, this decoration also appears to be a fine-ware decoration. No examples were found at Urbárske Sedliská.

8.3.4. Type IV. Incised Lines with Dots This type of decoration has a frieze made up of a zigzagging line that has a dot at the apex of each line (or it can be described as a series of triangles with dots at the corners of each triangle) (Fig. 8.31:IV,a). Examples were found by Varsik at Bratislava-Trnávka, dating to the second century (Varsik 2002, 135). Droberjar does not have this type of decoration included in his typology as a specific sub-type, but it would fit into the broad 200 type: line ornaments, or 920 type: different combined ornaments. This type does not appear at Urbárske Sedliská.

8.3.3. Type III. Filled Triangles (Droberjar Type 240) The decoration has an incised line running along the shoulder from which triangles are drawn hanging down upon the body of the vessel; each triangle has shallow circular depressions (dimples) filling up the interior space (Fig. 8.31:III,a–c). Sometimes stray depressions can be found outside the triangles. There are three forms of this decoration, with two forms (IIIa and b) dating to the second century AD, while the third (IIIc) dates to the third century (Varsik 2002, 135). The first form has triangles that are better made, subtly attached to the shoulder line, and that usually appear on highly polished fine ware (Varsik 2002, 135). The second form is similar but the triangles’ attachment to the shoulder line is poorly executed and the overall appearance is rougher (Varsik 2002, 135). The third is much like the second in quality of execution but on a larger scale. Type III decorations seem to appear only on fine-ware vessels.

8.3.5. Type V. Dimples (Droberjar Type 570). This type of decoration has circular depressions larger than the dots made by the small wheels and that have the appearance of dimples (Fig. 8.31:V,a and b). These dimples can be in rows (Va) or placed randomly on the surface of a vessel (Vb). It is possible that a punch was used to create the dimples, which are usually equal in shape, size, and depth. Varsik dates this type of decoration to the second century, based on the evidence available from various sites (Varsik 2002, 135). The second century date also applies to examples found in Moravia (Droberjar 1997, 85, 86, 185, 187; Varsik 2002, 135). Type V decorations seem to have been used only on fine-ware vessels.

One example that possibly fits into the first form (IIIa) was found in Hut 1 at Urbárske Sedliská on sherd 78, from a polished fine-ware vessel (Fig. 8.31:78). The shoulder line is made up of a center raised line flanked on either side by incised lines, from which hang the triangles. The triangles do not have incised lines defining their shape, but instead the depressed circles themselves are formed into triangles. These depressions were probably made by small toothed wheels similar to those used in Type I and II decorations. The absence of lines defining the triangles marks sherd 78 as different from the types put forward by Varsik and also for Droberjar’s type 240. The decoration on the sherd has similarities to Droberjar’s type 154, which is a dotted line with hanging triangles made of dotted lines, but with their interiors empty (Droberjar 1997, 32, Abb. 17:154). Type 154 was found on a vessel from the German settlement at Hradčovice and dates to the B1 period (Droberjar 1995, 25, Abb. 6:6). It would seem that the motif on sherd 78 is an intermediary or cross between decoration type 154 and 240 (Varsik’s IIIa), and so may date to the early part of the B2 period or very late B1. There is a vessel with triangles made of dots from central Germany dating to the first/second century AD (Berlekamp 1988, 412, Taf. 63:a). The vessel comes from the site of Nordhausen, near Erfurt and is of the Rheinweser-Germanen culture. This vessel has triangles covering the body of the vessel, both hanging from the shoulder and coming up from the base, and they are not outlined, but are defined by the dots. This vessel can be seen as another early example of this type of decoration; it possibly influenced development among the Marcomanni and spread to the Quadi.

One example at Urbárske Sedliská comes from Hut 1. Vessel 1 has dimples on the surface placed in a random manner (Figs. 8.9:1 and 8.31:1). This vessel has already been mentioned as imitating a Roman glass vessel or beaker and dating to the B2 period. 8.3.6. Type VI. Fine Punctures (Droberjar Types 130, 140, 160 and 190) There are two types of fine punctures/pricks which I list here, while V. Varsik only lists the first type (Varsik 2002, 136) (Fig. 8.31:VI,a). The first (Varsik’s VIa) is a triangular shape pressed into the clay by a punch, while the second (VIb) can have a circular, narrow oblong oval or other type of vaguely ovular shape (possibly done with a stylus). The VIb type is different from the above incised dots and dimples by not being always circular and by the size and depth of the punches. These punctures/pricks seen in VIb need not have been made by a punch, but instead some may have been done by using the tip of a stylus. I have added this sub-type here because it is different not only from decoration Types I–V, but also from Type XI: Wedges and XII: Thumb Impressions. Both VIa and VIb appear in rows, or randomly, on the outer surface of the vessel, usually on the body. Only one example of Type VIa was found by V. Varsik at Trnávka, but more have been found in Moravia (Varsik 2002, 136). The examples from Moravia have been dated to the second half of the second century AD (Varsik 2002, 136). From hut US82 at Urbárske Sedliská, two examples were 203

found dating to the middle of the second century AD (Turčan 1985, 114). One sherd had small oval punches placed randomly and another had three vertical rows of oval punches (Turčan 1985, Obr. 4:7 and 3 respectively).

8.3.8. Type VIII. Grooved Lines/Furrows (Droberjar Type 200) These lines or furrows appear on the body of a vessel and Varsik identifies two versions (Fig. 8.32:VIII,a and b). The first, VIIIa, has the furrows going down the body at an angle from a base line (Varsik 2002, Taf. IV:VIIIa), while the second has a center line from which the grooved lines extend perpendicularly, which he calls grill-hatching (Varsik 2002, Taf. IV: VIIIb). Even though the lines are closely spaced, they are different from the comb-line decoration seen in Type XIII (Figs. 8.37–7.40). These grooved lines seem to have appeared only on fineware vessels. There are few examples of Type VIII decoration in Slovakia and the dates are not exact, falling into the second to third centuries (Varsik 2002, 136). Droberjar does not have a specific sub-type identified for these types of decorations but they do fall under the broad category 200: line ornaments. Type VIII decoration was not found at Urbárske Sedliská.

Four sherds were found with Type VI decoration, with a Type VIb from Hut 1 and one VIa and two of VIb from Hut 2. The VIa decorated sherd 79 from Hut 2, is from a coarse-ware vessel (Fig. 8.31:79). The sherd is very small and only two triangular punches survive in what appears to have been a row. In regards to the VIb type, 80 from Hut 1 is from a fine-ware vessel (Fig. 8.31:80). There are three half-circle punches at an angle on the body-sherd, seemingly placed at random because there are no other groupings on the sherd. From Hut 2, fine ware sherd 81 has three rows of oval punches preserved (Fig. 8.31:81). The second sherd, 82, from a coarse-ware vessel, has three offset rows of half-circle punches (Fig. 8.31:82). 8.3.7. Type VII. Barbotine (Droberjar Type 520) The decoration has raised pieces of clay tightly spaced on the outer surface of the vessel, usually on the belly (Fig. 8.32:VII,a). These “spikes” of clay have various shapes, such as pyramidal or rounded protrusions, and may also be combined with other types of decorations. Barbotine starts in the second century, appearing at almost all settlements dated to that period, and continues into the B2/C1 phase, with none having been found from the C1 period in Slovakia (Varsik 2002, 136). Varsik found five sherds with barbotine at Zadné (part of the find-complex at Bratislava-Trnávka), including a small cup, dating to the second century (Varsik 2002, 136, Abb. 2:30 and Taf. VIII: 14). This type of decoration rarely appears on ceramics from graves, but is common on settlement ceramics (Varsik 2002, 136). Only three examples have been found from graves, all dating to the early part of the B2 phase: one from a grave at the cemetery at Abrahám and two from a grave at Sládkovičovo (Kolník 1980, 68, 69, Taf. LIV: 194b, c; 138, Taf. CXXXVI: 28a, c1, c2; Varsik 2002, 136). Type VII decoration appears only on fine- ware vessels.

8.3.9. Type IX. Fluted/Impressed Lines (Droberjar Types 550 and 560) The decoration has lines placed on the body of the vessel that can be described as fluted (Fig. 8.32:IX,a–d). These lines are deeper and usually wider than Types VIII and XIV and some appear to have been created differently from incising. V. Varsik identified four sub-types (IXa– d). Varsik mentions that small fine-ware cups and dishes that have radiating flutes from the base (Type IXa) are evidence of Przeworsk influence and date to the second half of the second century AD based on their occurrence in graves (Varsik 2002, 137). These flutes appear to go diagonally on the body of the vessels. Similar to Type IXa are vessels with U-shaped arcs (Type IXb) covering the lower part of dishes. Examples from the settlement at Křižanovice in Moravia date to the B2 and B2/C1 period (Varsik 2002, 137). The third type (IXc) has flutes running parallel to the axis of the vessels, closely spaced, while the fourth type (IXd) has the flutes placed diagonally on the body of the vessel, but do not run to the base of the vessels setting this type as different from Type IXa (Varsik 2002, 137). These last two types, IXc and d, are seen as having a modest beginning in the second century AD, but expand in popularity by the third century during the C1 period (Varsik 2002, 137). One example was found in Hut US82 by V. Turčan (Turčan 1985, 100, Obr. 4:15). The sherd is from a fine-ware vessel and has four preserved flutes in the style of Type IXc.

Two examples of barbotine were recovered from Hut 2. Both are body-sherds and come from fine-ware vessels. The first, no. 83, is a small sherd with three of the clay spikes preserved (Fig. 8.32:83). The barbotine on sherd 83 is similar to that on a sherd found in hut US82 (Turčan 1985, 100, Obr. 5:11). From Moravia, in Hut 1 at Blučina and dating to B2/C1, a portion of a dish also has a similar type of barbotine; while a dish from hut Z-3 at Komořany, dating to B2, also has this type of pronounced barbotine (Droberjar 1997, Taf. 5:21). The second example, no. 84, has a bumpy surface, but not as pronounced as seen with sherd 83 (Fig. 8.32:84). This type of less pronounced barbotine can be seen also on a sherd from Hut 19 at Bratislava-Trnávka (Varsik 2002, Taf. VII:4). From Moravia, a similar type can be seen on a sherd from Blučina, Hut 3, dating to B2/C1 and from Feature 16 at Blučina, dating to B2 (Droberjar 1997, Taf. 9:12 and Taf. 22:4).

Three fine-ware sherds were found with fluting at Urbárske Sedliská, with one from Hut 1 and two from Hut 2. Sherd 86 from Hut 1 has two fluted lines fitting into Type IXc (Fig. 8.32:86). Sherd 2 from Hut 2 is a small rim-sherd with one fluted line (Fig. 8.9:2 and 8.32:2). The placement of the flute near the rim may signify a small cup or dish, and the flute’s design indicates the decoration is Type IXc. The third sherd, 85, from Hut 2 is a larger body sherd with three flutes 204

preserved. One flute’s end is also preserved, while the others continue off the edges of the sherd, indicating that they were placed diagonally on the body, but it appears to have been Type IXc, based on the size and shape of the flutes.

There are different ways in which the impressions were created. An early method, dating to the first century AD and found at Dúbravka, was to use two fingers and pinch the clay, creating two impressions separated by a mound of clay in the middle (Elschek 1995, 40, 41, Abb. 5:13 and 17; Varsik 2002, 138). The impressions from Urbárske Sedliská are from the more usual method of using one finger to create the impressions. From experimenting in the lab (holding vessels and trying various fingers on the surface), I believe it is most likely that the thumb was the primary digit for creating the impressions. Many of the thumb impressions also had a line bisecting them, as if the potter had an overly long thumbnail. The impressions with the added line were usually not intermixed at random with non-lined thumb impressions and so would appear to have been done for effect. It is difficult to say, however, whether the line was intentionally meant to be there or it was a natural byproduct of the potter’s having actually a fingernail long enough to matter. The impressions were often the sole decorative element on vessels, but there are also many examples of their being used in conjunction with incised lines, such as Type XIVb: Incised lines in a crisscross pattern.

8.3.10. Type X. Roughening These vessels are undecorated coarse-ware vessels with a portion of the body purposefully left or made rougher than other parts of the vessels (Fig. 8.33:Xa). The decoration is commonly seen in the first century AD and rapidly fades in the second century (Varsik 2002, 137). No clear examples were found at Urbárske Sedliská. 8.3.11. Type XI. Wedges (Droberjar Type 170) This decorative type has triangular wedge shapes pressed into the body of vessels (Fig. 8.33:XIa–f). The wedges were placed randomly (XIa), in rows (XIb, c, f), or in a herring-bone pattern (XId and e). These wedges are different in shape from the triangles seen in Type VI decoration, with the Type XI triangles being isosceles and Type VI being equilateral. Types XId and XIe differ from each other in the size of the wedges forming the herring bone pattern, with XIe having thinner, more delicate wedges. Type XI decoration appears more commonly in the second century, gaining in frequency as the century progresses (Varsik 2002, 137). Examples of XId in Slovakia have been found dating to the first half of the second century AD, and the type lasts with variations into the third and fourth centuries (Varsik 2002, 137). The examples of this decorative type in Moravia span a similar time period as in Slovakia (Droberjar 1997, 66– 67, 140). No examples were found at Urbárske Sedliská.

Varsik created six subtypes in his decorative typology, but with the examples from Urbárske Sedliská I have increased the total to fourteen subtypes. For greater clarity, instead of listing the new subtypes after Varsik’s, thereby separating similar types, I have rearranged them so that similar ones are placed near each other. In parentheses after the subtype I have placed Varsik’s subtypes, if applicable, for easier reference to his typology; I have also cited Droberjar’s where possible. The differences between the various subtypes are based on the depth of the impressions, the treatment of the mounds, and the placement of the impressions on the body of the vessel.

8.3.12. Type XII. Nail/Finger Impressions (Droberjar Types 110, 160 and 170)

The first type, XIIa, has the pinched impressions mentioned above (Fig. 8.33:87), and was recognized at Dúbravka in a first century AD context. Two fingers are pushed in and then together to create a pair of impressions sharing a pinched mound, usually in the shape of an hourglass. The pairs of impressions are then repeated over the body of the vessel, but also appear as rows, closely packed and overlapping. This subtype continues into the second century; there is an example from Trnávka (Varsik 138, Abb. V:11) and another example from hut US82 (cat. no. AP32382). There are possibly three examples of this type at Urbárske Sedliská. Sherd 88 from Feature G2 has what appear to be two sets of pinched impressions, while a sherd (Inv. 603-4-3/4G4) from Hut 2 has closely spaced impressions with the intervening mounds appearing to have been pinched. The third example, 89, comes from Feature G3 and has a pair of pinched impressions (Figs. 8.33:88, 89). One other possible example may be vessel 60 from Hut 2 (Fig. 8.26). 60’s row of thumb impressions on the neck has some impressions with intervening mounds appearing pinched rather than built up from finger pressure. These pinched mounds, however, could be the result of the

The decoration in its simplest form was created by the potter pressing a finger into the body of the vessel, thereby leaving a depression and mound of clay to one side; the combination is then repeated either in rows or at random (Figs. 8.33:XIIa and b; 8.34–7.36). There are many variations in the appearance of the impressions and mounds, depending in part on how much of the finger and fingernail was used. The presence of this type of decoration is common on sites all across southwest Slovakia and Moravia (Droberjar 2002, 120; Varsik 2002, 137). This decorative type seems to be primarily used on coarse-ware vessels. From all the examples found at Urbárske Sedliská, only a small amount of fine-ware sherds had finger impressions and these were of the specialized sub-types XIIl–n (see below). No examples of the very common finger impressions, such as sub-types XIIb–i, were found on fine-ware sherds.

205

potter’s using his thumb to make the initial impression and mound, but then modifying the decoration by reforming the mounds if too large or deformed.

examples were found in either of the Huts or features, but one example was found in Hut US82 (cat. no. AP32518). Type XIIk (Varsik Type XIIf) has thumb impressions in rows connected by incised lines (Fig. 8.36:XIIk). No examples were found at Urbárske Sedliská.

The second type, XIIb (Varsik XIIa; Droberjar 112a–c), has deep impressions in tightly packed rows; the mounds are curtailed because of the close placement of the impressions (Fig. 8.33:60, 90, 91). The impressions are usually circular or oval in shape. These sometimes have mounds in an hourglass shape, if the thumb was pushed in and then twisted back toward the mound of the previous impression. Thirteen examples of this type were found, with the majority from Hut 2 (seven total, including vessels 59 and 60, the latter already mentioned above as a possible example of Type XIIa), four from Hut 1 and one from Feature G3.

Type XIIl (Droberjar 160) has the impressions in rows, but they are created by pressing in only the very tip of the finger or thumb instead of pressing in more deeply (Fig. 8.36:106). This type of shallow impression creates very narrow, half circle or crescent forms with little or no mound on either side. They are different from the shallow impressions seen in Types XIIf–i by being smaller and narrower. Type XIIl also seems to rely more on the fingernail to help create the narrow, sharply defined shape on some examples from fine ware vessels. A total of six examples were found: four from Hut 1, one from Feature G3, and one from Hut 2.

Type XIIc (Varsik XIId; Droberjar 111) has deep impressions tightly packed, but placed randomly on the vessel’s body (Fig. 8.34:92 and 93). There is a total of eight examples: three from Hut 1, four from feature G2, and one from Hut 2.

Type XIIm (Droberjar 170) has the same half-circle impressions as in Type XIIl, but placed in a herring bone pattern (Fig. 8.36:107–110). A total of eight examples were found: seven from Hut 1 and one from Hut 2.

Type XIId (Droberjar 112d) has deep impressions widely spaced in rows (Fig. 8.34:94, 95). A total of five sherds were found with this type: four from Hut 1 and one from Feature G2.

Type XIIn has half-circle impressions in rows with very controlled mounds on one side (Fig. 8.36:111). This decoration appears as if the finger or thumb was pressed in and then scraped to one side to create a very controlled mound of clay (which creates a different effect from the usual impression and having the mound naturally form to one side). Only two examples were found, one from each hut.

Type XIIe has deep impressions widely spaced randomly on the body (Fig. 8.34:96). Seven sherds were found with this type of decoration: three from Hut 1 and four from Hut 2. Type XIIf (Varsik XIIc) has shallow impressions in tightly packed rows, and the mounds are small or nonexistent (Fig. 8.35:97 and 98). The shape of the impressions can still by circular or oval, but often the shape is a half-circle. Ten sherds with this decoration were found: one from Hut 1, two from Feature G2, three from Feature G3, and four from Hut 2.

8.3.13. Type XIII. Comb Lines (Droberjar Types 330, 350, 360, 370 and 390) The comb-line decoration is another common decorative type seen on German pottery in the central Danube area. The type is seen frequently in the first and second centuries AD, but then disappears in the third century (Varsik 2002 138). Much as with Type XII, there are many variations of the comb-line motif. Varsik created ten subtypes, and, based on the examples from Urbárske Sedliská, I have created an additional eight. As with Type XII, I have arranged the subtypes so that similar ones are near each other; I have also placed those seen on fineware vessels together (listed first – Types XIIIa–d), separate from those seen on coarse wares. The major difference between comb lines on fine-ware and coarseware vessels is the execution of the lines. Fine-ware versions were cleanly applied in a controlled manner to create well defined decorations and the comb itself had finer-sized bristles. The coarse-ware comb lines were applied roughly and have the appearance of having been made hastily. The comb used on coarse-ware vessels appears to have had much thicker bristles. Varsik’s type number will be given in parentheses along with the corresponding Droberjar type.

Type XIIg has shallow impressions tightly packed and randomly placed (Fig. 8.35:99). Two sherds were found with this decoration in Hut 2. Type XIIh has shallow impressions widely spaced in rows (Fig. 8.35:100–102). A total of ten sherds was found with this type of decoration: two from Hut 1, one from Feature G3, and seven from Hut 2. Type XIIi (Varsik Type XIIe) has shallow impressions widely and randomly spaced on the vessel (Fig. 8.36:103–105). Twenty sherds were found with this decoration type: nine from Hut 1, three from Feature G3, and eight from Hut 2. Type XIIj (Varsik Type XIIb) has a pattern on the body of the vessel with one or more rows of impressions going horizontally across the shoulder of the vessel from which then descend vertical rows of impressions, in pairs, with empty spaces between them (Fig. 8.36:XIIj). No

206

Type XIIIa (Varsik XIIIc; Droberjar 353 and 356) uses comb lines to create triangular shapes formed by the lines themselves (Fig. 8.37:112). The decoration usually appears in rows and is used on fine ware vessels. One example was found of this type on a fine-ware sherd from Hut 1.

Type XIIIj (Varsik XIIId; Droberjar 357) has comb lines that create arches instead of half-circles, because the lines are dragged down and away from the apex of the curved lines (Fig. 8.38:124). Type XIIIj has arches that appear separately in rows. No examples were found at the site. Type XIIIk has the same arches seen in XIIIj but joined to create wavy lines, sometimes in rows one on top of another (Fig. 8.38:125 and 126). Four examples were found at the site: two from Hut 1, one from G2 and one from G3.

Type XIIIb has very fine comb lines creating half-circles on fine-ware vessels (Fig. 8.37:113 and 114). The decoration usually appears in rows, but random placement is sometimes also seen on vessels. Two examples were found on fine-ware sherds from Hut 2.

Type XIIIl (Varsik XIIIa) has the same arches as in XIIIk in rows, but while one row may have the arches dragged downwards, the lower row may have arches applied with an orientation opposite to that of the upper row, so that the comb lines join, and creates the appearance of long wavy lines (Fig. 8.39:127). No examples were found of this type at the site.

Type XIIIc has comb lines creating long narrow bands of lines that connect to other bands of lines, curving at their apexes (Fig. 8.37:115). The comb lines number only half of the total number of lines seen in Type XIIIb or other types. One example was found, from Hut 2. Type XIIId was created by a comb with fewer bristles than seen on other types and forms an almost complete circle of lines, but stops short of actually joining the lines to finish the circle (Fig. 8.37:116). The type seems to be created by first applying the comb to create one-third of a circle, then using the same center point, create another third of the circle opposite the first third. The lines do not join, but create the illusion of a full circle. This type is seen both on fine-ware and coarse-ware vessels. One example was found in Hut 1.

Type XIIIm (Droberjar 340) has comb lines that create arches, the ends of which are then dragged far down the vessel, creating bands of lines (Fig. 8.39:128–130). The lines and arches are applied so that few overlap. The examples from Urbárske Sedliská are all from coarseware vessels. Ten examples were found of this type: seven from Hut 1 and three from Hut 2. Type XIIIn has a pattern similar to XIIIm, but here the arches are not fully completed and the long bands of lines are narrower, with blank spaces between them (Fig. 8.39:131 and 132). These half-formed arches are different from the arcs in XIIIi because of the greater number of lines in the XIIIm examples. The sherds with this type from Urbárske Sedliská are all from coarse-ware vessels. This decoration is reminiscent of Type XIIIc; XIIIn may be the coarse-ware version. Three examples were found, all from Hut 1.

Type XIIIe uses comb lines to create half-circles in rows, but the half-circles are kept separate (Fig. 8.37:117). This type is seen on both fine-ware and coarse-ware vessels. Four examples of this type were found: two from Hut 1 and two from Hut 2. Type XIIIf (Varsik XIIIb; Droberjar 350) has comb lines of connecting half-circles in rows, which create the illusion of a wavy line (Fig. 8.37:118–120). Six sherds with this decoration were found: four from Hut 1 and two from Hut 2.

Type XIIIo (Varsik XIIIf; Droberjar 370) has narrow bands of comb lines creating rows of wavy lines. No examples of this type were found at the site (Fig. 8.40:133).

Type XIIIg (Varsik XIIIe) has comb lines of XIIIf arranged in rows of half-cirlces of alternating orientation (Fig. 8.38:121). The pattern of the comb lines is of an exaggerated wavy line. No examples were found of this type at the site.

Type XIIIp (Varsik XIIIg; Droberjar 370) has only one or two comb lines creating rows of wavy lines. No examples of this type were found at the site (Fig. 8.40:134).

Type XIIIh has comb lines of half-circles applied randomly to the vessel so that their axes go in all directions and the lines overlap (7.38:122). The examples from Urbárske Sedliská are all from coarse-ware vessels. Two examples were found of this type, one from each hut.

Type XIIIq (Varsik XIIIh; Droberjar 330) has comb lines applied to create boxes of lines that are placed perpendicularly to each either in alternating placement. No examples of this type were found at the site (Fig. 8.41:135). Type XIIIr (Varsik XIIIi) has comb lines applied as lines on a vessel, with no arcs or arches present (Fig. 8.40:136). Three possible examples were found of this type, with one from Hut 1 and two from Hut 2.

Type XIIIi (Droberjar 351) has comb lines that do not create half-circles, but arcs, which are applied randomly so that lines overlap and disrupt arcs (Fig. 8.38:123). The examples from Urbárske Sedliská are all from coarseware vessels. Two examples were found, one from each hut.

Type XIIIs (Varsik XIIIj) has narrow bands of comb lines applied vertically down a vessel which are then overlapped by even narrower bands of comb lines at an 207

angle to the other bands (Fig. 8.40:137). No examples of this type were found at the site.

Type XIVe is an even more chaotic variant of XIVd and appears to have two types of incising overlying each other (Fig. 8.43:144 and 145). It appears that deeply incised lines, similar to XIVb in form, were applied first, and then the chaotic hatching of XIVd was applied over the other incised lines. This two-step process leaves behind distinctive dashes that contrast sharply with the chaotic hatching. I do not believe that this type was created the other way around, where the chaotic hatching was applied first and then the potter came back and put in the dashes. The dashes appear in precise rows, and were applied first; they were then interrupted by the placement of the chaotic hatching. Four examples were found: three from Hut 1 and one from Hut 2.

There were many other sherds with Type XIII decoration found at Urbárske Sedliská, but were too small or fragmentary to enable accurate identification of the subtype. A total of 48 sherds were indeterminate: 28 from Hut 1, four from G3, and 16 from Hut 2. 8.3.14. Type XIV. Incised Lines (Droberjar Types 200, 210 and 220) The decoration has a stylus dragged across the body of the vessel to create patterns. This decoration lasts from the first to third centuries AD (Varsik 2002, 138). Varsik identifies this type as “Hatching” (Varsik 2002, 138), but I have labeled type XIV as “Incised Lines” to be more inclusive, because upon examining the hatching done on vessels at Urbárske Sedliská, I noticed that the size and depth of the incised lines of the hatched variety also matched the incised lines seen on other sherds, which had a series of vertical lines on the body. Varsik’s first subtype is in fact vertical lines, but placed at an angle descending from a horizontal line, which is different from the vertical-line examples seen at Urbárske Sedliská. Varsik lists three sub-types (Varsik 2002, Taf. IV: XIVa– b); I have created three additional subtypes that are based on the vertical line differences and examples at Urbárske Sedliská.

Type XIVf is very different from the other incised lines. The decoration is a herring-bone pattern created by jabbing the stylus into the clay and then dragging it either down-and-out or up-and-out on the surface of the vessel (Fig. 8.43:60). The lines go from the deeply placed original jab by the stylus to a tapering shallow end. Type XIVf is similar to XId and e: Wedges in a herring bone pattern and XIIm: Fine impressions in a herring bone pattern. This herring bone pattern, however, is clearly made by a stylus, probably the same used to create the other Type XIV decorations; for this reason it is included here in this category. Only one example comes from Urbárske Sedliská: the pattern is on vessel 60, which is a unique example, combining four different types of decoration.

Type XIVa (Varsik XIVa; Droberjar 213), as already mentioned, has a horizontal incised line on a vessel with incised lines descending from it at an angle (Fig. 8.40:138). No examples of this type were found at Urbárske Sedliská.

8.3.15. Type XV. Finger Fluting The potter dragged a finger on the surface of the vessel to create this fluted-line decoration (Fig. 8.44:60). This type of line is different from the other types of fluted lines already discussed because of the roughness of execution. The finger fluting can be one line, such as the wavy line seen on vessel 60, or it can be rows of finger flutes. Type XV is a coarse-ware decoration and only one example comes from the site, on the aforesaid vessel 60. No example of the rows of finger flutes were found at the site, but Hut US82 had an example of this decoration (Turčan 1985, Obr. 6:3 and 4).

Type XIVb (Droberjar 210) has vertical incised lines on the body of a vessel, but they do not descend from a horizontal line (Fig. 8.40:139 and 140). The lines are not equally spaced on the body of the vessel and appear to have been applied quickly in a free-hand style. Some of the sherds have some cross-hatch lines appearing at one side (see 139) and it may be that the potter just did not finish hatching. This arrangement may indicate that Type XIVb is merely unfinished hatching and not a distinct decorative subtype. Nineteen examples were found at the site: eleven from Hut 1, three from Feature G2, one from G3, and four from Hut 2.

8.3.16. Type XVI. Rim Incising As the name implies, this decoration has lines incised into the rim of vessels (Fig. 8.44:33 and 72). There were two examples from the features: one from Hut 1 and the other from Feature G2. This decoration appears alone on a vessel or in combination with other types of decorations.

Type XIVc (Varsik XIVb; Droberjar 220) is the classic hatch decoration found on the body of German pottery (Fig. 8.41:59, 141 and 142). Forty-eight examples were found: eleven from Hut 1, two from Feature G2, and thirty-five from Hut 2. Type XIVd (Varsik XIVc) is a variant of XIVc with the hatching placed in a chaotic pattern, with lines crisscrossing at odd angles (Fig. 8.42:67 and 143). Three examples were found: one from Hut 1 and two from Hut 2.

8.3.17. Type XVII. Combination Patterns (Droberjar 920–980) The final type has different decorations combined on a vessel. As mentioned above, the most common is the finger impressions of Type XII combined with the 208

hatching of XIV (Fig. 8.44:60 and 94). Seven examples were found: three from Hut 1 and four from Hut 2.7 Vessel 60 also had other decorations added to the impressions and hatching, including the incised herring bone and finger fluting, creating a wavy line. Droberjar provides a whole range of combination types, beyond what was found at Urbárske Sedliská (Droberjar 1997, 41–42). Combining decorative patterns is common and various examples have been found at many sites both in Slovakia and Moravia.

common at 44% of the total for fine ware than the form is for coarse ware at only 31% of the total. When the huts are examined separately, the distribution of forms is slightly different (see Table 8.7). In Hut 1, Form II makes up 33% of the fine-ware total and Form III is at 50%, while among coarse wares Form II clearly dominates. In Hut 2, the numbers are more evenly distributed, with fine ware have an equal representation of Forms II and III, but among coarse wares Form III is slightly higher. What is clear from the data, however, is that Form I is consistently less than the other two forms.

8.4. Analysis of Forms and Decorations

Table 8.7. Summary of the Form Totals in Fine and Coarse Wares by Amount and Percentage from Huts 1 and 2.

For the following analysis of forms and decorations, please refer to Table 8.6 and Charts 8.1–7.8 at the end of the chapter. In regards to the section on the decorations used, please refer to Table 8.10 and Charts 8.9–7.15. 8.4.1. Discussion of the Form Data and Comparison to Other Huts

Hut 1

Fine Ware

Coarse Ware

I

1 (17%)

3 (21%)

II

2 (33%)

8 (58%)

III

3 (50%)

3 (21%)

I

1 (14%)

4 (29%)

II

3 (43%)

5 (33%)

III

3 (43%)

6 (40%)

Hut 2

A total of 1,079 German sherds were collected from the two huts. Out of this total, 915 were coarse ware and 168 were fine ware. Hut 1 yielded 399 coarse-ware sherds and 91 fine-ware, while Hut 2 yielded 516 coarse-ware and 77 fine-ware. These amounts are even more pronounced when total weights are considered, with coarse-ware sherds having a combined weight of 15.97 kg and the fine-ware at 1.37 kg.

Based on the totals and percentages from Huts 1 and 2, the most common forms were: 1) IB1 – wide bowls/dishes with rims rounded inward; 2) IIA2 – vessels with cylindrical necks slanted inward and a conical body with slight bulge at shoulder; 3) IIB2 – vessels with necks slanted inward and the body having a conical form with a slight bulge at the shoulder; 4) IIIA – narrow-mouthed containers; 5) IIIB – wide-mouthed containers. When looking at Table 8.4, one can see that the total amount of sherds for all three forms was not large, being only 29 sherds. The sherds belonging to IBI and IIB2 are all coarse ware, while IA was a fine-ware sherd. For IIA2, IIIA, and IIIB the sherds were mixed, but fine-ware forms outnumber coarse-ware forms. It is not surprising that these forms are the most common, since they are primarily dishes for the preparation and consumption of food.

In section 8.2, where the forms were defined, it was said that Form I vessels (one-piece containers) were the most common vessels found on German sites in southwest Slovakia. Form I is not the most common, however, in the data from the two huts at Urbárske Sedliská. Overall, Form II vessels are the most common at 43%, with Hut 1 having 50% as Form II, and Hut 2 with 37% of the total (See Charts 8.1, 8.6 and 8.8). One can accept this finding as is, but there is a potential problem. As mentioned in section 7.2, Forms IIA2–IIC2 can be considered an intermediate group betweens Form II and III, since they have a slight bulge at the shoulder. Defining the bulge can be somewhat subjective and some of the IIA2–IIC2 vessels may in fact fall under the Form III category. If, for the sake of argument, the IIA2–IIC2 vessels are moved to the Form III category, then form III is the most common at the site, making up 62% of the total and dropping Form II to only 14% of the total. I am inclined, in any case, to believe that the percentages for the Form II vessels are too high. There can be no doubt, however, that at least at Urbárske Sedliská, Form I vessels are not the most common.

In comparison to other huts, such as the assemblage from Hut US82, we see that Form I vessels occur at a higher frequency than the other two forms, which fits in with the statement made in Section 7.2 (see Table 8.8).8 The pottery from the hut had 41% as Form I, 36% as Form II, and 23% as Form III.9 These amounts are interesting in comparison to our two huts, with Hut I having 20%, 50%, and 30% respectively, while Hut 2 has 23%, 37%, and 40% respectively.

When we consider the occurrence of forms in fine ware and coarse ware some differences are discernible. As seen in Charts 8.3 and 8.4, Form II is slightly less common in fine ware (42% of the total), while for coarse ware Form II is at 45% of the total (this is using the data as presented without the modifications discussed above, moving some Form II to the Form III column). Form III is more

For comparison to huts from Moravia, five huts were chosen dating from the B2 period at random: one from Blučina (Hut 4), one from Komořany (Hut Z-9), two from Křepice (Huts 2 and 7), and one from Rajhrad (Hut 2) (Droberjar 1997, 179, 185–189 and 195).10 All five huts vary in size amongst themselves and with Huts 1 and 2 209

from Urbárske Sedliská, but the differences are not great. The inventory of ceramics and items also vary among the Moravian huts, with the lowest amount from the hut at Rajhrad, 98 German sherds, while Hut 2 from Křepice had the greatest at 308 sherds (Droberjar 1997, 186–187 and 195). Table 8.8 provides a summary of the total occurrence of each form from the various huts. It is immediately clear that Form II is the most commonly found (even with hut US82 included), but we must keep in mind the previous discussion where it was mentioned that some Form II may in fact be Form III for Huts 1 and 2. It is also interesting to note that, once again, Form I vessels are not the most commonly found in Moravia and in fact are usually less than the Form III vessels.

Table 8.9. Frequency of Occurrence for the Various Decorative Types by Feature Type Hut 1 G2 G3 Hut 2 Total I II 1 1 III IV 1 1 V 1 3 4 VI 2 2 VII VIII 1 2 3 IX X XI 38 8 8 38 92 XII 51 1 1 30 83 XIII 26 5 1 43 75 XIV 1 1 XV 1 1 2 XVI 3 4 7 XVII 123 15 10 123 271 Total 367 79 9 466 921 Undec.

In my opinion, these findings indicate that German potters were skilled in their craft. The usual assumption for German pottery has been local production, with each hut making their own supply of coarse wares. If this situation was the case, one would expect a preference for quick, easily made one-piece vessels, since they are the easiest to create. The findings, however, show a preference for two-piece and even more so for three-piece forms, which require more work and skill. It would seem, therefore, that whoever was making the pottery at each hut was skilled or that there was a full-time potter for each community or even several communities.

Once again in comparison to Hut US82 and the five huts chosen from Moravia, Huts 1 and 2 do not differ greatly from these other six huts. The three types commonly seen in Huts 1 and 2, Types XII, XIII, and XIV appear consistently at all the sites and in large numbers as well. Other types that appeared frequently were VI (Fine Punctures), VII (Barbotine), XI (Wedges), and XVII (Combination Patterns). These four types do not compare in quantity to the more common three types, but they make up most of the remaining percentage of the total. When the amounts from all eight huts are combined, as can be seen in Chart 8.11, it is clear just how popular XII, XIII and XIV were on German ceramics. Combined, the three types are 79% of the total, with XII at 34%, XIII at 24%, and XIV at 21%.

Table 8.8. Total Number of Each Form Found in the Various Huts Forms Huts I II III 4 10 6 Hut 1 5 8 9 Hut 2 19 17 11 US82 6 43 22 Blučina 4 Komořany Z-9 9 12 13 4 36 13 Křepice 2 15 56 10 Křepice 7 5 7 3 Rajhrad 2

Near Stupava, the site of Dúbravka was found to have a relatively large settlement dating from the first to fourth centuries AD. K. Elschek in his 1995 publication provided a useful page of pie-charts summarizing the amount and types of pottery found in five features, two of which (129/93 and 74/89) were six-post huts dating to the same period as the huts at Urbárske Sedliská (Elschek 1995, 51, Abb. 7:Objekt 129/93 and Objekt 74/89). Both of these huts had the same three common decorative types, XII, XIII, and XIV, appearing most frequently, much as the other huts already mentioned (Elschek 1995, 51: Abb. 7). An interesting difference between the two huts at Dúbravka and the others is the higher occurrence of decoration Type X, Roughening, which does not appear at Urbárske Sedliská but does in four of the huts from Moravia.

8.4.2. Discussion of the Decoration Data and Comparison to Other Huts In describing the various decoration types above it was mentioned that Types XII (Nail/Finger Impressions), XIII (Comb-Lines) and XIV (Incised Lines) were the most commonly found. As may be seen in Table 8.9 and the various charts, this statement held true at Urbárske Sedliská. A large percentage of the total was apportioned to these three types, and as may be seen in Chart 8.9, decoration type XII was the most common at 34%. It should be mentioned that if the undecorated sherds are included then they hold the majority at 77.3%, according to Chart 8.10.

210

8.5. Conclusion Form als die Funktion einzelner Gefäße in den Vordergrund gestellt. Dies führte zur Definition von vier Grundformen“ (Varsik 2002, 132).

The two huts from Urbárske Sedliská do not stray from what would be expected in terms of the German pottery. The types of forms and decorations are common ones found throughout southwest Slovakia and southeast Moravia. There are slight differences, but none is of such magnitude to cause Huts 1 and 2 to be labeled unique.

5. They bring to mind beer flagons and it can be argued these are specialized vessels for beer drinking. Beer was a staple drink of the ancient Germans (Tac. Ger. 23; Todd 2004, 79).

According to the data from Huts 1 and 2, Form II was the most commonly seen German pottery shape. This finding also holds true for assemblages from five huts chosen at random from Moravia. In regards to decorations found on German pottery, Types XII, XIII, XIV were the most common at Urbárske Sedliská, having 33.9%, 30.6%, and 27.7% of the total respectively. This finding holds true when the assemblages from the same five huts from Moravia were examined as well, with Type XII at 33.9%, Type XIII at 23.6% and Type XIV at 20.7%.

6. The possibility that vessel 57 was from the Przeworsk was first mentioned by V. Turčan, when reviewing the pottery from the two huts. 7. The total here includes the sherds that have already been listed under the individual decorations as well. 8. At the moment, only Hut US82’s assemblage was examined in enough detail to be able to apply the new typology I have created. It is hoped that in the near future a detailed examination of pottery from other sites in the area can be done to expand the typology and see how useful it is for the examination of pottery on a broader scale.

It is clear that the site of Urbárske Sedliská was a connected part of the greater Germanic culture that made up the Quadi territory, during the early second century AD (see Chapter 7 for a discussion of the dates for the huts).

9. The analysis of the pottery for the identification of the forms was conducted by me. The pottery was examined through images taken of them, not by examination of the material itself. The percentages given, therefore, should be seen as preliminary ones.

Endnotes 1. T. Kolník, with V. Varsik and J. Vladár, has created the catalog of finds for the settlement at Branč, but it remains unpublished (Kolník et al., 2005, 313).

10. The five huts come from Droberjar’s Studien zu den Germanischen Siedlungen (1997), in which there is a section summarizing the inventory of all items found in the huts used in his study. The inventories of the German pottery in his work divide the sherds by form and decoration established by Droberjar. These inventories were then adapted by me to fit into the form and decoration typologies established by V. Varsik and expanded in this chapter.

2. In putting together the list of sherds analyzed for the forms, samples were not taken from Feature G3, both because there were too few good examples and problems at the time of identifying the age of the feature. Sherds with various types of decoration from Feature G3, however, are included in section 8.3. A full analysis of the sherds from Feature G3 will be published in the near future. 3. The typology lists as many possible forms as there might be, but not every type is represented by the pottery excavated at Urbárske Sedliská. 4. Varsik’s Form III is defined as vessels with three more or less independent members: the neck, shoulder and lower part (Varsik 2002, 133 and 152). These vessels come in all sizes and function, with the S-shape most common, and the surface is always carefully treated by polishing and frequently graphite is applied (Varsik 2002, 133). The polishing and application of graphite is usually reserved for fine ware vessels and therefore should not be considered part of the definition of a form. In regards to From IV, Varsik defines it as being three-piece vessels also with an S-shape, but specifically coarse-ware vessels with rougher clays (Varsik 2002, 134). That is, Form IV is based on the quality of clay, not on the form. Varsik’s intention, however, is clear when he states, “Bei der typologischen Auswertung habe ich deshalb mehr die

211

Tables and Charts Table 8.4. List of Abbreviations Used in the Catalog Inventory references: Inv. No. = sherd number, followed by unit and stratum numbers Ft. = feature from where the sherd came Fig. = figure, located at the end of the chapter Sherd/Vessel Descriptions: Fr. = form (see Tables 7.1–3) Un. = unknown H. = height in cm P.H. = preserved height in cm D. = diameter in cm B.D. = base diameter Est. D. = estimated diameter in cm Th. = thickness in mm or cm L. = length in cm P.L. = preserved length in cm W. = weight in grams Sherd Colors (Munsell colors used): O. = outer surface I. = interior surface C. = core/fabric of the sherd M. = margins of the fabric O.M. = outer margin I.M. = inner margin

Vessel Types: Narrow-Mouthed Containers = h>d Wide-Mouthed Containers = d>h Deep Bowls/Stew Pots = d>h or d=h Cups Flagons Beakers

Inclusion Amounts: Abundant Moderate Sparse

Feel of Sherd (outer surface of sherd): Harsh = abrasive to the fingers Rough = irregularities can be felt Smooth = no irregularities can be felt Texture Descriptions: Subconchoidal = breaks somewhat like glass or flint Smooth = flat or slight curved, no visible irregularities Fine = small, closely spaced irregularities Irregular = larger, more widely spaced irregularities Hackly = large and generally angular irregularities Laminated = ‘stepped’ effect

212

Table 8.6. Total Frequency of Occurrence for the Various Forms Hut 1 G2 Hut 2 Form Total Fine Coarse Fine Coarse Fine Coarse 1 1 IA IB 2 1 2 5 IB1 1 1 1 3 IB2 IC 1 1 2 IC1 IC2 ID IIA IIA1 1 1 2 1 5 IIA2 1 1 IIA3 1 1 IIB 1 1 2 IIB1 4 2 6 IIB2 1 1 IIB3 2 2 IIC IIC1 1 1 IIC2 IIC3 1 1 4 6 IIIA 3 1 1 2 7 IIIB 1 1 2 IIIC IIID 5 2 3 4 14 Bases 3 1 4 5 5 18 Unident. 14 17 1 6 15 24 77 Total

Table: 8.10. Summary of the Occurrence of the Various Decoration Types at the Various Sites Mentioned in the Text. Type Hut Hut Hut Blučina Komořany Křepice 2 Křepice 7 Rajhrad 1 2 US82 4 Z-9 2 1 7 1 I II 1 1 III 1 IV 1 V 1 3 8 13 13 13 2 VI 2 1 4 9 13 5 3 VII 2 VIII 1 2 4 13 IX 2 1 4 1 X 3 9 3 4 4 XI 38 38 89 47 15 33 19 22 XII 51 30 46 7 26 31 15 4 XIII 26 43 37 21 12 28 15 2 XIV 1 2 XV 1 1 XVI 3 4 6 4 11 5 XVII

213

Charts Chart 8.1. Frequency of Forms I - III by Percentage

I 24% III 33%

II 43%

Chart 8.2. Frequency of Forms shown by Percentage (Only Those Forms Shown which had Representative Sherds from the Features)

IIIC 7%

IA1 IA2 2% 2%

IB 5% IB1 11%

IIIB 16%

IB2 2% IC1 2% IIA1 5%

IIIA 9% IIC2 2%

IIA2 7%

IIC 5% IIB3 2%

IIA3 2% IIB 2%

IIB2 14% IIB1 5%

214

Chart 8.3. Frequency of Forms found in Fine Ware by Percentage (Without Bases and Unidentified)

IIIC 8%

0%

IA 8%

IB2 8%

IIIB 30% IIA2 22%

IIIA 8%

0%

IIB 8%

IIB1 8%

Chart 8.4. Frequency of Forms found in Coarse Ware by Percentage (Without Bases and Unidentified)

IB 4%

IIIC 9%

IB1 18%

IIIB 13%

IB2 4% IC1 4%

IIIA 14%

IIA1 4% IIB3 4% IIB1 4%

IIB2 14%

215

IIA3 4%

IIA2 4%

Chart 8.5. Frequency of Forms found in Hut 1 by Percentage IA1 3%

Unidentified 20%

IB1 10% IIA1 6%

IIB1 6% Bases 17% IIB2 13% IIIC 3% IIIB 13%

IIIA 3%

IIC 6%

Chart 8.6. Frequency of Forms from Hut 1 by Percentage (Without the Bases and Unidentified Categories)

IA1 5%

IIIC 5%

IB1 15%

IIIB 20%

IIA1 10%

IIIA 5%

IIB1 10%

IIC 10%

IIB2 20%

216

Chart 8.7. Frequency of Forms from Hut 2 by Percentage IB1 3% IA2 IB 3% 3%

IB2 3% IC1 3%

Unidentified 25%

IIA2 8% IIA3 3% IIB 3% IIB2 5% IIB3 3%

Bases 17%

IIIA 8% IIIB 8%

IIIC 5%

Chart 8.8. Frequency of Forms from Hut 2 by Percentage (Without the Bases and Unidentified Categories)

IA2 5%

IIIC 10%

IB 5%

IB1 5% IB2 5%

IIIB 14%

IC1 5%

IIA2 13%

IIIA 13%

IIB3 5%

IIB 5%

IIB2 10%

217

IIA3 5%

Chart 8.9. Total by Percantage of Decoration Types Found in the Four Features

XV 0.4%

XVI 0.7%

XVII 2.6%

III 0.4%

V 0.4%

VI 1.5%

VII 0.7%

IX 1.1%

XIV 27.7% XII 33.9%

XIII 30.6%

Chart 8.10. Total by Percentage of Decoration Types Found in the Four Features including Undecorated V 0.1%

VI 0.3%

VII 0.2%

III 0.1%

IX 0.3%

XII 7.7% XIII 7.0% XIV 6.3% XV 0.1% XVI 0.2% XVII 0.6%

Undecorated 77.3%

218

Chart 8.11. Total by Percentage of Decoration Types Present in Hut 1 XVII 2%

XVI 1%

III V 1% 1%

VI 1%

IX 1%

XIV 21% XII 31%

XIII 41%

Chart 8.12. Total By Percentage of Decoration Types Including Undecorated Found in Hut 1 III 0.2%

V 0.2%

VI 0.2%

IX 0.2%

XII 7.8% XIII 10.4% XIV 5.3% XVI 0.2% XVII 0.6%

Undecorated 74.9%

219

Chart 8.13. Total by Percentage of Decoration Types Found in Hut 2 XVII 3%

VII 2%

VI 2%

XV 1%

IX 2%

XII 31%

XIV 35%

XIII 24%

Chart 8.14. Total by Percentage of Decoration Types Including Undecorated Present in Hut 2 VI 0.5%

VII 0.3%

IX 0.3%

XII 6.5%

XIII 5.1% XIV 7.3% XV 0.2% XVII 0.7%

Undecorated 79.1%

220

Chart 8.15. Shows the percentage breakdown of the decorations types from all eight huts combined.

XVII 3.7%

II 0.0% I 1.0%

IV III V 0.2% 0.1% 0.1%

VI 6.0%

VII 4.2% VIII 0.2% IX 2.2%

XVI 0.2%

X 0.9%

XV 0.3%

XI 2.6%

XIV 20.7%

XII 33.9%

XIII 23.6%

221

Figures

Figure 8.1. Map of sites mentioned in Chapters 8–10 (both in text and figure-captions), with Roman forts indicated along the Danube River by letter. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.

Stupava/Urbárske Sedliská Dúbravka Bratislava-Trnávka Sládkovičovo Abrahám Kostolná pri Dunaji Kráľová pri Senci Devínska Nová Ves Bohdanovce Báhoň Kuchyňa Branč Pobedim

A. Vindobona B. Carnuntum

14. Blučina 15. Křepice 16. Horní Věstonice 17. Mikulov 18. Mikulčice 19. Hrušky 20. Rajhrad 21. Křižanovice 22. Komořany 23. Mušov 24. Zohor 25. Komjatice 26. Cífer

27. Bíňa 28. Krížovany nad Dudváhom 29. Trnava-Hrnčiarovce 30. Šarovce 31. Horná Streda 32. Janova Ves 33. Bratislava 34. Bratislava-Devín 35. Krakovany-Stráže 36. Vysoká pri Morave 37. Mlékojedy

C. Gerulata D. Brigetio

E. Celemantia

222

Figure 8.2. Varsik’s Form I typology (modified image from Varsik 2002, 141, Taf. II:1).

223

Figure 8.3. Examples of Form I vessels. Not to scale. (IC2 and ID examples are modified from Droberjar 1997, Abb. 38:2424 and Abb. 46:5301, 5305, and 5307)

224

Figure 8.4. Varsik’s Form II typology (modified image from Varsik 2002, 141, Taf II:2).

225

Figure 8.5. Examples of Form II vessels. Not to scale. (IIA1 example is a modified image from Varsik 2002, Taf. II:2,b; IIC1 and IIC3 examples are modified images from Droberjar 1997, Abb. 36:2322 and 2328; Abb. 40:2755)

226

Figure 8.6. Varsik’s Form III typology (modified image from Varsik 2002, 142, Taf III:3).

227

Figure 8.7. Varsik’s Form IV typology (modified image from Varisk 2002, 142, Taf III:4).

228

Figure 8.8. Examples of Form III vessels. Not to scale. (IIIA and IIID examples are modified images from Varsik 2002, Taf. III:1,a,b,f–h)

229

Figure 8.9. Fine-ware Form I. 1 IA, 2 IB2.

230

Figure 8.10. Fine-ware Form II. 3–5 IIA2, 6 IIB1, 7 IIB, and 8 IIC2.

231

Figure 8.11. Fine-ware Form III. 9 IIIA, 10 and 11 IIIB.

232

Figure 8.12. Fine-ware Form III. 12 and 13 IIIB, 14 IIIC.

233

Figure 8.13. Fine-ware Bases.

234

Figure 8.14. Fine-ware Unidentified Rims.

235

Figure 8.15. Coarse-ware Form I. 31–34 IB1.

236

Figure 8.16. Coarse-ware Form I. 35 IB1, 36 and 37 IB2.

237

Figure 8.17. Coarse-ware Form I. 38 and 39 IC1.

238

Figure 8.18. Coarse-ware Form II. 40 and 41 IIA2, 42 IIA3, 43 IIB1.

239

Figure 8.19. Coarse-ware Form II. 44–46 IIB2.

240

Figure 8.20. Coarse-ware Form II. 47–49 IIB2.

241

Figure 8.21. Coarse-ware Form II. 50 IIB3, 51 and 52 IIC.

242

Figure 8.22. Coarse-ware Form III. 53–55 IIIA.

243

Figure 8.23. Coarse-ware Form IIIA and IIIB. 56 and 57 IIIA, 58 IIIB.

244

Figure 8.24. Coarse-ware Form IIIB. (Drawing by Peter Šimčík)

245

Figure 8.25. Coarse-ware From IIIB. Vessel 60 shown without ornamentation.

246

Figure 8.26. Coarse-ware Form IIIB. Vessel 60 shown with ornamentation. (Drawing by Peter Šimčík)

247

Figure 8.27. Coarse-ware Form IIIC and Bases. 61 IIIC. (64 drawn by Peter Šimčík)

248

Figure 8.28. Coarse-ware Bases. (67 drawn by Peter Šimčík)

249

Figure 8.29. Coarse-ware Unidentified Rims.

250

Figure 8.30. Coarse-ware Unidentified Rims.

251

Figure 8.31. Ornament Typology I–VI. Letters refer to sub-types. Not to scale. Sherds or vessels with letters by them are modified images taken from Varsik 2002, Taf. IV. Drawings 1, 78, 81, and 82 are by Peter Šimčík. Sherds 79 and 80 are from photographs.

252

Figure 8.32. Ornament Typology VII–IX. Letters refer to sub-types. Not to scale. Sherds or vessels with letters by them are modified images taken from Varsik 2002, Taf. IV. Drawings of 83, 84, 85 and 86 are by Peter Šimčík. Sherd 2 is by the author.

253

Figure 8.33. Ornament Typology X–XII. Letters refer to sub-types. Not to scale. Sherds or vessels with letters by them are modified images taken from Varsik 2002, Taf. IV. Vessel 87 is a modified image from Elschek 1995, Abb. 5:17. Drawings of 60, 90, and 91 are by Peter Šimčík.

254

Figure 8.34. Ornament Typology XIIc–e. Not to scale. Drawings 94 and 95 are by Peter Šimčík.

255

Figure 8.35. Ornament Typology XIIf–h. Scale usable for all pieces shown. Drawings by Peter Šimčík.

256

Figure 8.36. Ornament Typology XIIi–n. Scale usable for 106–110. Drawings by Peter Šimčík. (Types XIIj and XIIk are modified images from Varsik 2002, Taf. IV:XII,b and f)

257

Figure 8.37. Ornament Typology XIIIa–f. Scale for all examples shown. Drawings by Peter Šimčík.

258

Figure 8.38. Ornament Typology XIIIg–k. Scale usable for 122, 123, 125, and 126, which were drawn by Peter Šimčík. (121 and 124 are modified images from Varsik 2002, Taf. IV:XIII,e, and d)

259

Figure 8.39. Ornament Typology XIIIl–m. Scale usable for 128, 129, 131, and 132, which were drawn by Peter Šimčík. (127 is a modified image from Varsik 2002, Taf. IV:XIII,a)

260

Figure 8.40. Ornament Typology XIIIo–s and XIVa and b. Scale for examples 139 and 140, which were drawn by Peter Šimčík. (133–138 are modified images from Varsik 2002, Taf. IV:XIII,f–j and XIV,a)

261

Figure 8.41. Ornament Typology XIVc. Drawings by Peter Šimčík.

262

Figure 8.42. Ornament Typology XIVd. Drawings by Peter Šimčík. Example 143 is listed as Vessel 67 under the Form section.

263

Figure 8.43. Ornament Typology XIVe and f. Drawings by Peter Šimčík.

264

Figure 8.44. Ornament Typology XV–XVII. Drawings 60 and 94 by Peter Šimčík.

265

been found since 1981 and the number of sites increased, the types of pottery have not changed.1 In fact, when the catalog of material created by Krekovič is compared to the general discussion of Roman pottery found in Pannonia by E. Bónis in The Archaeology of Roman Pannonia (1980), there is almost no difference in the types of pottery mentioned. This finding is strengthened further when looking at the catalog of Roman pottery found at the auxiliary fort and vicus of Gerulata, located on the Danube frontier (Fig. 9.1) (Krekovič 1998). It would seem safe to say that whatever pottery was available in Pannonia was therefore available to the Germans living beyond the Danube (Krekovič 1997a, 233).

Chapter 9 The Roman and La Tène Pottery from Urbárske Sedliská The following chapter deals with the Roman and La Tène pottery recovered from Urbárske Sedliská. There are three main sections to the chapter: 1) the Roman pottery; 2) a general discussion of the Roman pottery at the site and reference to other huts beyond Urbárske Sedliská; 3) the La Tène pottery. The Roman and La Tène sections will first have general introductions to the pottery, followed by a catalog of what was recovered. The Roman pottery is presented first because it is much more substantial than the La Tène pottery.

In regards to Roman pottery found in Pannonia, a series of important works have appeared over the years.2 There have been many regional studies, but the first real study using material from a large portion of ancient Pannonia (Hungary and Slovenia) was by E. Bónis (1942), providing a look at early imperial production, while K. Póczy (1957) developed a chronological framework of the development of forms and techniques based on the material from the site of Intercisa (Fig. 9.1) (Brukner 1981, 169). More recently, O. Brukner (1981) created an extensive catalog of Roman ceramics found in the portion of Lower Panonnia that was in the former Yugoslavia. Another important catalog of pottery was published by D. Gabler (1989) from excavations of the fort at ÁcsVaspuszta (Fig. 9.1).3 For Carnuntum (Fig. 9.1), V. Gassner (1989, 1990, 1991, 1993) and M. Grünewald (1979, 1981, 1986) have dealt with the pottery.

9.1. Types of Roman Pottery found in Southwest Slovakia Pottery from the Roman Empire makes up the largest portion of imported goods in the barbaricum (Krekovič 1981, 341; Kolník 1998b, 74). The types of pottery are varied and while some come from various parts of the empire, especially Northern Italy, most were manufactured near the frontier. Along the Middle Danube, the largest number comes from Pannonia, but also some from Raetia. Even though pottery starts to be imported into southwest Slovakia in the second quarter of the first century AD, the most intense period of importation was from the second half of the second century to the mid-third century AD (Krekovič 1981, 367; 1987, 270; 2000b, 260). Indeed, at the start of the second century there is a dramatic rise in the amount of pottery being imported, which may correspond to the establishment of production centers in Pannonia itself during this same period (Krekovič 1981, 367). These centers would have increased the supply of pottery available for export from Pannonia and driven down prices in the markets both for provincials and visiting German buyers. By the mid-third century there is a marked decrease in the amount of pottery imported into the barbaricum, which may be a reflection of the series of crises that struck the empire during this period, disrupting trade (Krekovič 1981, 367; 2000b, 260). Importation of Roman pottery continued until the beginning of the fifth century (Krekovič 2000b, 260). Southwest Slovakia has the highest occurrence of Roman pottery along the Middle Danube, both in number of pieces and variety of types (Krekovič 2000b, 260). In regards to find-spots, the majority of Roman pottery is found in German settlements, with only a few examples coming from cemeteries (Krekovič 2000b, 259).

For Slovakia, one of the first discussions of Roman pottery was by J. Eisner in Slovensko v Pravěku, in which he provides basic descriptions of the types of Roman pottery found at various sites (Eisner 1933, 222–230). E. Beninger provided detailed descriptions of Roman pottery found in German context in his publication Die germanischen Bodenfunde in der Slowakei (1937). Another, but brief, summary of Roman pottery found in Slovakia was created by V. Ondrouch in Limes Romanus na Slovensku (1938, 90–100). More recent publications include T. Kolník’s catalog (1980) of grave goods from the three Roman-Period cemeteries at Abrahám, Kostolná pri Dunaji, and Sládkovičovo (Fig. 8.1), providing an important overview of the types of imported pottery found in the barbaricum in mortuary context. A more recent publication from Slovakia dealing with Roman pottery from the limes was published by E. Krekovič (1998), cataloging the wares found at the auxiliary fort and vicus of Gerulata (already mentioned above).4 A basic description of Roman pottery found on the territory of Slovakia was provided by T. Kolník in Územie Slovenska pred Príchodom Slovanov (1998b, 74–76). In regards to terra sigillata, the first real catalog for Slovakia was made by F. Křížek (1939), which he then expanded upon with later finds (1961 and 1966). Recent works on terra sigillata were made by K. Kuzmová and P. Roth (1988), and a very extensive study was published in 1997 by K. Kuzmová, providing a detailed discussion and catalog of finds in southwest Slovakia.

E. Krekovič wrote a seminal article on Roman imported pottery in 1981, which is still useful today as a general overview of the pottery types seen in southwest Slovakia and Moravia as well. Even though more material has 266

In the following discussion, I shall be using Krekovič’s catalog extensively with additions where needed, especially if chronologies or other information have changed. My discussion will not go into the same detail provided by Krekovič, because the goal is merely to provide a basic introduction of what is found in southwest Slovakia and Moravia. It should be noted that Krekovič did not include a discussion of terra sigillata in his 1981 article; the publications by K. Kuzmová and P. Roth will be used for this pottery.

Handled pitchers were imported from northern Italy during the first century, since the local Pannonian population did not make handled vessels (Bónis 1980, 362; Krekovič 1981, 360). By the second century, Pannonian workshops were creating their own pitcher types. Gassner identified various types of pitchers at Carnuntum, grouping them by their differences in funnel shape and rim (Krekovič 1998, 19). Many of these pitchers were in use during the same periods and/or over several periods, sometimes without a clear change in form (Krekovič 1998, 19).5

9.1.1. Brick-Colored Wares

2) Pots. These have a wide range in form and appearance, thereby limiting their use for dating (Krekovič 1981, 361).

These vessels have a varied color, ranging from yellow to orange-red (depending on the temper and/or firing) and so E. Krekovič has labeled them Brick-Colored Wares, but they are also referred to as Yellow Wares (Krekovič 1981, 360; 2000b, 260). The clay is well levigated with only small inclusions (Kolník 1998b, 75). Under this group are also those vessels that have a carefully applied slip of similar color, varying from orange-red to brown, but there are also many unslipped vessels (Krekovič 1981, 360; Kolník 1998b, 75). Two specific subgroups of this ware, Pannonian Striped Wares and Marbled Wares, are treated separately below in sections 8.1.2 and 8.1.3 respectively. Krekovič listed them separately because of their specific appearance, making them stand out in comparison to other Brick-colored Wares and therefore having greater chronological significance (Krekovič 1981, 360).

3) Bowls. There are various forms which bowls can take. One form found in Slovakia with parallels in Pannonia is a single-piece hemispherical bowl with a horizontal rim and an orange-red slip (Fig. 9.2:4), in use from the first to the third centuries AD (Krekovič 1981, 361–362; 1998, 17). Another bowl type similar to the above is a low profile design with horizontal rims decorated with grooves, imitating terra sigillata forms (Krekovič 1981, 362) (Fig. 9.2:5). There are also bowls that have collars, located on the body of the vessel, imitating the Drag. 38 shape (Krekovič 1981, 362; 1998, 23) (Fig. 9.2:6). These are similar in shape to smooth, gray ring-bowls (mentioned below), but smaller (Krekovič 1998, 23). Collared bowls appear through the four periods devised by Gassner, but during the first period, these bowls are similar to Italian terra sigillata forms (Krekovič 1998, 23). This form is similar to that seen among the mortaria and so these bowls have been labeled sometimes as a form of mortaria made in the terra-sigillata technique (Krekovič 1981, 362).

The regular Brick-Colored Wares were manufactured throughout the Roman Period and therefore are difficult to place in a chronological framework merely based on technical appearance (Krekovič 1981, 460). Originally, the only chronological changes well identified were that by the Late Roman Period the fabric became coarser and the glossy slip disappeared, while the matte slip still appeared into the fourth century, but in smaller amounts, and the Brick-Colored Wares replaced the Coarse-Gray and Glazed Wares (Krekovič 1981, 360). V. Gassner (1989), however, classified this group of pottery through the finds at Carnuntum, which were in datable contexts, dividing the pottery into four periods: 1) AD 50–120, 2) 120–170/180, 3) 170/180–230, and 4) 230–300 (Krekovič 1998, 19). He also classified the various forms used in creating Brick-Colored Ware, providing dates of usage (Krekovič 1998, 19–27).

4) Plates. The form is a rare find in Slovakia. One example from the site of Bíňa (Figs. 9.1, 9.2:7) has a parrellel found at Carnuntum, which imitates either Dragendorf 24/25 or 6 (Krekovič 1981, 362). Other plate types seen at Gerulata or Carnuntum are plates with rims straight or bent inwards, which are similar to Pompeian versions dating to the first century (Krekovič 1998, 25). A plate type dating to Gassner’s periods 2–4 is one with the rim bent out and set with a groove (Krekovič 1998, 25). 5) Goblets/Mugs. These usually survive only in small fragments, making it hard to identify precisely their parallels in Pannonia.

In Slovakia, this ceramic group is one of the most commonly found both in amount of sherds and in the various forms represented, with most of the examples found dating to the second century (Krekovič 1981, 363). For ease of discussion, the wares can be grouped by forms, but it should be mentioned here that pitchers and bowls are the most common forms found in Slovakia (Krekovič 1981, 363).

6) Cups. As with mugs, cups usually survive in small fragments. One type that has been identified is a twoeared bell-shaped form (Fig. 9.2:8), which has been found throughout Pannonia, and one production site was identified at Aquincum (Fig. 9.1), operating during the principates of Trajan and Hadrian (Krekovič 1981, 363).

1) Pitchers (jugs) (Fig. 9.2:1–3). These have either narrow or wide necks, with one to three handles, as seen on examples from Pannonia proper (Krekovič 1981, 360). 267

9.1.2. Panonnian-Striped Wares

(1950, 21) with material from the Rhineland: the first spans the last third of the first century to the first third of the second century and the second period spans from the end of second century to the end of the Roman Period (Krekovič 1981, 348–349; Brukner 1981, 179; Krekovič 1997b, 42). During the first period, the slip has a softer orange-red marbling, while for the later period the marbling is thicker, muted red-brown (Droberjar 2002, 126).

Along with Pannonian-Stamped Wares (discussed below) this is a typical Pannonian ceramic. The most popular shapes were pitchers and pots, which had on the upper 2/3 or in the middle of the vessel a wide band of red or brown paint applied by brush (Krekovič 1981, 353; 1998, 17). The most common form of pot has a bulging body, with a horizontal rim, and dates to middle of the first century to the first third of the third century AD (Krekovič 1998, 17) (Fig. 9.3:1–3). The band of paint was the only decoration on the vessels, but the band itself might be divided by etched-in lines or there may be etched wavy lines or a series of notches, made by a toothed wheel (Krekovič 1981, 353–354) (Fig. 9.3:4). This ware has La Tène roots and the oldest known examples from the Roman Period come from Poetovia (Fig. 9.1), from the time of the Flavians (Bónis 1980, 359; Krekovič 1981, 354). The typology and chronology of this ware was worked out by E. Bónis (1970) (Gabler 1989, 477). There were production centers at Acquincum, dating from the Hadrianic period to the first half of the third century, as well as at Gorsium in Pannonia (Fig. 9.1); but the main center of production appears to have been Brigetio, starting at the beginning of the second century and reaching its height during the second half of the second century and first half of the third century (Bónis 1980, 359; Krekovič 1981, 354; 1998, 17). In southwest Slovakia, almost every settlement site that dates to the Roman Period has some type of this ware (Krekovič 1981, 354). Krekovič believes it most likely that since Brigetio has such large-scale production facilities for this ware, most of the examples found in Slovakia originate from there (Krekovič 1981, 354–355).

The origin of marbled ware is still debated. One line of thought is that the type was developed not in the western provinces but in the east, and that it spread westward by slaves who made pottery for the legions. This argument is weakened by the discovery of marbled wares in civilian contexts and the fact that no eastern slave names are found as graffiti on marble-ware vessels (Krekovič 1997b, 41; 1998, 14). Marbled wares have been found, however, at many legionary forts and so it does appear that there was military production (Krekovič 1998, 15). At the moment, based on the sketchy evidence, it would seem that marbled wares were produced both by military and civilian potters, but which was first is still unknown (Krekovič 1997b, 42; 1998, 15). In Pannonia, marbled wares first appear at Emona during the middle of the first century AD, apparently connected to local pottery production (Krekovič 1997b, 41; Krekovič 1998, 14). Aquincum has marbled wares with late La Tène forms discovered in the civilian settlement (Krekovič 1998, 15). Marbled Wares in Pannonia were divided into two periods by K. Póczy, with the first period having finer examples produced from the Flavian period to the first third of the second century, while the second period is in the first half of the third century with rougher marbling appearing at Aquincum (Krekovič 1997b, 41). During the third century there is a gradual fading away of the technique, disappearing by the fourth century (Krekovič 1997b, 41–42). For Pannonia, marbled vessels were primarily bowl shaped, but pitchers, jugs, and cups are also present (Krekovič 1997b, 42). There are only a few examples from Slovakia.

9.1.3. Marbled Wares These vessels are identified by their surface treatment, which involved the application of a red or brown slip with a brush or sponge to imitate the appearance of marble (Krekovič 1981, 348; Droberjar 2002, 126). The slip itself is similar in composition to those found on Pannonian-Striped Wares and Brick-Colored Wares. There are two forms of marbling based on techniques spoken of in German literature: 1) true marbling (Fig. 9.3:5 and 6), which used a sponge dipped in the color, and 2) where a brush is used creating an effect called “geflammte,” which has tongues of red paint on the vessel resembling flames (Fig. 9.3:7) (Krekovič 1981, 348). Krekovič notes, however, that a brush could be used to create true marbling as well (Krekovič 1981, 348).

9.1.4. Mortaria These are thick-walled vessels, with wide and downturned horizontal rims, used for the grinding and preparing of food (Hayes 1997, 80) (Figs. 9.3:8, 9). Many of them have roughened interiors with the addition of a layer of sand or stones before the vessels were fired (Droberjar 2002, 187). There has been some debate about their exact use, because some vessels are either smaller, thinner-walled, or have a smooth, or glazed interior (Krekovič 1981, 355). The consensus is that smaller mortaria, or those with thinner walls, or smooth interiors were used not for grinding but for the curdling of milk or creating other milk products (Krekovič 1981, 355). There are other smaller mortaria with roughened interiors, which are thought to have been used for grinding salt or for squeezing out the juices of fruit (Krekovič 1981, 355). It must be said, however, that

Archaeologically it has also been shown that marbling is seen first in the west in the pre-Flavian period, with South-Gaulish sigillata producers wanting to imitate glass vessels (Krekovič 1997b, 42). This marbling technique was then soon transferred to other types of vessels and spreads out from Gaul, so that by the mid-first century marbled wares are being manufactured in Raetia (Krekovič 1997b, 41). There are two periods for non terra-sigillata marbled wares first determined by E. Gose 268

many of the roughened mortaria could have been used for multiple tasks.

357). The other vessel was found in a Sarmatian settlement dating to the first half of the fourth century (Krekovič 1981, 357). Krekovič explains this as either the Sarmatians prepared food differently than the Germans or they had weaker ties to the provincials in Pannonia (Krekovič 1981, 357). Again it is worth considering the role of beer in the popularity of mortaria among Germans and not among the Iazyges.

For Pannonia, mortaria production appeared first at the workshops in Poetovio and Aquincum during the time of the Flavians and Trajan (Brukner 1981, 181). These vessels were shallow, larger in size, thicker-walled, and made of yellowish-whitish clay (Brukner 1981, 181). In Slovakia two types of mortaria occur, and they are used in different periods as well. The first type are unglazed mortaria in use during the second century AD, while the second type are mortaria with glazed interiors in use from the third to fourth centuries (Krekovič 1981, 356; 2000b, 260). Another dating difference that has been determined is wall thickness, with solidly made mortaria being of the Early Roman Period (especially in Gassner’s first period, mentioned above under Brick-Colored Wares, section 8.1.1). Vessels that are smaller and thinwalled belong to the Later Roman Period (Krekovič 1998, 24).

9.1.5. Smooth-Gray Wares The ceramic is readily identified by the fine-gray, sometimes even almost white, material making up the fabric, and having been fired in a reducing atmosphere (Krekovič 1981, 363–364). The clay is well levigated, without a lot of sand mixed in, and the firing is of a good quality (Kolník 1998b, 75). The surface of these vessels at first is left alone, but by the mid-third century there is often a gray-to-black slip, which becomes more matte over time (Krekovič 1981, 364). This ware can be referred to either as smooth-gray or fine-gray wares. Overall, the group is very similar to Brick-Colored Wares in terms of material used and the forms of the vessels (Krekovič 1998, 28). Chronologically, the wares are coeval with Brick-Colored Wares, but fall out of use before the start of the fourth century (Krekovič 1981, 364). Pannonian-Stamped Wares, mentioned below in section 8.1.6, are a specialized form of the Smooth-Gray Wares (Krekovič 1998, 13).

Mortaria are usually brick colored (thereby many can be considered part of the Brick-Colored Wares), sometimes with a gray core, which is usually associated with the glazed examples (Krekovič 1981, 356). The glaze itself is green tinted, usually green-brown (see section 8.1.12 for more information on Glazed Wares) (Krekovič 1981, 356). Mortaria fragments have been found all over southwest Slovakia and appear to have been a popular vessel among the Germans; the phenomenon is interesting since these are a specialized vessel type. Krekovič is of the opinion that Germans must have had close relations with provincials first to see the use of mortaria and then to realize their potential uses in German settings (Krekovič 1981, 356). In fact, by the Late Roman Period, when other imported glazed wares drop out of use in the barbaricum, glazed mortaria are still used in great numbers, thereby implying that these vessels played an important role in German homes (Krekovič 1981, 356– 357). One wonders if in fact these mortaria were not used for the creation of milk products, but instead for the processing of hops and barley toward the creation of beer, which has already been mentioned as a popular drink among the Germans. Another explanation for the number of mortaria found in the barbaricum is the making of a flavorful vegetable-herb sauce known as moretum, which may have become popular among the Germans (Droberjar 2002, 187).

As with Brick-Colored Wares, there is a large variety of fine-gray ware examples from Slovakia, from various types of sites. Ring Bowls (Ringschüssel). This is a very common Roman form popular with Germans, with examples from almost every German settlement dating from the second to third centuries AD (Krekovič 1981, 364; Kolník 1998b, 75; Krekovič 2000b, 260). These bowls get their name from the outward projection of clay at the join between neck and body (Fig: 9.4:1–4), which is different from the collared bowls mentioned above. The form is also seen among the Brick-Colored and Marbled Wares (Krekovič 1981, 364; 1998, 29). In regards to the Smooth-Gray Ware versions, these vessels often have a slip or their surface has been colored by the smoke from the firing process (Krekovič 1981, 364). These bowls are sometimes decorated by a toothed wheel. There are many variants to the ring bowls, depending on the shape of the bowl and placement of the ring(s) (Krekovič 1981, 364). In Moravia, M. R. Pernička identified five bowl types (A, B1–B4) of which A is of the Early Roman Period having a low profile and wide shape, while B1–B4 belong to the Later Roman Period (Krekovič 1981, 364). In Hungary, E. Bónis identifies their origin in terra-sigillata forms, but E. B Thomas, having identified some similarity to Drag. 24/25, notes late La Tène influence as well (Krekovič 1981, 364). It appears that the bowl type was so popular with Germans that Roman production was geared toward satisfying this market (Krekovič 1997a, 233).

In comparison to Germans in southwest Slovakia, the Iazyges living east of the Danube (Fig. 9.1), who also would have had similar access to provincial markets on the limes, did not import mortaria in the same number (Krekovič 1981, 357; 2000b, 262). Krekovič, using the data available in 1981, mentions that while 43 examples have been found in southwest Slovakia, only two come from the Sarmatian Iazyge territories, one of which was found in the grave of a German warrior (Krekovič 1981,

269

Conical (Tapered) Bowls. These bowls have a thickened beveled rim below which is a border or ring (Fig. 9.4:5 and 6). In fact, these bowls are sometimes hard to distinguish from ring bowls, a difficulty further compounded by the existence of transitional forms between the two and other bowls that have a third ring in the upper register (Krekovič 1981, 365–366). These bowls are not as widely found as ring bowls (Krekovič 1981, 366).

9.1.7. Coarse-Gray Wares Even though the type is labeled as gray, the surface of these vessels can have brownish, reddish, to blackish grays, while the fabric can even be whitish (Krekovič 1981, 366; Gabler 1989, 484; Kolník 1998b, 75; Krekovič 1998, 30). Often, the surface of these vessels has been blackened by fire/smoke. The fabric can be similar to the Smooth-Gray Wares, but often there is an added assortment of sand and/or pebbles in great quantity, which makes this group easily identifiable because of the rough surfaces (Gabler 1989, 484; Kolník 1998b, 75; Krekovič 1998, 30). Common forms are Beakers with Impressed Walls and Containers with Horizontal Rims, discussed below as separate groups (see 8.1.8 and 8.1.9), along with plate-shaped bowls with inturned rims (Fig. 9.5:1–4) and pitchers (Fig. 9.5:5 and 6) (Kolník 1998b, 75–76; Krekovič 1998, 31). CoarseGray Wares were made throughout the Roman Period but, besides the beakers and containers with horizontal rims, are rarely found in Slovakia (Krekovič 1981, 366). Krekovič believes that this ware may have suffered a similar fate as the glazed wares, where many of them may have been misidentified as medieval (Krekovič 1981, 366). The few identified sherds have mostly come from pitchers (Krekovič 1981, 366).

Hemispherical bowls. These vessels have thick rims and their forms imitate the Drag. 37 type (Fig. 9.4:7) (Krekovič 1981, 366; Gabler 1989, 472). The fabric and decoration of these bowls are similar to ring bowls, and often have fine channels around the vessel made from toothed wheels (Krekovič 1981, 366; Gabler 1989, 472). As in the case of ring bowls, these bowls have a wide variety of forms, with some even losing their spherical aspect, having a sharp carination between neck and body (Krekovič 1981, 366). Pots and Pitchers (Jugs). The fine gray-ware pots and pitchers are similar in form and appearance to their counterparts in the Brick-Colored Wares (Krekovič 1981, 366). 9.1.6. Pannonian-Stamped Wares

9.1.8. Beakers (Cups) with Impressed Walls and ThinWalled Mugs

This is a typical ware of Pannonia having gray vessels made of well-levigated clay, some having a gray or black slip, decorated with stamped ornamentation of ordinary vegetal motifs, rouletting, or rosettes (Fig. 9.4:8) (Bónis 1980, 357–358; Krekovič 1981, 350). As with Raetian Wares (see section 8.1.11), many of the forms and decorations used on the stamped wares have their origins in the La Tène period for the region (Bónis 1980, 357– 358; Gabler 1989, 471; Krekovič 1981, 350). As mentioned before, this ware is a subtype of Smooth-Gray Wares, with the only difference being the use of the stamped decorations (Krekovič 1998, 13).

The beakers are a widespread Roman pottery form, in which a beaker has the walls pressed in at regular intervals creating ovular impressions along the middle register—also known as Faltenbechers (Fig. 9.5:7–9) (Krekovič 1981, 357; Gabler 1989, 478). The design was probably influenced by glassware with similar decoration. In the Rhineland, the vessel type appears under the Flavians and expands from there throughout the empire, lasting until the end of the Roman Period (Krekovič 1981, 357). Because of their long usage and many variations in appearance and technical construct, it is hard to create a chronology (Krekovič 1981, 357). These vessels are considered a subgroup of Coarse-Gray Ware as already mentioned because of their coarse fabric and gray to black color. There are also some examples that fall under the category of Brick-Colored Wares, but occur less frequently than the gray examples; those few that have been found correspond in form to the tall beakers mentioned below (Krekovič 1998, 26). In regards to Pannonia, there was a widespread tendency to say that smaller vessels or ones that are shorter, with slips and made of a fine brick- or gray-colored material are the older type and the taller forms made of a rougher grayblack material are the younger type (Krekovič 1981, 358). This view has changed with the discovery of tall beakers of rough material from the older period as well. Production centers at Aquincum were already producing tall beakers with impressed walls by the first half of the second century AD (Krekovič 1981, 358). It seems that the tall vessels with rough surface appear throughout the Roman Period, but the finer short forms disappear by the

These vessels were thick-walled, with bowls and pots the most popular forms (Bónis 1980, 357). The first seminal work on Pannonian-Stamped Wares was by L. Nagy (1928) and later expanded by K. Póczy (1957) (Krekovič 1981, 350). A more recent work was written by E. Maróti (1991), who identified many local variants as well as the well-known manufacturers (Krekovič 1998, 13). The first identified producer of this style in Pannonia was the potter Resatus and his apprentices, working AD 75–135. The height of stamped-ware production was in the midsecond century, and by the mid-third century the style was in decline (Krekovič 1981, 350–351; Kolník 1998b, 75). For the barbaricum, southwest Slovakia has the highest concentration of stamped wares, with many examples from various types of sites (Krekovič 1981, 352). Pannonian-Stamped Wares found in Slovakia were first studied as a group by T. Kolník (1958) and much has been added since (Krekovič 1981, 352). The ware was popular among the Germans, who even created handmade imitations (Krekovič 1981, 353; Kolník 1998b, 75). 270

third century (Krekovič 1981, 358). In the barbaricum not only are there direct imports of beakers with impressed walls, but Germans also produced their own imitations. In Slovakia, mugs with impressed walls are found often both in settlements and cemeteries (Krekovič 1981, 359).

9.1.10. Barbotine Ware The pottery has applied decoration on the surface of the vessels such as leaf motifs, vines, and small balls, which probably represented grapes or berries (Fig. 9.6:6). The vessels were made to imitate metal forms first appearing in the Hellenistic Period and later Roman Period glassware (Krekovič 1981, 342; Hayes 1997, 66). For Slovakia, the term Barbotine Ware is used for a specific group of thin-walled bowls, cups, and flagons made of a fine-grey, mixed with brick-red, material, usually having a grey, red-to-brown surface (Krekovič 1981, 342). For Pannonia, E. T. Szőnyi (1972) published a study of the form and types of decoration, but these are not chronologically significant (Krekovič 1981, 342). Based on the description, these vessels fit into a category that J. Hayes labels as Thin-Walled Wares, which were made to compliment terra sigillata for table use and are also labeled as such by Bónis (Bónis 1980, 360; Hayes 1997, 66). They originated in Northern Italy according to Hayes probably in the second century AD, but based on the findings from Hungary and Slovakia, it would appear that they first start being made in the later half of the first century and continue into the second century, by which time they were solely made for export to the barbaricum (Bónis 1980, 360; Krekovič 342; Hayes 1997, 66). More recent studies have shown that already in the first century there were also production centers in Pannonia (Krekovič 1998, 7). The vessels are rare finds in the empire and in the barbaricum they first appear along the Amber Route then along the limes (Krekovič 1981, 342; 1998, 7). The most popular form found in Slovakia is a two-eared cup, similar in shape to glazed skyphoi examples mentioned below (Krekovič 1981, 342).

Thin-walled pottery was a popular type in the Danubian provinces, with most imports coming from northern Italy, probably Cremona and Ravenna (Gabler 1986, 99). In Pannonia, a specific thin-walled ware, the thin-walled mugs with a bronze-like shine, mostly made with a brown slip, are usually identified as imports from the western provinces (Krekovič 1981, 358; Gabler 1989, 479). This identification has been challenged by some scholars, with problems providing adequate evidence of local production, but a kiln site at Savaria (Fig. 9.1) has yielded many distorted vessels with the bronze-like surface (Gabler 1989, 479; Krekovič 1998, 9). It is also accepted that production took place at Carnuntum sometime after AD 100, most likely influenced by the arrival of Legion XIV from the Rhine country (Krekovič 1998, 9). The surface of these vessels can also be roughened by the sprinkling of fine sand across the surface on the slip before firing; the technique is first seen in the Rhineland during the Flavian Period, lasting to the end of the Roman Period (Bónis 1980, 362; Krekovič 1981, 358–359).6 In comparison to beakers with impressed walls, thin-walled wares are a rarer find in Slovakia (Krekovič 1981, 359). 9.1.9. Containers with Horizontal Rims These vessels have a characteristic wide horizontal rim, which is decorated with grooves (Fig. 9.6:1–4) (Krekovič 1981, 359). Below the rim, the vessels can be decorated with dense groupings of grooves or wavy lines (Fig. 9.6:5), while the color of the vessels can vary greatly from brown to gray, sometimes with the surface grayed through smoke (Krekovič 1981, 359–360; Kolník 1998b, 76). The fabric is coarse, often mixed with sand, while at the rim of the vessels there is sometimes an applied coating of black pitch (Krekovič 1981, 360). The pitch coating is thought to have either been used to ward off insects or to seal the vessels (Krekovič 1981, 360). There is evidence that some of the vessels were finished in the barbaricum itself (a kiln and wasters were found at CíferPác), which makes it difficult to determine which vessels were the imports (Krekovič 1981, 360; Kolník 1998b, 76). This ware is a subgroup of Coarse-Gray Wares. The vessels appear to have La Tène origins and their use was at first thought to be only in the first two centuries (primarily second) of the Roman Period (Krekovič 1981, 360; Gabler 1989, 485). In Pannonia these vessels are often found at villa sites, some dating to the third century, thereby pushing the manufacturing of these vessels also into that century (Krekovič 1981, 360). A great number of these bowls have also been found at the Carnuntum legionary fortress (Gabler 1989, 485). In Slovakia, various find spots have been identified (Krekovič 1981, 360).

9.1.11. Raetian Wares The vessels making up this group have the characteristic decoration of barbotine geometric motifs and notches made by toothed wheels first seen in the province of Raetia (Fig. 9.6:7–11) (Krekovič 1981, 342). The vessel fabric is reddish in color, always having a glossy/shiny surface (which can appear bronze-like), ranging in color from red-brown to red (Bónis 1980, 362; Krekovič 1981, 342, Droberjar 2002, 126). The origins of these vessels are based on local La Tène traditions in Raetia which gradually became mixed with Roman ones (Krekovič 1981, 342). Raetian wares in Pannonia were first collated by E. Bónis (1960) and E. T. Szőnyi (1973) (Gabler 1989, 466). The most popular form found in Slovakia is a small pear-shaped pot with recessed feet. The barbotine decorations can be geometric designs similar in appearance to the letter X, rosettes, crescents, beads, as well as patterns made by scratching into the surface or using toothed wheels (Droberjar 2002, 126). The typology first created by Drexel (1911, 80–84) is still usable today with three styles evident based on the decorations: 1) thick geometric decoration, first half of second century AD; 2) application of plastic horseshoe/crescent shapes in rows and also rows of incised lines, second half of second century AD; and 3) 271

decoration made by toothed-wheel, first half of third century AD (Krekovič 1981, 342–343; Gabler 1989, 466; Droberjar 2002, 126). There is however a more recent finer organization proposed by T. Fischer (1983/1984) (Krekovič 1998, 10). At first these vessels were directly imported from Raetia, but by the second half of the second century, local Pannonian potters were imitating the Raetian vessels, with production at Poetovio, Savaria, Vindobona, Aquincum, and Gorsium (Fig. 9.1) (Krekovič 1981, 343; 1998, 10). There is some debate, but it is thought the true Raetian imports have wall thickness not exceeding 2–3 mm according to E. T. Szőnyi, but D. Gabler accepts up to 4 mm (Krekovič 1998, 10). In Slovakia, these vessels are a rare find, mostly discovered near the limes (Krekovič 1981, 343).

likely problem is that many glazed sherds from Late Roman sites have been misidentified as Medieval in date; 2) weak importation by the third and fourth centuries; but this does not seem to work since glazed mortaria importation remains high throughout the period; and 3) it is possible that Germans only wanted the glazed mortaria, not having a particular fondness for the other forms (Krekovič 1981, 347–348). 9.1.13. Terra Sigillata The final group of pottery deals with terra sigillata. Even though this pottery is a fine ware and one would expect it to be a very rare find in the barbaricum, the reality is far different. As of 1997, 1,246 fragments had been found at 114 sites (these numbers have gone up since then), divided among settlements, cemeteries, and the Romanstyle buildings (Kuzmová 1997, 11). This concentration of terra sigillata is the highest found in the barbaricum (Krekovič 1997a, 233). When the find-spots are mapped, their distribution is seen to follow along the valleys of the major waterways, Morava, Váh, Nitra, and Hron Rivers; there is a denser concentration in the area of Trnava (Kuzmová 1997, 11). These waterways were the major trade routes and it is not surprising to see concentrations along them. The majority of the find spots are within 50 km of the limes (Kuzmová and Roth 1988, 144).

9.1.12. Glazed Wares Glazed wares have a long history in the Mediterranean region and the Roman examples appear to imitate earlier Hellenistic forms (Bónis 1980, 361; Krekovič 1981, 344). The start of Roman glazed wares dates to the principate of Augustus in Italy, and there is evidence to show it started in the provinces between 30–25 BC as well (Bónis 1980, 361; Krekovič 1981, 344). Glazed wares were first imported into Pannonia in the first century AD and came from Northern Italy, but by the second century there were local Pannonian production centers (Bónis 1980, 361; Gabler 1986, 96–97; Krekovič 1981, 344–345; Hayes 1997, 64). There are roughly two periods for glazed wares: 1) dates to the first century AD with yellow glazed small cups being the most popular; the glaze itself was lead based; 2) dates to the third to fifth centuries AD, with the fourth century seen as the height of production in Pannonia; the glaze had more copper in it and the quality deteriorated over time so that by the fifth century the glaze does not hold well to the vessels (Krekovič 1981, 344–345). Popular forms in Pannonia during the second period are vessels with crescent ornaments placed in vertical rows on the body, one-handled jugs, beakers, bottles, mortaria, and there are also rare examples of anthropomorphic vessels (Krekovič 1998, 38–40). In the second period, glazed mortaria take over as the most commonly found glazed ware in Slovakia.

Of the 1,246 fragments, 1,005 come from German sites, while the remainder (241) comes from the Roman-style structures (Bratislava-Devín, Bratsilava-Dúbravka, Stupava, and Cífer-Pác) (Kuzmová 1997, 16). In regards to those found at German sites, a large portion (98.4%) of the sherds comes from settlements, with only 1.6% from graves (Kuzmová 1997, 16). Based on the evidence, almost every major settlement has yielded some amount of terra sigillata in southwest Slovakia (Krekovič 1987, 250). Most of the fragments (54.5%) are from dishes having relief decorations, while 28% are from undecorated vessels and 17.5% could not be identified (Kuzmová 1997, 16). The majority of terra sigillata can be identified as South Gaulish, Rheinzabern, Westerndorf, and Pfaffenhofen (Kuzmová and Roth 1988, 130–139; Kuzmová 1998, 16). These types mirror the terra sigillata found in the province of Pannonia, where by the second decade of the first century AD, terra sigillata was coming from southern Gaul, primarily Montans, Banassac and La Graufesenque (Bónis 1980, 362). By the second century, imports from Lezoux in Central Gaul had gained dominance and, as the market in Pannonia strengthened, the factories moved eastward so that by the middle second century there were factories at Rheinzabern, while in the late second and third centuries factories opened at Westerndorf and Pfaffenhofen (Bónis 1980, 362; Kuzmová and Roth 1988, 138–139).

There are examples in Slovakia of glazed vessels with relief decoration rarely found even in the provinces (Bónis 1980, 361; Krekovič 1981, 345). Three very well preserved early examples were discovered in graves, two being green-glazed skyphoi (Fig. 9.6:12) dating to AD 35–50 and the third a green-glazed kantharos (Fig. 9.6:13) dating to AD 50–80, with all three having relief decoration of leaves (Krekovič 1981, 345). These vessels need to be seen as luxury vessels, alternatives to silver and gold vessels (Krekovič 1981, 346; Hayes 1997, 64). The remaining material from various sites is made up of small fragments. There is a low number of glazed examples dating to the Late Roman Period, even though glazed forms reached their height of popularity in the fourth century in Pannonia (Krekovič 1981, 347). Krekovič provides three possible reasons: 1) the most

It should be mentioned that the oldest terra sigillata (79 pieces) found in southwest Slovakia were found in the Roman-style structures at Bratislava-Devín, and were 272

primarily from Arretine workshops in Italy dating to the Augustan period (Kuzmová 1997, 42 and 53). The location of Bratislava-Devín at the juncture of the Morava and Danube Rivers, near Carnuntum (c. 10 km), makes the site a possible anomaly and probably should not be considered part of Quadi territory. The site is thought to have been Roman controlled by the late Augustan period as a strategic strong-point at the confluence of two important rivers (Kolník et al 1993, 242–243; Kuzmová 1997, 41–42; Pieta and Plachá 2000, 6–7).

9.2. Catalog of the Roman Pottery from Urbárske Sedliská The majority of sherds identified as Roman were heavily worn body sherds that could not be matched to specific vessel shapes. Out of the 84 Roman sherds, 11 are diagnostic and out of those only six were found in the four features (one from Hut 1, four from Hut 2, and 1 from G3). Since the number is small, the material will be presented below as a whole for the entire site, and the huts will not be treated separately from the other units. The format for the presentation of the Roman sherds will have them organized by type, such as Brick-colored and Pannonian Striped Ware. Only the more important pieces will be cataloged individually, with each piece’s description regarding color, feel, inclusions, and texture following the terminology used for the German pottery in Chapter 8, except there will be no “Form” line (see Table 8.4). The numbering of pottery continues from Chapter 8.

The oldest terra sigillata (24 pieces) found at German settlements in southwest Slovakia were from south-Gaul, dating to the late first and early second century AD (Kuzmová and Roth 1988, 130; Kuzmová 1997, 16). Stamps have been found from the master Mercata working out of La Graufesenque and from the late southGaul potter, master Natalisa working out of Banassac (Kuzmová and Roth 1988, 130). The greatest amount of terra sigillata (843 pieces) from German settlements, however, dates between the middle second and early third centuries, coming primarily from Rheinzabern (401 pieces) followed closely by Westendorf (335 pieces) (Krekovič 1987, 250; Kuzmová and Roth 1988, 132; Kuzmová 1997, 32: Table 8). If we combine the totals found from German sites with those pieces from the Roman-style structures, Rheinzabern still dominates (555 pieces) as the most commonly found, but Westendorf falls off slightly (380 pieces) (Kuzmová 1997, 32: Table 8, 53: Table 14).

9.2.1. Brick-Colored Wares Out of the 84 sherds recovered, 54 could be identified as Brick-Colored Wares, clearly making this group the dominant type at the site (see Table 9.1 and Chart 9.1, located at the end of the chapter). A majority of the diagnostic pieces (8) come from this group as well. 144. Base. Inv. No. 16-3-3. Ft. G1. Fig. 9.7:144. B.D. 15 cm. P.L. 5.8 cm. Th. 7 mm. W. 26 g. O., I., and C. very pale brown (10YR 7/4).

In regards to the most popular forms, Drag. 37 dominates the archaeological record from German settlements (532 pieces), with Drag.18/31 (128) and Drag. 32 (71) distant second and third in quantities (Kuzmová 1997, 32: Table 8). It is interesting to note that this proportion holds true with the Roman-style structures as well, with Drag. 37 at 130 pieces and Drag. 18/31 a very distant second at 24 pieces (Kuzmová 1997, 54: Table 16). Drag. 32 at the Roman-style structures is represented, however, by only 9 pieces, which is surpassed by Drag. 43 (10 pieces) and Drag. 54 (12 pieces) (Kuzmová 1997, 54: Table 16).

Base of what was possibly a pitcher/jug, based on the lip at the bottom and thickness of the body. The outer surface is smooth and the texture is fine. There is a moderate amount of fine sand and mica inclusions. 145. Handle. Inv. No. 348-3-2. Fig. 9.7:145. P.L. 4.4 cm. Th. 1.3 cm, while the body of the vessel was 7 mm. Handle width is 3.85 cm. W. 40 g. O. and I. reddish yellow (7.5YR 7/6) and C. reddish yellow (5YR 6/6).

The amount of terra sigillata found in southwest Slovakia is even more pronounced when comparisons are made with other areas of the barbaricum. In the territory of the Sarmatian Iazyges, much like with the mortaria mentioned above, terra sigillata is found in lesser quantities. Only two pieces (one from northern Italy and one from Gaul) have been found, both dating to the end of the first and beginning of the second century, compared to the 152 pieces from southwest Slovakia for the same period (Kuzmová and Roth 1988, 143; Kuzmová 1997, 32: Table 8). For the mid-second to early third centuries, when Rheinzabern is found in such large quantities in southwest Slovakia, only 39 pieces have been discovered in Sarmatian territory (Kuzmová and Roth 1988, 136: Table 8). Even in comparison to Quadi sites further north in Moravia there is a marked difference, with only 411 pieces having been discovered at 65 sites there (Kuzmová 1997, 14: Table 1).

Handle of a pitcher or jug with a portion of the body preserved. The handle is made up of three coils of clay that have been joined and smoothed to create two grooves on the outer side. There are remnants of the slip (red, 10R 5/8) where the handle joins the body. The outer surface is rough to the feel, probably because of weathering, while the texture is fine. There is a moderate amount of mica and sand inclusions. 146. Rim. Inv. No. 736-4-3/4. Ft. G4. Fig. 9.7:146. D. 9 cm. P.L. 4.5 cm. Th. 5.5 mm. W. 10 g. O., I., and C. light red (2.5YR 6/6). The rim of a probable pitcher or jug, but not enough survives for precise identification. A slip was applied to the vessel and is red (2.5YR 5/6). The outer surface is smooth and the texture is fine. There is a moderate amount of sand and mica inclusions. 273

The shape of the out-turned lip bares some resemblance to the jugs mentioned by Krekovič as belonging to Gassner’s Type (8–11): Jugs with out-turned rims and wide mouths (Krekovič 1998, 21, Taf. 6: 4 and 10). These jugs are found in all four periods (Krekovič 1998, 21) and therefore cannot be used for dating purposes.

Rim sherd of what was probably a pitcher. The mouth has a horizontal rim and below there is a raised border. There is a remnant of the slip, but badly worn, leaving a color of reddish yellow (5YR 6/6). The surface feels smooth and the texture is fine. There is a sparse amount of sand and mica inclusions.

147. Rim. Inv. No. 740-4-3/4. Ft. G4. Fig. 9.7:147. D. 16 cm. P.L. 3.8 cm. Th. 5.5 mm. W. 5 g. O., I., and C. light brown (7.5YR 6/4).

There is some similarity in the shape of the rim and the placement of the ring below the rim with a pitcher from Gerulata (discovered at the fort, but no assigned date), belonging to Gassner Type (2–5): Jugs with singleprofiled rims, but that one has a mouth diameter of only 6 cm, making it half the size of sherd 150 (Krekovič 1998, 20, 62, T. 5:8). Based on the mouth width, sherd 150 is more suitable in Gassner Types 8–11: Jugs with outturned rim and wide mouth, and so may just be a wider mouth variant of the Gerulata jug.

The rim fragment of a bowl, pot, or possibly pitcher/jug (see below). A slip was applied and on the outside is a light brown (7.5YR 6/4), while on the inside a reddish yellow (5YR 6/6). The outer surface is smooth and the texture is fine. There is a moderate amount of mica inclusions. The sherd possibly belongs to a single-piece hemispherical bowl with horizontal rim of the type mentioned by Krekovič (1981, 361–362). The shape of the rim, however, with its squared edge, has no parallel found by the author among bowls. There are strong similarities to sherd 150 (rim from a pitcher) listed below. It is possible that the width measured is too great and may be more in line with that of sherd 150, thereby making sherd 147 belong to a pitcher as well.

9.2.2. Pannonian-Striped Wares This is the second group in number (15 pieces) found at the site. Even though none was diagnostic, three body sherds are presented here: two because of their decoration and the other for both size and decoration. 151. Body. Inv. No. 744-4-3/4. Ft. G4. Fig. 9.8:151. P.L. 4.5 cm. Th. 4 mm. W. 5 g. O. and O.M. pink (7.5YR 7/4), I.M. light red (2.5YR 6/8) and C. gray.

148. Rim. Inv. No. 743-4-3/4. Ft. G4. Fig. 9.7:148. D. 9 cm. P.L. 6 cm. Th. 5.5 mm. W. 10 g. O., I., and C. light red (2.5YR 6/6).

Body sherd of an indeterminate vessel. Outer surface has two horizontal incised lines and two rows of vertical notches as decoration. The outer and interior surfaces are slipped, being red (2.5YR 5/6). The outer surface is smooth and the texture is fine. There is a moderate amount of sand and mica inclusions.

Rim fragment of a pitcher or jug. Very little survives which hampers a more precise identification. There appears to have been a red slip applied to the vessel (2.5YR 5/6). The outer surface feels smooth, while the texture is fine. There is a moderate amount of sand and mica inclusions.

152. Body. Inv. No. 885-4-2. Fig. 9.8:152. P.L. 4.2 cm. Th. at middle 4 mm, at bottom 6 mm. O. and C. light red (10R 6/6) and I. pink (7.5YR 7/4).

The shape of the rim and width of the opening bear some resemblance with sherd 146 and Gassner’s Types 8–11 mentioned above.

Body sherd preserving the decoration of a vessel that was possibly a container. The decoration shows the band of slip applied to the body, which is divided into two registers. The upper register has a wavy line incised into the paint, while the second register has rows of incised vertical lines. The slip is red (10R 5/8) in color. The outer surface is smooth and the texture is fine. There is a sparse amount of sand inclusions.

149. Base. Inv. No. 745-4-3/4. Ft. G4. Fig. 9.7:149. B.D. 9 cm. P.H. 8.7 cm. Th. 9.5 mm. W. 80 g. O. reddish yellow (5YR 6/6), I. yellowish red (5YR 5/6), O.M. and I.M. yellowish brown (10YR 5/6), C. red (2.5YR 5/6). The base of a probable pitcher, jug, or pot. The outer surface is moderately pockmarked, while on the interior surface near the edge of one break is an empty space that appears to have been there when fired, which may indicate the use of an organic temper or other material that has since disappeared. There appears to have been no slip. The outer surface is smooth and the texture is fine. There is a moderate amount of mica inclusions.

The decoration on sherd 152 is the same type seen on a vessel from Gerulata. The vessel is a hemispherical container with horizontal rim and on the body there is a band of paint in which is inscribed a wavy line, below this is another band of paint that has rows of incised vertical lines (Krekovič 1998, 62, T. 7:4). Krekovič remarks that the form is typical and in Carnuntum they cover a period from the middle of the first century up to the first third of the third century, without strong morphological changes (Krekovič 1998, 17).

150. Rim. Inv. No. 886-4-2. Fig. 9.7:150. D. 12 cm. P.L. 4.3 cm. Th. 5 mm. O. and I. pink (7.5YR 7/4), C. yellowish red (5YR 5/8).

274

153. Body. Inv. No. 956-4-3/4. Ft. G4. Fig. 9.8:153A and 153B. D. at belly 14 cm. P.L. 12.4 cm. Th. 5 mm. O. and C. reddish yellow (7.5YR 6/6) and I. reddish yellow (5YR 6/6).

9.2.6. Beaker with Impressed Walls and Thin-Walled Wares Two small body fragments were identified as belonging to this group. The identification was based on the thinness of the walls and the surface treatment of the sherds. A sherd from Hut 1, Inv. No. 10-3-3, has a shiny bronze-like slip with fine sand particles in it, possibly identifying it as a thin-walled cup fragment. The second, a sherd from Hut 2, Inv. No. 747-4-3/4, has a very dark gray- to black-slipped surface with black sand particles. It may come from a beaker with impressed walls, but being so small positive identification is not possible.

Two large sherds belong to the belly of a pot. The outer surface has the characteristic belly slip which has been divided by horizontal incised lines, while the separate slipped bands have rows of notches as decoration. The slip is a red (10R 5/8). The slipped area is smooth to the feel and the texture is fine. There is a sparse amount of fine sand and mica inclusions. Sherd 153 probably belongs to the same type of vessel, a hemispherical container with horizontal rim, mentioned in the previous entry, 152. A similar pot was found at Trnava-Hrnčiarovce, dating to between the 2 and first half of the third century (Krekovič 1981, 354, Obr. 1:9). The vessel from Trnava-Hrnčiarovce is more complete with surviving rim, below which are two grooves in the vessel and then the band containing the rows of incised lines. The incised lines have a slant to them, from left to right. Sherd 153 has a similar pattern on the body, and may have been drawn upside down.

9.3. Discussion of the Roman Pottery from the Site in Comparison to That of Other Sites As is readily apparent, the Roman pottery from the site is not of the best quality and the quantity is small. Pannonian-Striped Ware and possible Thin-Walled Wares indicate the presence of some table ware, but the majority of sherds belonged to day-to-day vessels. As presented in the discussion of Roman pottery in section 8.1, one expects to find a mixture of forms, such as pitchers, bowls, pots, and cups. The material from the two Huts and the site overall roughly falls in with that expectation. The distribution of the pottery and percentage representation can be seen in Table 9.1 (presented at the end of the chapter) and Charts 9.1–8.3. Even though many of the sherds are too fragmentary, the overall impression is that of a standard representation of Roman pottery usually discovered in German settlements. Pitchers/Jugs were the most common form found, at 7% of the total, with pots/bowls, mortaria, and beakers/flagons making up the remainder of identified forms, while the unknown sherds made up 85% (see Chart 9.3). If we remove the unknown group, then Pitchers/Jugs comprise 47% of the group, with Pots at 23%, Beakers/Flagons 15%, and Mortaria at 15%. The dominance of the Pitchers/Jugs may only be a reflection of the survival of diagnostic sherds that could be ascribed to this group. Many of the unidentifiable sherds probably belong to pots/bowls.

9.2.3. Mortaria Two fragments were identified at the site, but only one, Inv. No. 474-3-3-G3, came from a feature, the ditch portion of G3. The other fragment, Inv. No. IM 37, came from stratum 2 of Unit 3, so that it may have originated from the fill of Hut 1. Both are unglazed and very thick (over 1 cm), therefore probably dating to the second century. 9.2.4. Smooth-Gray Ware Two sherds were identified from this group, both from Unit 4: one found in stratum 1 and the other in Hut 2. A sherd, Inv. No. 894-4-1, from stratum 1 is a very small body fragment, but interesting in the dark gray fabric and evident black slip. The sherd’s appearance is more reminiscent of terra nigra (Black Arretine), but too little survives for better identification. The other sherd, Inv. No. 734-4-3/4, came from Hut 2 and is also a very small body fragment, with what appear to be two lighter lines decorating it.

When the findings from Huts 1 and 2 are compared to Roman pottery from the same group of huts used in Chapter 8 in regards to the German pottery, we see that there is no significant difference in the Roman pottery numbers and types found at Urbárske Sedliská (see Table 9.2). Huts 1 and 2 from Urbárske Sedliská are at the low end of the amount of Roman pottery present at 4% each, but they are not unique, with the Dúbravka, Blučina and Rajhrad huts having even less. The highest totals were seen in Huts 2 and 7 from Křepice, where the Roman pottery made up 31% and 40% of the totals respectively. As can be seen in Table 9.3 and Chart 9.4, Brick-Colored Wares make up the largest group (343 pieces) of Roman pottery found in the various huts, at 76.6% of the total, followed at a distance by Pannonian-Striped Wares (56 pieces, 12.5% of the total).

9.2.5. Coarse-Gray Ware Nine body fragments were identified as belonging to this group. Their identification was based on the interior color of the vessel because the outer surface was light brown in color. Eight sherds came from Hut 2, of which three were some of the largest Roman pottery fragments found during the excavation, and were from the same vessel, which was probably a pitcher or other large vessel.

275

As may be seen in Table 9.4, the majority of sherds from Huts 1 and 2 belonging to pitchers/jugs at 47% is not unusual. Overall, pitchers/jugs make up 43% of the total at 35 pieces, with bowls at 37%, pots 14%, and storage vessels 6%. What is surprising is the lack of bowls, which seem to be more common at the other sites and huts used in the comparison (Table 9.4).

Table 9.4. Vessel Forms Present at the Various German Huts Hut Bowls Pots Pitchers/ Storage Jugs Vessels 1 Hut 1 1 4 Hut 2 6 7 9 US82 Blučina 1 1 4 Komořany 3 2 Z-9 Křepice 10 6 5 2 Křepice 10 3 12 7 Rajhrad 2 30 11 35 5 Totals

Table 9.2. Pottery Present in the Various German Huts Hut German Roman Other 493 18 3 Hut 1 661 28 6 Hut 2 753 39 0 US82 Dúbravka 503 3 6 13/88 Dúbravka 282 0 8 129/93 Blučina 211 17 0 4 Komořany 196 31 0 Z-9 Křepice 308 141 0 2 Křepice 249 168 0 7 Rajhrad 130 5 0 2

9.4. The La Tène Pottery As mentioned in Chapters 3.1 and 8.1, the Germans started to move into Celtic territory in southwest Slovakia during the very last years of the first century BC and the first quarter of the first century AD. Celtic groups were still in the area and during this period, the German settlements and some material culture copied Celtic types, such as hut architecture – Kolník’s Type II form (see Chapter 7.1) (Elschek 1995, 39; Kolník 1998b, 60). German pottery of the Early Roman Period also showed Celtic influence, as in the form of some barrel-shaped pots, amphorae, jugs, and various decorations (Kolník 1998b, 73). There were also influences from further abroad, not only Roman, but also from the Celto-Dacian groups of eastern Slovakia and also the Przeworsk culture of southern Poland (Kolník 1998b, 73).

Table 9.3. Frequency of Occurrence of Roman Pottery Types Found in German Huts Hut BC PS MW FG CG B G TS 13 4 1 Hut 1 15 4 1 8 1 Hut 2 30 2 2 4 1 US82 Blučina 14 3 - 4 Komořany 25 4 - 1 1 Z-9 Křepice 113 14 6 6 - 2 2 Křepice 131 27 1 6 1 - 2 7 Rajhrad 2 1 2 - 2 343 56 3 18 19 3 1 5 Totals BC = Brick-Colored Wares B = Bearkers/Thin-Walled PS = Pannonian-Striped Wares Wares MW = Marbled Wares FG = Fine-Gray Wares CG = Coarse-Gray Wares G = Glazed Wares TS = Terra Sigillata

9.4.1. Late La Tène Pottery in Southwest Slovakia Celtic pottery by the Late La Tène period showed a high quality of manufacture, produced by skilled craftsmen in the Bratislava region (Pieta and Zachar 1993, 157). Much of the pottery was made on the potter’s wheel, but many coarse-ware vessels were still handmade (Kolník 1998b, 73). In regards to type, utilitarian ware dominated the record, being used for a wide variety of uses in the household and other places. The pottery differed from Middle La Tène in that the marked horizontal partitions of vessels that indicated the way of manufacture through circular ribbing disappeared along with pronounced sshaped vessels and stamped decorations (Pieta and Zachar 1993, 157). The Late La Tène pottery showed a marked strengthening of the vessels’ mouths and pronounced everted rims. The surfaces of the vessels were polished, and a popular type of decoration was a colored slip. There was also a growth in the variety of shapes for specific uses.

276

The typology of La Tène pottery has at least ten types of ceramic shapes (Pieta and Zachar 1993, 157). (1) The first type is large cooking pots with strengthened rims, decorated on their bodies with comb lines going down the body, while the clay often has graphite mixed in, giving a shiny/sparkle appearance (Fig. 9.9:1–3) (Pieta and Zachar 1993, 157). Other vessels of this shape sometimes have horizontal banding for decoration instead of comb lines. (2) The second group is containers, with a very wide body, sometimes described as egg-shaped, with a narrow neck and out-turned rim (Fig. 9.9:4) (Pieta and Zachar 1993, 157). Some of these are decorated in horizontal bands of colored slips. (3) The third type is a barrelshaped vessel, having a slip decoration on the body (Fig. 9.9:5–7). (4) The fourth type is comprised of cups either barrel shaped or in the form of a chalice (Fig. 9.9:8). Most of these are made of well refined clay and are polished. (5) The fifth type includes a wide variety of bowl shapes (Fig. 9.9:9–11). A very common bowl has the lip turned slightly inward (Fig. 9.9:9 and 10), while another form has the lip flattening out and forming a wide rim to the bowl. Another bowl form is called Békásmegyer, with a strongly defined S-shaped form to the body, and a thick rim (Fig. 9.9:12) (Pieta and Zachar 1993, 160). Other shapes were biconical or even globular. (6) The sixth form is a tea-pot shaped vessel on three legs, sometimes made of clay with added graphite, with the legs sometimes decorated in cross-hatching (Fig. 9.10:1–3 only show the legs) (Pieta and Zachar 1993, 161). (7) The seventh form includes large vessels narrow in shape with reinforced rims and showing on the interior that they were wheel-turned (Fig. 9.10:4) (Pieta and Zachar 1993, 161). On the body handles are sometimes placed to help in transport. Many are made with clay that has grog mixed in, while the surface of the vessel might be either left rough or polished. (8) The eighth type includes various lids that can fit on some of the vessels already mentioned (Fig. 9.10:5). The lids have a lip to help them sit well on the vessels and have a handle for ease of use. (9) The ninth type includes shallow bowls or globular forms which have their bottoms punctured to create colanders or were used for making cheese (Fig. 9.10:6). (10) The final form is a conical-shaped cup with an ear-shaped handle placed below the rim and above the base (Fig. 9.10:7 and 8). This form appears in the Late La Tène and shares the form with a cup type seen among the Dacians during this period, being hand made and rougher, and showing the mixing of cultures in the region by the last half of the first century BC (Pieta and Zachar 1993, 163 and 165).

154. Rim. Inv. No. -3-1. P.L. 5.7 cm. Th. 5 mm. O., I., and C. dark grayish brown (10YR 4/2). Rim from probably a type five globular bowl with outward turned lip. The clay has fine sand and mica inclusions. 155. Base. Inv. No. 67-3-3. Ft. G1. Fig. 9.11:155. B.D. 18 cm. P.H. 10 cm. Th. 8 mm. W. 135 g. O. gray (10YR 6/1), I. very dark gray (10YR 3/1), C. very pale brown (10YR 7/3). Base of a probable Type 1 cooking pot with combed lines for decoration vertically placed. There are abundant quantities of sand, mica, and graphite, creating a sparkle to the surface and ranging in size up to 2 mm. On the exterior is a circular depression and on the interior is a square depression. The locations of these two depressions seem odd and are probably an example of repair work to the vessel. As said in Chapter 3.1, the graphite decorated vessels were highly prized and kept in circulation even after repairs had been done. 9.5. Conclusion Roman pottery was a widespread and popular import among the Germanic Quadi, being the most commonly found Roman item on German sites. The variety of Roman pottery used by the Quadi shows that Germans were knowledgeable in the usage of the various vessels and applied them to their own daily life (such as pitchers/jugs), or they took Roman vessels and applied new use-patterns to them (such as mortaria). The types of pottery seen were mostly direct imports from Pannonia and were actually produced in the province as well, such as Brick-Colored Ware, Fine-Gray Ware, Coarse-Grey Ware, Pannonian-Striped and Stamped-Wares, and others. Those pieces from further afield within the empire, such as South-Gaulish terra sigillata, were probably the same pieces available in Pannonian markets. Even though Roman pottery is a common find on German sites and the variety of types found is great, the overall number of pieces found within specific contexts, such as individual huts or burials, is quite small in comparison to the total amount of artifacts recovered, especially in comparison to German pottery. In the Early Roman Period, especially in the early part, Celtic influence was seen on German hut forms as well as pottery types. This influence is comparable to the La Tène forms and decorations that survive on the Roman provincial pottery as detailed in section 8.1. Overtime, while Roman pottery forms and decoration continued to grow in influence, the Celtic influence diminishes. This dichotomy can be seen as a reflection of the reality on the ground, where the Germans became more and more entangled with various aspects of the Roman Empire nearby, while Celtic culture is absorbed either by the Romans or eastern Dacians.

9.4.2. La Tène Pottery at Urbárske Sedliská At Urbárske Sedliská, five La Tène-period sherds were recovered, all coming from Unit 3: two pieces from the fill of Hut 1 and the other three from stratum 1. Four pieces are diagnostic: three rims and a base; only two are presented here.

277

Endnotes 1. In his 1987 article dealing with a general discussion of Roman imports in Slovakia, Krekovič does provide an update to the types and amounts of Roman pottery found, including terra sigillata (1987, 247–254). 2. There are too many publications to list here in the chapter. The ones mentioned in the text are just a sampling of the extensive publications. 3. Gabler also published the ceramics for Vindobona (Gabler 1978). 4. There have been many publications in Slovakia dealing with the pottery from Gerulata within works cataloging the excavations of cemeteries, the fort itself, and various find spots in the town of Rusovce (see Kraskovská 1974a, 1974b; Pichlerová 1981; Kuzmová et al. 1996; Varsik 1996). 5. For a detailed description of all the various pitcher forms see Krekovič 1998, 19–22; Gassner 1989. 6. The use of sand to decorate the surface of thin-walled wares in general goes back to the third quarter of the first century BC and is first seen in central Italy in the area of Cosa; it is thought the use of sand originates from La Tène traditions (Anderson-Stojanović 1992, 39).

278

Table Table 9.1. Distribution of Roman Pottery at Urbárske Sedliská G1 Hut 1

G2

G3

G4 Hut 2

Unit 2

Unit 3 Sts. 1 & 2

Unit 4 Sts. 1 & 2

Unit 8

STPS

13

1

3

14

8

5

4

3

3

4

-

1

4

-

1

3

1

1

Mortaria

-

-

1

-

-

1

-

-

-

Fine Gray

-

-

-

1

-

-

1

-

-

Coarse Gray Beaker/Thin Walled

-

-

-

8

-

-

1

-

-

1

-

-

1

-

-

-

-

-

18

1

5

28

8

7

9

4

4

Pottery Type Brick Colored Pannonian Striped

Totals: Charts

Chart 9.1. Roman Pottery Found at Urbarske sedliska by Percentage Beaker/Thin-Walled 2% Coarse Gray 11% Smooth Gray 2% Mortaria 2%

Pannonian Striped 18% Brick-Colored 65%

279

Chart 9.2. Distribution by Percentage of the Roman Pottery at Urbárske Sedliská

Unit 8 5%

STPs 5%

G1 - Hut 1 21%

Unit 4 Sts. 1&2 11%

G2 1%

Unit 3 Sts. 1 & 2 8%

G3 6%

Unit 2 10%

G4-Hut 2 33%

Chart 9.3. Distribution by Percentage of Roman Pottery Vessel Forms Found at Urbárske Sedliská Pitchers/Jugs 7%

Pots 4% Beakers/Flagons 2% Mortaria 2%

Unknown 85%

280

Figures

Figure 9.1. Map showing the location of sites mentioned in Chapters 8–10 (both in text and figure-captions). 1. Vindobona 2. Carnuntum 3. Gerulata 4. Vaspuszta (Ad Mures) 5. Ács (Ad Statuas) 6. Brigetio 7. Celemantia A. Devín

B. Stupava

281

8. Aquincum 9. Savaria 10. Gorsium 11. Intercisa 12. Poetovio 13. Emona 14. Sirmium C. Mušov

Figure 9.2. Examples of Brick-Colored Wares. Various scales. 1–3:Gerulata; 4, 5, 7 and 8: Bíňa; 6: Cífer (1–3 are modified images from Krekovič 1998, Taf. 25:1–3; 4–8 are modified images from Krekovič 1981, Obr. 2:1, 6:8, 7:1, 11:1, 12:5).

282

Figure 9.3. Examples of various Roman wares. Not to scale. 1–5: Gerulata; 6: Carnuntum; 7: Sirmium; 8 and 9:Moravia (1–4 are modified images from Krekovič 1998, Taf. 7:1–4; 5–7 are modified images from Krekovič 1997b, Figure 4, 9, and 13; 8 and 9 are modified images from Droberjar 2002, 187).

283

Figure 9.4. Examples of various Roman wares. Not to scale. 1 and 3: Pobedim; 2: Krížovany nad Dudváhom; 4: Trnava-Hrnčiarovce: 5: Bíňa; 6: Šarovce; 7: Horná Streda; 8: Stupava (1–7 are modified images from Krekovič 1981, Obr. 10:3–6, Obr. 12:8, Obr. 15:2 and 5; 8 is a modified image from Turčan 2005, Obr. 3:1).

284

Figure 9.5. Examples of various Roman wares. Not to scale. 1–9: Gerulata (Modified images from Krekovič, 1998, Taf. 27:7–9, Taf. 28: 5 and 6, Taf. 30:1–4).

285

Figure 9.6. Examples of various Roman wares. Not to scale. 1 and 4: Bíňa; 2: Trnava-Hrnčiarovce; 3: Janova Ves; 5– 11: Gerulata; 12: Kostolná pri Dunaji; and 13: Sládkovičovo (Images 1–4 modified from Krekovič 1981, Obr. 4:1–4; images 5–11 modified from Krekovič 1998, Taf. 1:1–5 and 8, Taf. 13:4; 12 and 13 from Kolník 1979, Image 27).

286

Figure 9.7. Brick-Colored Wares from Urbárske Sedliská.

287

Figure 9.8. Pannonian-Striped Ware from Urbárske Sedliská. 153A reconstructs the vessel width at the body, while 153B provides a more accurate view of the surface. (153B drawn by Peter Šimčík)

288

Figure 9.9. La Tène pottery: Types 1–5. Not to scale. All vessels are from the oppidum at Bratislava (Modified images from Pieta and Zachar 1993, Obr. 86:1 and 8, Obr. 87:1–3 and 9, Obr. 88:1 and 3, Obr. 89:3, Obr. 90:1–3).

289

Figure 9.10. La Tène pottery: Types 6–10. Not to scale. All vessels are from the oppidum at Bratislava (Modified images from Pieta and Zachar 1993, Obr. 90:4 and 6, Obr. 91:1–3, Obr. 92:2, Obr. 93:1 and 2).

290

Figure 9.11. La Tène base from Urbárske Sedliská.

291

10.1.1. Metal Imports

Chapter 10

After pottery, but in much smaller numbers, metal import items are the most commonly found at German sites, especially cemeteries (Krekovič 1997a, 233, 234; Tab. 4.2). Bronze objects, such as fibulae, buckets, bowls, and various types of jewelry, are the most frequent among metal imports (Kolník 1998b, 77).

Miscellaneous Cultural Material and Faunal Analysis Chapter 10 details the remaining cultural items found in the features at Urbárske Sedliská. There are six sections to the chapter: 1) a summary of other items beside pottery found in German Huts for southwest Slovakia; 2) description of items recovered from the huts and features in the following order: Bronze Objects, Iron Objects, Bone Hairpin, Bead, Stone, and Slag; 3) spindle whorls found at the site; 4) a discussion of animal remains found at German settlements; 5) an analysis of the bones found at Urbárske Sedliská; 6) a summary of the findings.

1) Bronze vessels (which include buckets, ladles and strainers, frying-pans, kettles, and bowls). The identification and dating of the various vessel types were done at the centers of production in Italy and other provinces by A. Radnóti (1938), and H. Eggers (1951) catalogued a large number of bronze vessels found beyond the frontiers of the empire (Kraskovská 1978, 2; Droberjar 2002, 31). The typologies and classification created by Radnóti and Eggers form the basis of identification of bronze vessels found in Slovakia. Two groups have been identified. The first group of bronze vessels date to the AD 1–79 and were produced in Campania, primarily from Capua (Kolník 1998b, 77; Droberjar 2002, 31). By the late first century, bronze vessels from southern Gaul start to appear; in the second century bronze items arrive from the Rhine region, and by the third century the majority came from Pannonia (Kolník 1998b, 77). Most bronze vessels have been found in graves (Fig. 10.1:1–3, 9 and 10), with only a few examples from settlements (Krekovič 1997a, 233; Droberjar 2002, 31).

10.1. Roman Imports Other Than Pottery Recovered from German Huts Pottery is the most common cultural item found in German huts as seen in Chapters 8 and 9. Other items, such as jewelry, metal vessels, or glassware are found in huts, but in much smaller numbers. There are whole categories of items that were probably present but rarely preserved such as various household goods manufactured from wood, textile, or leather. As discussed in Chapter 7, we know that wooden furniture was present from the black stains left in hut floors. The preservation of textiles is a rare occurrence, but some have survived in enough quantity to reconstruct sewing techniques and the appearance of various clothing (see Furmánek and Pieta 1985, 79–81); evidence also comes from pottery fragments that have cloth impressions on them, and we know spinning took place in huts from the discovery of spindle whorls. Some bronze and iron fittings, especially belt parts, indicate the presence of leather goods.

2) Silver vessels. These are a rare find in Slovakia and the majority date to the Later Roman Period (Krekovič 1987, 245). They have all come from elite burials (Fig. 10.1:5), with the most well known at Krakovany-Stráže (Fig. 10.2), where one burial yielded an elaborate silver lanx, along with various bowls and cups (Holčík and Turčan 1998, 76; Kolník 1998b, 77; Turčan 2000, 52). A total of 13–14 silver vessels have been found in Slovakia (Krekovič 1987, 245). These were highly prized items and were probably given as gifts by Romans to German elites as diplomatic gifts (Droberjar 2002, 316), but see chapters 2 and 4 for a more detailed discussion.

The items that do survive in the archaeological record help to show a varied material culture, especially with the inclusion of imports. In the barbaricum, many items were produced locally by Germans, but there are various Roman imports present as well, along with items from further abroad, such as pottery from the Przeworsk culture in Poland and the Celto-Dacian groups in northern and eastern Slovakia.

3) Fibulae. Roman-produced fibulae have been found at over 62 sites in Slovakia, both in settlements and cemeteries, with the majority from graves (Krekovič 2001a, 95, 99). The majority of fibulae were made from bronze and iron, with some brass examples, but there are also isolated examples of gold, silver (Fig. 10.1:7), and enameled fibulae (Droberjar 2002, 303). Around 206 pieces had been found as of 2001, of which the majority (148 pieces) date to the first century AD and are the strongly profiled type A1 67–68 (Figs. 10.3:1 and 2) (Krekovič 2001a, 95). Other types that have been found include the Noric-Pannonian fibulae A 236–237 (Fig. 10.3:3), A 238 (Fig. 10.3:4) and Okorag type (Fig. 10.3:5), along with various types of fibulae from other parts of the empire, such as Aucissa types (Fig. 10.3:6), A

The following discussion of Roman import items found on German sites is not meant to be exhaustive; I do not list every type ever discovered in southwest Slovakia. The goal is to provide a general introduction to the more common types of artifacts discovered both in graves and settlements. It should be noted that some of the best preserved examples, and more rare artifacts, have been discovered in cemeteries; especially at Abrahám, Sládkovičovo, and Kostolná pri Dunaji (Fig. 8.1) (see Kolník 1980). 292

19 (Fig. 10.3:7) and Langton-Down types (Fig. 10.3:8) (Krekovič 2001a, 95). There is a decline in fibulae numbers starting in the second century, with only 48 pieces having been identified (Krekovič 2001a, 95). The most common forms are from Pannonia, including the knee fibula (Figs. 10.3:9 and 10) and the late strongprofiled fibula (Fig. 10.3:11) (Kolník 1998b, 81; Krekovič 2001a, 95). The decline in numbers continues into the third and fourth centuries, with only 3 known from the latter and 14 to the former (Krekovič 2001a, 98). Among these later fibulae, the most common form is the crossbow fibulae (Fig. 10.3:12) (Krekovič 2001a, 98).

as a type of adornment (Fig. 10.5). These strips could be strengthened and decorated with rivets, while their ends usually had a bronze end-piece (Figs. 10.5 and 10.7:f1 and f2), which could also be of iron or, more rarely, silver or gold (Droberjar 2002, 206). There are also some examples that were decorated with precious stones (Droberjar 2002, 206). These end pieces could have various styles and were either made by Germans or imported from the Roman Empire as part of finished belts. The most common form in the Roman Period had a vase-like body and a knob shaped end (Figs. 10.7:f1 and f2), but there were also some with spoon-like ends (Fig. 10.8) (Droberjar 2002, 206). The end-pieces had the portion that attached to the leather of the belt split to fit over the leather and there was a rivet that attached the split end together.

Fibulae were also produced by locally Germans, sometimes copying Roman types, but additional forms were imported from other parts of the barbaricum as well (Kolník 1998b, 80).

6) Day-to-day items such as small knives, scissors or needles. Small knives were an important part of life, having various uses much as they do today. Knives were usually of iron and are common finds on archaeological sites (Figs. 10.1:6 and 10.9) (Kolník 1998b, 82). The most frequent form is a straight blade with both sides sharpened, but in the Early Roman period there are also scalpels and curved blades that are thought to have functioned as razors (Kolník 1998b, 82). Scissors have been found quite often in graves, both male and female (Droberjar 2002, 219). Iron scissors are more commonly found than bronze versions (Droberjar 2002, 219). The most frequent form has the scissors as one piece of metal that has the blades at either end, which is then bent in the middle, creating a loop, so that the blades come together to form the scissors (Fig. 10.10). One held the scissors near the blades, and squeezed the blades together to make them cut. These scissors are similar in design to those used today for the sheering of sheep and are in fact referred to as “Shepherd’s Scissors” (Kolník 1998b, 82). Needles for sewing have been found in women’s graves, but they also turn up in settlements and as a common find (Kolník 1998b, 81; Droberjar 2002, 109). They are usually made of bronze, but iron examples are found as well. There are various sizes to the needles, from ones that are similar to modern examples, to ones that were very long, up to 11 cm, which were used probably for the sewing of specific types of items, whether leather or thick textiles.

4) Jewelry (which includes hairpins, rings, earrings, and bracelets/armbands). A less commonly found item are hairpins, which in the Early Roman Period were usually of bronze and by the Late Roman Period were more often made from bone (Kolník 1998b, 81). Most examples of hairpins have come from graves (Fig. 10.7:c), but there are examples from settlement contexts. Finger rings are almost unknown from the Early Roman Period in Slovakia, but for the Late Roman Period various examples have been found, some of gold with gem insets (Kolník 1998b, 82). Earrings are another rare find in German context and do not become a more common find until the Great Migration Period, starting in the fourth century AD (Kolník 1998b, 82). Bracelets and armbands were worn both by men and women and could be made of bronze (Fig. 10.4), iron, gold (Fig. 10.1:8), or silver (Droberjar 2002, 207). Most often these were made of wire with spiral twined ends having simple hook clasps. 5) Belt decorations. The various metal parts of belts are common finds in German graves and they also turn up in settlements, but in much smaller quantities (Kolník 1998b, 82). Most of the metal belt parts were either of iron or bronze (Droberjar 2002, 224). There are various parts to a belt, with the buckle being the most identifiable, which had a variety of forms with varying degrees of decoration. In the older Roman Period, some of the female graves have Noric-Pannonian-style belts, with bronze ornamentation in the form of fasteners with crimped plates, hinged parts with zoomorphic ends, and rivets and crescents decorating the belt itself (Figs. 10.5 and 10.7:d–j) (Kolník 1998b, 82). Most often, the buckles were of universal designs, made of bronze or iron, and various types have been identified: from the simplest circular or oval fasteners to square and more elaborately shaped buckles with reinforced fasteners, reminiscent of modern-day belt-buckles (Fig. 10.6) (Kolník 1998b, 82). Another part of the belt often found in archaeological contexts is the metal end-piece. Certain belts were not fastened by buckles but instead by straps that were twisted through a fastener, with the excess strip hanging

10.1.2. Lamps Roman ceramic lamps were common throughout the empire, but are a rare find in German context, where torches and lanterns were more common (Kolník 1998b, 76). There are some examples with relief decoration, but more often they were undecorated and the most common were Loeschke types IV and V (Krekovič 1983, 515). Some had maker’s stamps on the bottom with names such as FORTIS, LITOGENES, CRESCES, or VIBIANI (Krekovič 1983, 511, 512; Kolník 1998b, 76; Droberjar 293

150) and the amounts and sites continue to climb into the Great Migration Period where they peak (Krekovič 1987, 261 and 263). A majority of the beads discovered date to the Late Roman Period, are of glass and come from graves (Krekovič 1987, 263).

2002, 114). There are isolated examples from the fourth century with Christian markings (Kolník 1998b, 76). According to 1987 figures, only 11 ceramic lamps have been found in Slovakia on German sites, while a great many have been found at the Roman sites of Gerulata and Kelemantia (Krekovič 1987, 263; Kolník 1998b, 76). In Moravia, most of the lamps date to the second century AD and come from northern Italy (Droberjar 2002, 114).

10.1.5. Items from bone As previously mentioned, some items could be made from bone, such as hairpins and beads, while antlers could be used for knife-handles. Other items made from bone include needles, awls, combs, game-pieces, and other specialized items (Kolník 1998b, 83; Droberjar 2002, 134). The working of bone and antler was a widespread industry in the Roman and Great Migration Periods (Droberjar 2002, 134).

10.1.3. Glass Products The use of glass for day-to-day items and their production in mass-quantities started in the late first century BC and early first century AD within the Roman Empire, quickly making glassware a common item in Roman households (Harden 1987, 1). During the same period, glassware also spread beyond the frontier into German contexts usually as luxury items (Kolník 1998b, 78; Droberjar 2002, 295). There are finely made glass bowls found in elite burials in southwest Slovakia at Zohor (Fig. 10.1:4) and Vysoká pri Morave dating to the first century AD (Kolník 1998b, 78). A unique example is a glass alabaster vase used as a burial urn for an affluent buyer (Kolník 1998b, 78). Many of the best examples with large portions surviving come from graves, while examples from settlements are usually isolated shards (Droberjar 2002, 295). Various types of glassware have been found, including bowls, cups, platters, and, vases as well as small vessels for holding oils, perfumes, and cosmetics (Kolník 1998b, 78).

10.2. Artifacts Recovered from Huts 1–2 and Features G2–G3 10.2.1. Bronze Artifacts 1. Fibula. Inv. No. 945-3-3. Ft. G1. Fig. 10.11:1. The fibula is not entirely preserved, missing the lower part of the bow with the catch-plate. The break occurs at the nodule which divides the bow. The upper portion of the bow, spring, and pin are well preserved. The fibula is made of bronze, with no evidence of etched designs or other decoration. The pin is 3.2 cm in length; the spring is 1.4 cm wide, with the coil 6 mm in width. The upper bow is 1.2 cm at its widest portion and 4 mm at its narrowest, while the nodule separating the upper and lower portions of the bow is 8 mm wide and 3 mm thick.

10.1.4. Beads The most common type of jewelry found, especially in the Late Roman Period, are necklaces made with beads, usually of glass, but there are examples from elite graves of gold beads decorated with granulation (Kolník 1998b, 82). Other material could be used to make the beads including amber, clay, bronze, silver, iron, stone, and even bone (Droberjar 2002, 134). The beads can have various shapes and forms, but in Moravia for example, the most commonly found are the rib beads of Type XVIII (Droberjar 1997, 125).

The length of the pin (which is fully preserved) and the shape of the upper part of the bow, nodule and spring indicate that the fibula is most likely an A 68.3 The A 67, which has a similar design in the upper bow section (but with a more pronounced s-shape), has an elongated lower bow with catch-plate, thereby lengthening the pin as well. Since the pin is not overly long on the fibula found in Hut 1, it would seem unlikely that it would be an A 67, while the A 68 is a better fit.

Before the research done by M. Maczynska,2 it was thought that beads were not usable for dating, but her work showed otherwise (Krekovič 1987, 261). She studied 50,000 beads and was able to create 29 categories and 387 types by color, how manufactured and their shapes (Krekovič 1987, 261). According to the study done by Maczynska, few beads were found at sites that were very close to the frontier in comparison to the amounts found as you went further north, and the author concludes that the wearing of beads was not as popular near the frontier (Krekovič 1987, 261 and 263). In Slovakia, there is a jump in the amount of beads found and the number of sites starting in the B2 period (AD 90–

With the catch-plate missing, it is difficult to date this fibula. When looking at the portion of the bow that survives, it has a much wider s-shape to it, lengthening this portion of the bow. The early A 68 shares similar design to the A67 and the transitional A67/68, which all have the upper portion of the bow shorter and the curve more pronounced (Kolník 1971, 513 and Obr. 12:28 and 31). The A 68 with an elongated s-shape upper bow section has been identified as first coming in to use during the Claudian period (Kolník 1971, 518 and Obr. 16:21, 22 and 31). The elongated s-shape continues into the first half of the second century (Kolník 1971, 522). 294

The A 68 fibula is among the most commonly found fibulae in southwest Slovakia, starting in the mid-first century AD and lasting into the second century (Krekovič 1987, 256 and 261; Krekovič 2001a, 99). It is believed that the A 68 may have been made by the Germans as well, but a way of determining differences between the two has yet to be found (Krekovič 1987, 261).

distinguishing marks survive to identify its original function. 6. End-piece to belt strap. Inv. No. 938-4-3. Ft. G4. Fig. 10.11:6. Undecorated bronze end-piece to a belt strap. There are two bronze strips making the end-piece. Both halves are well preserved, along with a portion of the rivet that had held the end-piece to the strap. The end-piece is 4.4 cm long and 6.5 mm wide. Each half is 1 mm thick. The top portion, where the rivet is located, is slightly wider at 8 mm than the rest of the strip, and the end of the top portion has a rounded end. Where the rivet is located the two halves are bent out to make room for the leather of the belt strap. The backside of the rivet that would have held the backside of the bronze end-piece to the leather is broken off.

2. Hairpin. Inv. No. 943-4-4. Ft. G4. Fig. 10.11:2. Bronze one-piece hairpin with the shank bent at the last third of its length.4 The total length of the pin is around 14 cm, with a diameter of 2 mm mid-way down the shank, while the head itself is 2 cm long. The head of the pin is ornamented with a series of discs and globes. Above the shank are two wide discs (each 5 mm wide), followed by a large oblong sphere (5 mm wide and 7 mm long), above which are two more wide discs (each 5 mm wide), then one smaller disc (4 mm wide), then a small globe (3 mm in diameter), a small disc (3.5 mm wide) and finally a last half-globe (3 mm wide). No distinct designs could be seen in the oblong sphere and there is no indication that there was a hole through it either. All the discs have rounded edges.

7. End-piece to belt strap. Inv. No. 939-4-3. Ft. G4. Fig. 10.11:7. Half of an undecorated bronze end-piece to a belt strap. The piece is 6.3 cm long and has a tapered form. The top of the strip, where the hole for the rivet is located, is squared off and 7.5 cm wide, then tapers to a rounded end, where the width is 3.2 mm. The upper portion of the strip is 1 mm thick and the lower portion is 0.5 mm thick.

No known parallel with this hairpin has been found at other sites in southwest Slovakia or from sites further abroad in the barbaricum or from the provinces. There are similar hairpins, which provide a rough date for the Urbárske Sedliská hairpin. The Urbárske Sedliská hairpin fits into Group IVa of the groups created by B. Beckmann for hairpins of the Roman Period, which are the singlepiece pins with no hole and the head of which has a cone shape made up of various decorative elements (Beckmann 1966, 26, Tafel 2: Untergruppe IVa). This group dates to the Early Roman Period, mainly within Period B (Beckmann 1966, 26–30, 42–43). Group IV hairpins have a wide distribution in the barbaricum, with the earliest ones found in the Bohemian Basin (Beckmann 1966, 42).

8. Unidentified bronze. Inv. No. 940-4-4. Ft. G4. Fig. 10.11:8. A bronze piece that appears to have melted at some point. The original shape or function cannot be determined. 10.2.2. Iron Artifacts There were various iron fragments found in both huts and the two features. Their preservation was so poor, however, that their original shapes and functions could not be determined. The table below displays the numbers of iron artifacts found in each feature.

3. Needle. Inv. No. 944-3-3. Ft. G1. Fig. 10.11:3. Well preserved bronze needle with only a slight bend to the bottom portion. The length is 7.5 cm, 3 mm wide at the eye, and 1.2 mm diameter toward the bottom.

Table 10.1. Iron Artifacts Found by Feature Feature Number G1 (Hut 1) 1 G2 6 G3 4 G4 (Hut 2) 0

4. Needle. Inv. No. 942-4-3. Ft. G4. Fig. 10.11:4. Long needle bent toward the lower end and the eye is missing its upper portion. The needle is 11 cm long, 3 mm wide at the eye and 1.5 mm diameter toward the bottom. Beneath the eye of the needle for decoration are raised bands. 5. Portion of needle. Inv. No. IM239-3-3. Ft. G1. Badly corroded piece of a pin or needle. The portion that survives is 2 cm long and 4 mm wide, but no

295

10.2.3. Bone Hairpin

12. Gaming stone/sling-bullet. Inv. No. 927-4-3/4. Ft. G4. Fig. 10.13.

9. Bone Hairpin or awl. Inv. No. 941-4-4. Fig. 10.11:9. Well preserved reddish-brown stone spheroid. The long axis is 2.7 cm and the short axis is 2.2 cm. There are no imperfections or blemishes to indicate it had been mounted on or part of something else. The outer surface is perfectly smooth.

A partially preserved undecorated hairpin or awl. The bottom portion is broken off, but otherwise in good condition. The preserved length is 8.5 cm, with an originally length of probably 10–12 cm. The pin has a tapered design with the upper portion having a 6 mm diameter and a rounded end, while at the middle the pin has a 4 mm diameter and where it is broken off, 3 mm.

Most stone game pieces are more disc shaped and either light or dark colored (usually white and black) for use in games such as ludus latrunculorum or ludus duodecim scriptorum (Droberjar 2002, 84 and 92). Both stones (923 and 927) are more rounded and of colors that differ from the stones used in the mentioned games, but these differences do not preclude their being used for some type of game. On the other hand, their use as sling bullets is possible, but Germans were not known to use slings. There is also no other evidence for the presence of Roman soldiers to explain the stone’s presence.

10.2.4. Bead 10. Glass Bead. Inv. No. 937-4-3. Ft. G4. Fig. 10.11:10. An opaque glass bead decorated with ribbing on the sides. The diameter is 1.5 cm and the height is 1.3 cm, while the diameter of the hole is 4 mm. The glass has an olive green color and has milky discoloring with streaks of imperfection inside the glass, along with minute cracking. Based on the condition of the glass and being found in fill that appeared to have been burnt, the bead has been sintered. The green then may not be its original color and it may have been originally translucent.

13. Whetstone. Inv. No. 924-4-3/4. Ft. G4. Figs. 10.15:13 and 10.16:13. Rhomboidal shaped sandstone, probably used as a whetstone. The longer sides of the rhombus are 6.7 cm long and the shorter sides are 4 cm, while the stone is 1.5 cm thick. The identification of this stone as a whetstone is based on the discovery of a similar rectangular whetstone made from sandstone found in feature 4 at Kuchyňa (Turčan 1988, 117, 118, Obr. 8:1, 123).

The bead may fit into a typology developed by E. Riha for beads discovered at Augst/Kaiseraugst in Germany. Type 1 Melonenperlen, variant 3, seems to fit best with the glass bead from Urbárske Sedliská (Riha 1990, 82; Taf. 36:1160–1165). These beads are of a translucent glass, with air vesicles and milky-streaks (Riha 1990, 82). They were produced by the first century in Western Europe and in the Danube region, and were most popular between the Claudian and Antonine periods (Riha 1990, 82).

14. Pestle. Inv. No. 925-4-3/4. Ft. G4. Fig. 10.14. A rectangular ground stone, with one end broken off and the other rounded. The surface is very smooth except for the broken off end. There is percussion damage on the rounded end, indicating the stone’s use for pounding and grinding. The stone is 9.6 cm long and 4 cm wide.

10.2.5. Stone 15. Broken pestle. Inv. No. 926-4-3/4. Ft. G4. Figs. 10.15:15 and 10.16:15.

Various stones were found in the two huts. Hut 1 had three small rocks that bear no indication of anthropogenic modification. Hut 2, however, had rocks that appeared to have specific uses.

Stone fragment of what was most likely a pestle; note its similar shape to item 14. The stone is 11.3 cm long and 4.8 cm wide. The intact outer surface is smooth to the touch.

11. Gaming stone/sling-bullet. Inv. 923-4-3/4. Ft. G4. Fig. 10.12.

16. Mortar. Inv. No. 928-4-3/4. Ft. G4. Figs. 10.15:16 and 10.16:16.

Partially preserved spherical stone. The preserved diameter on the long axis is 2 cm, while the short axis is 1.4 cm at the wide end and 8 mm at the narrower end. This stone may have been either a gaming piece or sling bullet based on the similarity to the fully preserved stone, entry 12, below.

Smooth stone with concave surfaces on two sides. The stone is 10 cm long and 8.5 cm wide. The identification of the stone as a mortar is based on the concave sides and the presence of the two pestles.

296

17. Polishing stone. Inv. No. 930-4-3/4. Fig. 10.17.

common finds both in graves, usually of women, and in settlements (Droberjar 2002, 271).

Triangular shaped stone that appears to have a large quantity of graphite. The stone is 3.8 long, 3 cm wide and 1.6 cm thick. There are a series of incised lines on one side of the stone. The surface is very smooth. It is postulated that the stone was used to apply graphite to the surface of pottery vessels and to then polish their surface.

18. Spindle whorl. Inv. No. ?-3-1. Fig. 10.18:18. The spindle whorl is complete with a diameter of 3.3 cm, the hole having a diameter of 4 mm; the thickness of the whorl is 1.8 cm. The form is roughly biconical, having one side more cone-shaped than the other, showing its hand-made workmanship. The color is black and is decorated with vertical incised lines on the sides of the whorl, along the equator.

10.2.6. Slag Hut 1 had seven fragments of slag and Feature G2 had one piece of slag. As to whether or not they are true slag, i.e., being a once molten silicate or silicate mixture (Bachmann 1982, 1), remains to be determined by a specialist. A basic visual identification can be attempted, however, to try and determine if the pieces are from metallurgy or are they vitrified material created by other processes. Slag from metallurgy is usually black or dark grey on freshly exposed surfaces and show flow textures (Bachmann 1982, 3). Slag should have gas holes visible in a fresh surface, although some glassy slag may have no holes (Bachmann 1982, 4). Inclusions are common and if there is a high iron content, some slag may exhibit rust.

19. Spindle whorl. Inv. No. 443-3-2. Fig. 10.18:19. The spindle whorl is only 50% complete, weighing 2 grams. The diameter was probably 3.2 cm and the thickness was 8 mm. The diameter of the hole is ca. 2.5 mm. The whorl is disc shaped and undecorated, with a rough surface. 20. Spindle whorl. Inv. No. 935-4-2. Fig. 10.18:20. The spindle whorl is 75% complete, missing a portion of one side. The diameter is 4.2 cm, the thickness is 1.6 cm, with the diameter of the center hole at 6 mm. The weight is 20 grams. There are no decorations on the outside and the form is biconical, with one side being slightly larger. The center is depressed around the hole.

Of the eight pieces, seven appear to be slag of some type, having various gas holes and flow textures. The eighth piece, no. 448, is an oddly shaped lump of unknown material, which could be either stone or metal, with melted surfaces.

21. Spindle whorl. Inv. No. 936-4-2. Fig. 10.18:21. The spindle whorl is intact with a uniform shape, having a raised flange at the center around the central hole on both sides. The diameter of the whorl is 4.1 cm, while the thickness at the outer edge is 1 cm and from flange to flange it is 2 cm. The diameter of the central whole is 8 mm. The spindle whorl weighs 25 grams.

10.3. Spindle Whorls Possibly of the Roman Period Four spindle whorls were found during excavations, two of which (18 and 19) came from Unit 3 (strata 1 and 2) and the other two (20 and 21) from Unit 4 (stratum 2). They are included here because there is a possibility that they originated from either or both huts, especially the three whorls from stratum 2 of Units 3 and 4.

10.4. Animal Husbandry among the Germanic Quadi Animal husbandry was an important part of German agriculture, with much of a person’s wealth tied to the size of their flocks and herds (Todd 2004, 76–77). It is not surprising, therefore, to find great quantities of animal bones at settlement sites (Droberjar 1997, 149; 2002, 228). Studies at various sites in the Czech Republic and Slovakia have shown that cattle were the most widespread and important, followed by pigs; sheep and horse were a distant third, while goats and chickens were rarer (Droberjar 1997, 149; Droberjar 2002, 228; Todd 2004, 76, 77). Other animals found include the domesticated goose, dog, and even an isolated find of cat remains (Droberjar 2002, 228). Evidence from the bones has shown that while Germans did use some of these animals (excluding dogs and cats) as draught animals and for their secondary products such as wool and milk, meat was an important part of the diet. Studies have shown

Two of the spindle whorls (18 and 21) are complete, while the third is 75% (20) complete and the fourth (19) only 50% complete (Fig. 10.18). All four are made of fired clay and do not appear to be reused, thick-walled pottery fragments, which is not an uncommon occurrence (Furmanek et al 1991, 246). The whorls are medium sized (compared to the large variety of shapes) and lightweight, signifying that they were probably used in the creation of wool (larger, heavier whorls were used in the spinning of flax or yarn) (Furmanek et al 1991, 246). Spindle whorls were usually made from clay, but there are stone examples as well (Droberjar 2002, 270). Spindle whorls were made in a variety of shapes, following a basic discoidal or biconical shape. They may be undecorated (more common) or have appliqué, incised lines, or dimples (Droberjar 2002, 270–271). They are

297

large numbers of animals such as sheep and pigs slaughtered at a young age, but smaller numbers of cattle, which appear to have been more important for their milk and use as draught animals (Todd 2004, 77). Over time, because of non-selective breeding, German cattle and horses became smaller in comparison to those found in the Roman Empire and elsewhere, with the cattle being in height only 1.1 m at the shoulder, slender and with short horns, while horses were 1.4 m at the shoulder and solidly built (Droberjar 2002, 228; Todd 2004, 77). Wild animals appear to have had a minor role as a food supply, even though there was an abundance of wild boar, aurochs, roe, and red deer (Todd 2004, 77). There is evidence, however, to show that these animals were hunted, along with pheasant, bear and even elk, as discussed below (Droberjar 2002, 230). Fish and shellfish were also collected by ancient Gemrans, but their remains do not survive well (Droberjar 2002, 230; Todd 2004, 78).

Chapter 6 as being part of Hut 1. The amount of bone from sectors D1 and E1 is small, only 2 pieces. The bones of Units 3 and 4, stratum 1, and all the bones from Unit 8 were not given to the physical anthropologist for study. There were two reasons: 1) the bones were not from an intact cultural layer or could not be judged with any assurance to have come from one, and 2) the amount of bones from the huts that needed to be studied was already very great and it was thought best to focus on the groups that were deemed most important. Table 10.2. Bone/Teeth Totals from Units 2 – 4 and 8 Unit Stratum 2 Stratum 3 Hut 1 Hut 2 28 2 293 See Hut 1 830 3 143 See Hut 2 781 4 5 2 8

A detailed study was done of the animal remains from a German settlement excavated at Mlékojedy in the Czech Republic (Fig. 8.1) (Droberjar 2002, 184–185). Of the thousands of bone remains, it was found that 97% was made up of domestic animals, with cattle the most represented, followed by pig, sheep and goat, and only a few remains of dog and chicken (Droberjar 2002, 228). The cows weighed on average 236 kg and the bulls 278 kg (Droberjar 2002, 228). In regards to wild animals, deer and rabbit were the most common, with singular remains of wild boar, wild ox, badger, fox, and bear (Droberjar 2002, 228). Bird remains were also found such as crow, starling, and, unique to Mlékojedy, the remains of a crane (Droberjar 2002, 228).

Table 10.3. Animals Present and Their MNI and NISP from Units 2–4, Stratum 2 Unit and Animal Type MNI NISP Unit 2 Cattle (Bos p.f.t.) 1 2 Pig (Sus s.f.d.) 1 2 Sheep/Goat (Ovis/Capra) 1 1 Unidentified 20

10.5. Analysis of the Faunal Assemblage from Urbárske Sedliská5 The bones at Urbárske Sedliská were concentrated around and within Huts 1 and 2, thereby involving Units 2, 3, 4, and 8 (see Figure 6.2). Of the other units, only Hut 5 provided bone and that was only one fragment. Table 10.2 provides the amounts recovered from Units 2, 3, 4 and 8, with the bones from Huts 1 and 2 listed separately. The bones from stratum 2 of Units 3 and 4, even though they come from plough zone/disturbed strata, most likely originate from the huts (see Chapter 6 for more information). These bones, however, were not added to the totals of the huts and will be discussed separately in conjunction with the bones found in stratum 2 of Unit 2. As already discussed in Chapter 6, there is a strong probability that Roman-Period artifacts found in stratum 2 of Unit 2 originated in Hut 1 but because of plough action have been moved. The high number of bones found in Unit 2 is also a probable outcome of having been removed from Hut 1 along with the pottery. The only bones that could be added to the hut totals from Unit 2 are from sectors D1 and E1, which were described in

Unit 3 Cattle (Bos p.f.t.) Pig (Sus s.f.d.) Sheep/Goat (Ovis/Capra) Horse (Equus p.f.c.) Deer (Cervus e.) Unidentified

4 4 4 1 1 -

24 27 7 1 2 92

Unit 4 Cattle (Bos p.f.t.) Pig (Sus s.f.d.) Sheep/Goat (Ovis/Capra) Horse (Equus p.f.c.) Deer (Cervus e.) Unidentified

2 1 1 1 1 -

16 4 3 6 1 53

10.5.1. Analysis of Bones from Stratum 2 of Units 2–4 Table 10.3 shows the variety of animals found and their MNI (Minimum Number of Individuals) and NISP (Number of Identified Specimens) for Units 2–4, stratum 2, based on the number of teeth and bone that could be studied. The bones were in a varied state of preservation, some showing burning (14 bones), one having been heated by cooking (having a shiny sheen to the surface), while a large majority were broken or incomplete (Table 10.4 and 10.7). One bone had bite marks, while 11 had 298

evidence of gnawing (Table 10.5). Nine bones had cut marks on them (of which one also had evidence of gnawing) and one was chopped (Table 10.6). From Unit 3, the one horse bone (unidentified, 72 mm long) had evidence of having been worked, while a horn fragment appears to have been worked into an awl. Three items of pig from Unit 3 could be given a gender: one female mandible and one male maxilla and tooth. From Unit 2, a pig mandible showed the animal to have been around a year old. In Unit 3, a cow maxilla with an unerupted tooth (M3) shows it was under two years old, while an unidentified mandible belonged to an individual over two years old. Also from Unit 3, a pig maxilla with un-erupted M3 was probably under 1.5 years old.

bone, one is an unidentifiable fragment of bone from Hut 1 while the other is a section of deer antler from Hut 2, which also had a deer phalanx that may have been worked at the proximal end. Table 10.11 indicates male and female numbers, which all came from pigs (the determination of sex was based on the teeth). The general age could be determined for a number of animals. One cow could be aged at less than 1.5 years, one at around 1.5 years old, while a third was under two years old, and a fourth was around 3 years old (these ages were determined by epiphysial fusion on various bones: humerus, phalanx, metapodium, and tibia,). One sheep/goat could be aged at five months (based on epiphysial fusion in the phalanx 1). Most of the ages were obtained from pigs, of which one was under one year old, one greater than a year old, five at around 1.5 years old, and one was under two years old (these ages were based on epiphysial fusion seen on the humerus and tibia, and the dental eruption of the I1 and M3 teeth). Table 10.15 displays the ages of the various animals and shows that most animals were around 1.5 years old at time of death. In general, pigs were primarily kept for their meat and fat and were killed before the age of two, when the meat was still tender (Green 1992, 19), which could explain why all the aged pig bones were under 2 years old. Usually sheep are kept for their secondary products wool and milk, so that when killed for meat it was usually at an older age (Green 1992, 10 and 15). If sheep were killed for meat, it would be at a young age, 2 and under, when the meat would be most tender. The one sheep/goat bone aged at 5 months would appear then to have been killed for its meat. If cattle were killed specifically for their meat, it would be around the time they reached maturity, three to four years (Green 1992, 36). The four aged bones from the site, three under two years and the fourth at three years, would indicate that these cows were killed for their meat. The more usual practice was to keep cattle for their secondary uses until too old, at which point they were killed for the meat (Green 1992, 10 and 15).

10.5.2. Analysis of Bones from Huts 1 and 2 A total of 1,492 bones was collected along with 119 teeth from the two huts. Of these two totals, 556 bones and 107 teeth were well enough preserved to be commented on by the specialist (Table 10.8 provides the totals from the huts). The bones from the huts were in various states of preservation and as a general assemblage did not differ greatly from the stratum 2 bones. Table 10.9 shows the variety of animals found and their MNI and NISP from the two huts. Table 10.10 displays the total number of bones and teeth per animal found in Huts 1 and 2, showing that cattle and pig were the most important animals at Urbárske Sedliská. The presence of pig in almost equal number to cattle is surprising in view of the evidence from other sites as discussed above. When the unidentified category is included, the domestic animals only comprise 44% of the total and the wild only 2.8%. If the unidentified category is removed, then the domestic jumps to 94% and the wild is only 6% of the total, which corresponds well with what has been found at sites such as Mlékojedy discussed above. These totals and percentages can be seen in Chart 10.1 at the end of the chapter. Cattle and pigs make up just over 40% each of the total, with a combined percentage of 80.9%, clearly dominating the assemblage. Table 10.8. Bone and Teeth Totals from Huts 1 and 2 Hut Bone Teeth 1 282 69 2 274 38

Table 10.15. Ages of Animals Found at Urbárske Sedliská Age Cattle Sheep/Goat Pig Total 5 months 1 1 < 1 yr. 1 1 > 1 yr. 1 1 < 1.5 yrs. 1 1 c. 1.5 yrs. 1 5 6 < 2 yrs. 1 1 2 3 yrs. 1 1

Similar types of information could be determined from the Hut bones as already seen in the stratum 2 bones. Table 10.11 provides an overview of the findings, while Tables 10.12–14 provide more information on specific findings. A great many bones showed evidence of having been burned (93 total), most likely when the animal parts were cooked. Fourteen bones showed evidence of cut or chop marks from butchering, and the distribution is spread evenly among animal and bone types. Twelve bones have evidence of gnawing and these too are spread evenly among animal and bone types. Of the worked 299

coming from the frontier or with Romans living at Stupava.

10.5.3. Comparison of Faunal Findings from Urbárske Sedliská with Huts from That of Other Sites As already mentioned, the prevalence of domestic animals at Urbárske Sedliská compares with what was found at Mlékojedy, but it was surprising to find pig bones equal to the number of cattle bones. When we look at the faunal evidence from huts excavated in Moravia (Droberjar 1997, 150) dating to the B2 period, we see a difference in the number of different animals represented in comparison to the Urbárske Sedliská huts, but the general pattern is clear: cattle dominate the assemblage, followed by pig and sheep/goat. Even though we see at Urbárske Sedliská pigs equaling if not surpassing cattle, such equality is not the case when looking at German huts over a wide area.

10.6. Summary of Findings German huts have a wide assortment of material preserved within them. Not only did Germans enjoy locally produced goods but also imports from other cultures, especially from the Roman provinces. The most common import found in German huts was pottery of various types from the provinces, ranging from basic cooking ware to high-quality fine ware such as terra sigillata. The second most commonly found were metal items, ranging from bronze fibulae and pots to iron knives. Found less often were glassware and special items such as lamps.

While it is true that cattle still make up an important part of the faunal assemblage for Quadi German settlements, the amount is smaller than what is known at sites in Germany itself. At the site of Feddersen Wierde over half the faunal assemblage was made up of cattle, while at Wulfshof it was nearly 70 percent (Todd 2004, 78). Based on Table 10.16, cattle make up only 45% and pig 25% of the total for huts from Moravia and Slovakia as a whole. There are three explanations for the higher pig remains, and they need not have been independent of each other.

The huts from Urbárske Sedliská had a variety of items other than pottery. A bronze fibula from Hut 1 could be identified as an A 68, dating between the mid-first century and second century AD. Hut 2 yielded a bronze hairpin that could be placed within a general category of hairpins dating to the same period as the fibula. Other items included bronze needles, bronze belt fittings, a wooden hairpin, and stone gaming pieces. There were also stones from Hut 2 that could be identified as probable mortar and pestles used for day-to-day household activities.

The first explanation is the environment needed for pigs. Pigs are a forest animal and so where there is greater forest cover, there was usually a greater amount of pig raising (Green 1992, 18–19; Todd 2004, 77). Pigs were usually herded and not kept on site, being allowed to forage in the woodlands (Green 1992, 18). The region around Urbárske Sedliská, with the heavily forested Smaller Carpathian Mountains nearby, has populations of wild pig to this day, so that the area was ideal for the raising of pig.

The greatest amount of material found, even surpassing the pottery, was the faunal remains. The site showed that the local Germans had a meat diet dependent upon cattle and pig, with supplement from sheep/goats, bird, freshwater mussels, and wild animals such as deer. A surprising find was that pigs equaled the cattle remains at the site. Usually cattle remains dominate the faunal assemblage and the presence of pig in equal number at Urbárske Sedliská could be explained in three ways: 1) the natural environment was conducive toward the rearing of pig, 2) the previous Celtic population raised pigs and this practice may have been adopted by the Germans, and 3) the Roman army on the frontier and local Romans possibly at Stupava enjoyed pork, and so the local Germans may have raised more pigs to meet this demand.

The second explanation has to do with the Celtic population that had previously occupied the region, as discussed in Chapter 3.1. Pigs were closely associated with Celts and appear to have been as important if not more so than cattle (Green 1992, 5, 17–18; Jerem 1995, 594). It is possible then that as the Germans came into the region of southwest Slovakia, they picked up pig husbandry from the local Celts still living in the area, making the animal gain greater importance than it possessed in purely German territories.6

The finds from Urbárske Sedliská fit into the basic pattern of German material culture, as seen when compared to finds from other German huts.

The third explanation is the presence of Romans at the frontier and possibly at the Roman-style structure nearby at Stupava. The Roman army is known to have enjoyed pork and that local peoples sometimes geared their animal husbandry toward the presence of the Romans (Green 1992, 10). It is possible, then, that the Germans at Urbárske Sedliská may have been keeping more pigs than normal in order to do business with Roman traders 300

Endnotes 1. The “A” stands for O. Almgren, who in the late 19th century created a very useful typology for fibulae still in use today (see Almgren 1897). 2. See M. Maczyńksa 1977 and Tempelmann-Maczyńksa 1985. 3. Identification was done by Mgr. I. Bazovský, with the help of Dr. V. Turčan. 4. In describing the parts of the hairpin, I use the terms first put forward by B. Beckmann in his 1966 article (1966, 12–14). 5. I thank Dr. Alena Šefčáková of the Natural History Museum–SNM Bratislava for her analysis of the faunal remains from the site. 6. It is interesting to note, however, that in central Slovakia, where the Púchov culture (Celto-Dacian, lasting from ca. 200 BC to AD 175) was located, cattle dominated the faunal assemblage at Liptovská Mara followed by sheep/goats, with pigs the third most abundant (Pieta 1996, 43). This Celtic enclave may have had fewer pigs because of their location in higher terrain, where one could easily have cattle in the lowlands and sheep in the highlands, not competing with each other for pasture. By the Roman Period, however, the faunal assemblage showed a decline by almost half in the amount of sheep/goat present, but a steady increase in the pig bones (Pieta 1996, 43). K. Pieta believes the fall in sheep/goat occurred because of deteriorating environmental conditions (Pieta 1996, 43).

301

Tables Table 10.4. Information Gleaned from the Bones, Units 2–4, Stratum 2 Unit Bite Mark Gnawed Cut Worked Cooked Burned Female 1 1 2 1 5 8 1 1 1 1 3 6 1 14 4 11 10 1 1 16 1 Total: 1

Table 10.5. Bones Showing Teeth/Gnawing Marks, Units 2–4, Stratum 2 Mandible Metapodium Pelvis Ulna Humerus Tibia 1 1 1 2 1 1

Unident. 5

Table 10.6. Bones Showing Cut or Chopping Marks, Units 2–4, Stratum 2 Costa Humerus Mandible Scapula Phalanx 1 Antler 1 2 1 1 1 1

Unident. 3

Animal Sheep/Goat Unidentified

Animal Cattle Unident.

Male 2 2

Table 10.7. Bones Showing Evidence of Burning, Units 2–4, Stratum 2 Animal Os carpale Cranium Vertebra Phalanx 1 Tibia Unident. 1 Pig 1 Sheep/Goat 1 2 1 10 Unidentified -

Table 10.9. Animals Present and Their MNI and NISP from Huts 1 and 2 Hut and Animal Type MNI NISP Hut and Animal Type MNI Hut 1 Hut 1 Domestic Cattle (Bos p.f.t.) Pig (Sus s.f.d.) Sheep/Goat (Ovis/Capra) Horse (Equus p.f.c.)

7 10 3 1

75 95 17 3

Domestic bird (Gallus p.f.d.) Unidentified

1 -

1 152

Hut 2 Domestic Cattle (Bos p.f.t.) Pig (Sus s.f.d.)

1 2

50 31

Sheep/Goat (Ovis/Capra) Horse (Equus p.f.c.)

2 1

13 6

Unidentified

-

211

Wild Bird (Aves) Deer (Cervus e.) Wild Ox (Pratur) Fresh Water Mussel (Unio and Unio sp.)

Hut 2 Wild Bird (Aves) Deer (Cervus e. and Capreolus) Brown Bear (Ursus arctos) Snail (Cepaea vindobonensis)

302

NISP

1 1 1 2

1 2 1 4

1 2

1 8

1 1

1 1

Animal Type

Table 10.10. Total Number of Bones per Animal and Totals of Domestic and Wild Animals Found in the Two Huts Hut 1 Hut 2 Totals Total # # # # Bones* # Bones* # Teeth Amount Teeth Bones* Teeth

% Total Amount

Domestic Bird, Dom. Cattle Horse Pig Sheep/Goat

1 58 2 60 8

17 1 35 9

39 19 13

11 6 12 -

1 97 2 79 21

28 7 47 9

1 125 9 126 30 Total

.2% 18.9% 1.4% 19% 4.5% 44%

Wild Bear Bird, Wild Deer Mussel Wild Ox Snail

1 4 1 -

2 -

1 5 1

1 3 -

2 5 4 1 1

1 5 -

1 2 10 4 1 1 Total

.2% .3% 1.5% .6% .2% .2% 2.8%

147

5

10

353

53.2%

Unidentified

Hut 1 2 Total:

196 5 343 *Includes shell from mussels and snail

Table 10.11. Information Gleaned from the Bones, Huts 1 and 2 Gnawed Chopped Cut Marks Worked Burned Female 4 2 5 1 12 4 8 7 2 81 2 12 2 12 3 93 6

Male 6 1 7

Table 10.12. Bones Showing Teeth/Gnawing Marks, Huts 1 and 2 Animal Phalanx Tibia Calcaneus Scapula Ulna Femur Unknown 1 1 1 1 Cattle 1 Pig 1 1 5 Unidentified

Table 10.13. Bones Showing Cut or Chopping Marks, Huts 1 and 2 Bone Cattle Pig Sheep/Goat Deer Unknown 1 Pelvis 2* Vertebra 1 1 Mandibula 1 Cornu 1 Antler 1 Femur 1 Scapula 1 Metapodium 5 Unknown *These two were chopped

303

Table 10.14. Bones Showing Evidence of Burning, Huts 1 and 2 Bone Cattle Pig Sheep/Goat Horse Deer Mussel Unknown 1 1 Astragalus 1 Femur 1 Fibula 1 1 Humerus 1 2 Mandibula 1 Metacarpus 1 Metapodium 1 Metatarsus 3 Phalanx 3 Radius 1 1 Scapula 2 Shell 1 1 2 1 1 Teeth 3 Vertebra 62 Unknown 10 4 4 2 1 2 70 Total

Total 2 1 1 2 3 1 1 1 3 3 2 2 6 3 62 93

Table 10.16. Comparison of MNI Totals of Urbárske Sedliská Huts with Those from Moravia (with information adapted from Droberjar 1997, 150) Ur. Sed. Mikulčice Komořany Křepice Animal 1 2 II Z-9 VI X XIV XIV Cattle 7 1 8 9 6 4 8 5 Pig 10 2 2 4 5 1 1 1 Sheep/Goat 3 2 0 5 1 1 1 2 Horse 1 1 0 8 0 1 2 0 Bird, dom. 1 0 0 0 0 0 0 0 Deer 1 2 0 0 0 0 0 0 Total 23 8 10 26 12 7 12 8

304

Total 48 26 15 13 1 3 106

Charts Chart 10.1. Animals Present at Urbarske sedliska by number of bones found, without the unidentified category Mussel, 4, 1.3%

Snail, 1, 0.3%

Deer, 10, 3.2%

Wild Ox, 1, 0.3% Bird, Dom., 1, 0.3%

Bird, Wild, 2, 0.6% Bear, 1, 0.3% Sheep/Goat, 30, 9.7% Cattle, 125, 40.3%

Pig, 126, 40.6% Horse, 9, 2.9%

305

Figures

Figure 10.1. Grave-goods from a German elite burial at Zohor. Grave 5, first half of the second century AD 1. bronze jug – oinochoe (Eggers type 127), 2. bronze jug (type unknown, closest to Eggers type 122), 3. bronze patera (Eggers type 155), 4. glass bowl, 5. silver bowl, 6. bronze knife, 7. silver fibulae, Pannonian trumpet type (Hull type 162), 8. gold bracelet, 9. strainer, and 10. ladle (both 9 and 10 are Eggers type 162) (Kraskovská 1978, 14–15). Modified image from Krekovič 2000a, 48.

306

Figure 10.2. Silver lanx from Krakovany-Stráže (early second century AD, but found in an elite burial dating to thrid century AD). Top image, lanx without handles; bottom image, reconstructed with handles (top, image from Kolník 1979, Obr. 48; bottom, modified image from Holčík and Turčan 1998, Obr. 6).

307

Figure 10.3. Various Roman fibulae found in Slovakia. Fibulae 1–5 and 9–10 Noric-Pannonian origin, Fibulae 6–8 western origin (modified images from Krekovič 2001, Figs. 2 and 3).

Figure 10.4. Bronze armband from Grave 26, Abrahám (modified image from Kolník 1980, Taf. XVIII:26,c).

308

Figure 10.5. Reconstruction of a noric-pannonian style belt (modified image from Droberjar 2002, 225).

A B Figure 10.6. Two examples of belt-buckles. Both are bronze and from the cemetery at Abrahám, A is from Grave 40 and B is from Grave 67 (Kolník 1980, Taf. XIX:40,d; Taf. XXIII:67,h).

309

Figure 10.7. Partial inventory of items from Grave 35, Kostolná pri Dunaji. Items d–j are from a Noric-Pannonian style belt(s). Item c is a hairpin. All items shown are of bronze, except v, which is bone (Kolník 1980, Taf. C).

Figure 10.8. Belt end-piece with spoon-like end (Droberjar 2002, 207).

310

Figure 10.9. Iron knife blade with tong from Grave 44, Abrahám (Kolník 1980, Taf. XIX:44,d).

Figure 10.10. Iron scissors from Grave 40, Abrahám (Kolník 1980, Taf. XIX:40,b).

Figure 10.11. Inventory of items from Huts 1 and 2, Urbárske Sedliská (Drawings by Peter Šimčík). 311

Figure 10.12. Gaming stone/sling-bullet.

Figure 10.13. Gaming stone/sling-bullet.

Figure 10.14. Pestle.

312

Figure 10.15. Broken pestle (15), a mortar (16) and a whetstone (13).

Figure 10.16. Reverse of the broken pestle, mortar and whetstone.

313

Figure 10.17. A possible polishing stone for applying graphite.

Figure 10.18. The four spindle whorls recovered from Units 3 and 4 (Drawings by Peter Šimčík).

314

PART III SYNTHESIS

2.1 and 2.2 for chronology charts), primarily bronze vessels of the Eggers type 31 and fibulae of the Almgren 45, 47/48, 67, and 67/68 types (Kolník 1971, 513) (Fig. 3.4). There is a marked similarity between the early grave goods at these three cemeteries and those seen in the last phase of burials in the Czech Republic at the cemeteries of Tišice and Dobřichov-Pičhora, which helps support the view of a steady progression of Germans from west to east in this region (Kolník 1970, 519–520; 1980, 8). The earliest German settlement identified in southwest Slovakia is at Bratislava-Dúbravka, on a hillside, occupying an area roughly two hectares in size, and four kilometers east of the confluence of the Morava and Danube Rivers (Elschek 1995, 39) (Figs. 3.4 and 3.14). Two pit-houses were found dating to between AD 25/30 and 50/55 (the B1b period), with continued habitation into the following periods, with an unfinished Roman-style bathhouse added in the third century.

Chapter 11 German Identity in the Shadow of the Roman Empire The chapter has two objectives: 1) to provide a summary of what we know about the Quadi and what information was obtained from the project at Urbárske Sedliská; and 2) to place the findings within the context of the discussion in Chapter 4 about frontiers and German identity. 11.1. Southwest Slovakia: Archaeological Overview of the Quadi Presence

The Romans considered Vannius and the territory of southwest Slovakia to be within their sphere of influence, termed the hegemonia by C. R. Whittaker (2004, 40), and so meddled in German affairs on a regular basis (including the placement/replacement of various German kings). The frontier of the Roman Empire was not a static and closed defensive line, but instead a zone that straddled both sides of the Danube River and the Romans crossed the river whenever they thought necessary for political or military reasons (Wells 1999, 126; 2005, 50; Whittaker 2004, 5–6, 9). On a day-to-day basis, merchants and other provincials visited Quadi territory and Germans came to the markets at various frontier settlements in Pannonia. This contact along the Middle Danube enabled the creation of a complex interplay between the two groups and with others from various parts of the Mediterranean who were at the frontier as well.

The Germanic Quadi primarily lived in pit-houses, made of a wooden frame, with walls either of wattle-and-daub or wood and with a roof made of thatch (Fig. 7.1). The Germans would excavate an area to a depth between 20 cm and 50 cm, thereby creating a sunken living-floor; the huts usually covered an area from 9 m2 to 20 m2. The most common form is the six-post design, where there are two posts along the long walls, one in the middle of each short wall, while sometimes a seventh might be added at the center of the hut to prop up the ridge pole (Figs. 7.3 and 7.4). Settlements are usually found on terraces near rivers and streams; or on high ground, which are often buried sand dunes located near some water sources; or at the foot of low hills facing toward the south near some small stream. In general, German settlements were not large-scale urban efforts, but instead isolated farmsteads, hamlets, and villages. Most settlements ranged in number from two to forty huts, covering anywhere from a quarter hectare to ten hectares. Amidst the huts would be other cultural features such as storage and fire pits, as well as above-ground structures. The huts were not only used as homes but also as sites of industrial production, with some huts having kilns and furnaces. The settlements are usually multi-phase, with huts being abandoned and new ones built nearby, but rarely on top of previous huts. This pattern may explain why settlements cover large areas and are oblong, with huts being abandoned and new ones built heading in one direction, possibly following a watercourse. The Quadi had easy access to the Roman frontier, which was the southern border of their territory, with some settlements having been found closer than five miles from the border. These types of settlements and hut architecture continued relatively unchanged from the early first to fourth centuries AD, which shows almost a deliberate continuation of traditional ways by the Germans, even though they were well aware of the situation in Pannonia.

The earliest evidence for a German presence in southwest Slovakia comes from cemeteries, especially those at Sládkovičovo, Abrahám, and Kostolná pri Dunaji, which have material dating to AD 10–25/30 (B1a period; see Table 2.1 for an explanation of the periods and figures

The huts themselves usually lasted around twenty years, at which time any useable building material was removed and the rest burnt. The remaining pit was filled with the charred wood, sintered daub, and not only the trash left behind in the hut, but also whatever else was thrown in at

The archaeological evidence shows that German groups started to penetrate into southwest Slovakia by the first decade of the first century AD, having first come into Bohemia and Moravia during the first century BC (Fig. 2.6). Once in Slovakia, the Germans expanded quickly across the southern part of the country, creating settlements along the major rivers (Figs. 2.8, 3.3, and 3.4). This settlement was facilitated by the Romans, who placed the retinues of Maroboduus and Catualda in southwest Slovakia, between the Morava and Váh Rivers, under the rule of Vannius, a Quadi king whose power base had originally been centered in southeast Moravia. Vannius also gained power over neighboring tribal groups, such as the Cotini and Osi, while also accumulating wealth through his control of the portion of the Amber Route that crossed Quadi territory.

317

later dates. It appears that once a hut was abandoned and torn down, its pit was used as a trash dump by other huts in the area. Therefore, these huts are usually rich in material culture, albeit fragmentary, which may not be contemporaneous with the ruined hut. The huts have a mixture of pottery, bones, metal items, as well as various stones, jewelry, and spindle whorls. Bones from various domestic animals and pottery are by far the most common finds. When discovered, these huts appear as rectangular stains, dark in color, with the earth organic rich and even greasy to the touch because of the high concentration of animal remains (Fig. 7.25).

Bronze and other metal import items found in German huts and graves vary from the most common, such as fibulae and belt decorations, to less common hairpins, buckets (situlae), ladles (paterae), and strainers (Figs. 10.1, 10.3–10.7). The fibulae are usually of bronze and most found in Slovakia date to the first century AD. A common find in graves and in lesser numbers at settlements are the parts to belts, such as buckles, stud decorations, and other ornamentation (Figs. 10.5–10.7). The majority of Roman bronze vessels are found in graves, with 70% of the total amount coming from just three cemeteries: Sládkovičovo, Abrahám, and Kostolná pri Dunaji (Krekovič 1997a, 233). Very rare finds are Roman lamps, glassware, and beads. These items, however, are always in small quantities, unlike Roman pottery, but all the imports (including the pottery) comprise a very small portion of the total cultural remains from huts and most graves.

The most ubiquitous human-made find is German pottery, which is varied in form and decoration. During the first two centuries AD, they were all handmade, with wheel-made pottery coming into widespread use by the third century, eventually replacing the handmade vessels. German pottery can be divided into two broad groups: fine wares and coarse wares. Both types have similar vessel shapes, with most falling into three types: 1) onepiece forms, 2) two-piece forms, and 3) three-piece forms (Figs. 8.3, 8.5, and 8.8), with two-piece bowls, cooking pots, and storage vessels being the most common. Fineware vessels are usually dark gray to black or have an applied black finish, usually through either polishing or applied graphite, while their forms have sharper edges than coarse wares. Some German fine-ware forms appear to imitate Roman vessels, both ceramic and metal, as well as some glassware. It is thought that the fine-ware vessels may have been produced in centralized locations and then traded, while coarse ware may have been produced at each settlement or even individual households. Coarse-ware vessels are more commonly found on German sites than the fine wares. These vessels have a wide variety of shapes, colors and surface decoration. The quality of the fabric also varies from vessel to vessel. The most common surface decorations are thumb impressions, comb lines, and incised lines usually in a criss-cross pattern (Figs. 8.33–8.47).

Huts have provided much evidence for German animal husbandry and, according to Tacitus, a person’s wealth was measured in the number of cattle one owned (Tac. Germ. 5.1; Todd 2004, 76). The archaeological evidence bears out Tacitus’ statement, with huts in Moravia and Slovakia having yielded a large volume of animal remains of which cattle usually dominates the assemblages. Closely following cattle bones in number are pig and sheep/goats, while horse and chicken remains are fewer in number. Hunting does not seem to have been important to the Germans with deer bones being found always in small numbers, followed distantly by bones from wild boar, aurochs, pheasant, bear, and elk. There is also some evidence showing exploitation of marine resources such as fish and shellfish. In terms of cultural complexity, most scholars agree that the Germans by the first century AD were organized into complex chiefdoms, which by the fifth century AD had transformed into early kingdoms and state-level complexity. The Romans, however, labeled the German chiefs of the first century AD as “reges,” an act which has sometimes led to misleading discussions of kings in scholarly work. The scanty evidence, both literary and archaeological, shows that the ancient Germans most likely started as bands, but then because of contact with the Celts and Romans reformed into tribes. The tribes were originally led by a tribal council made up of elders, from which a peace-chief (Romans referred to him as the rex) and war-chief (Romans referred to him as the dux) could be elected.1 Over time, with the increasing contact with the more sedentary Celts and urbanized Romans, the German tribes disbanded and reformed usually in larger bodies and sometimes confederacies; as a result, the tribal council became unwieldy in effectively ruling the large tribal groups made up of various sub-tribes. This situation led to the usurpation of power by the rex and dux, a situation exacerbated by the Romans, who preferred to deal with one ruler as opposed to oligarchies. There was also a power struggle between the rex and dux, eventually leading to the melding of the two into the position of king as seen in Late Antiquity and the Early Middle Ages.

Huts and graves often contain various Roman imports. Roman pottery is the most common import found in huts. The pottery varies from the most mundane of cooking pots to high-quality terra sigillata. The most common pottery imports were produced in Pannonia, such as Brick-Colored Wares, Fine-Gray Wares and Coarse-Gray Wares, while specialized sub-forms such as Pannonian Stamped Wares, and Pannonian Striped Wares are rarer (Figs. 9.2–9.6). Imports from more distant places within the empire are also found, such as the already mentioned terra sigillata, some Raetian wares, and Italian thinwalled wares. A specialized Roman vessel very popular with the Quadi was the mortarium, which could be used not only to grind grains, but also, as postulated for some of the smaller vessels with a smooth interior surface, to produce various milk products, or a popular vegetablebased sauce called moretum. These could also have been used to make beer, a very popular beverage among ancient Germans.

318

During the Roman Period, German society became particularly warlike, with male identity tied to valor on the battlefield and the amount of loot he could obtain. While the position of rex was hereditary, the position of the dux was solely based on his abilities as a leader of men and how much loot he could obtain to distribute to his followers. The more able the dux, the greater his retinue, termed the comitatus by the Romans. The retinue could be made up of men from various tribes and so these men owed their allegiance only to the dux, having broken ties with the tribes they left; this development is similar to what was seen in the Late Republic, when Roman soldiers owed their allegiance more to the general they followed, such as Marius, Sulla, Pompey, or Caesar, than to the Republic. The retinue helped to destabilize further the original tribal structure of German society, with the dux being forced to obtain somehow the loot necessary to keep the retinue loyal and so enable him to hold onto power. The loot was obtained either through raiding other German tribes or the Roman provinces, or the redistribution of tribute and gifts the dux received from various sources. The need for loot helped to increase the role of warfare in German society. The Romans worsened the situation by wanting to deal with single rulers, either the rex or dux, thereby legitimizing the position of the dux; the Romans also provided gifts and tribute to the dux, further enhancing the power of the dux in the eyes of his retinue.

11.2. Summary of the Project at Stupava-Urbárske Sedliská Stupava was chosen because of an intriguing structure, referred to as the “Roman Station” by Slovak archaeologists, which was first excavated in the late nineteenth century, with work continuing there to this day. Stupava is located between the Morava River and the western foothills of the Little Carpathian Mountains, 16 km north of the Danube River, which was the Roman frontier (Figs. 3.12 and 3.14). Dr. Turčan and his partner, Ing. Staník, a specialist in ancient architecture and building technology, have been able to identify three phases to the site, with sub-periods of abandonment, reconstruction and improvement (Staník and Turčan 2000, 23) (Fig. 3.16). The complex started as a single large structure, with timber walls or wattle-and-daub, with plaster on the outside and the roof covered by Roman roof-tiles. The date for the start of this phase is disputed with some archaeologists saying late first century AD or early second, while Staník and Turčan believe it to be the middle of the second century (Bujna 1976, 511; Hečková 1986, 390–391; Staník and Turčan 2000, 23; Turčan 2002, 235).2 It appears that the builders first leveled parts of the area for better construction and filled in some German dwellings on the hilltop, one of which also had Roman legionary shoe-nails. The second phase saw the structure rebuilt into a two-wing building with seven rooms of varying size. The walls were now of mortar and stone, with thicknesses of up to 65–70 cm, and may have also supported a second floor. The foundations of the walls reused material from the first phase, including roof-tiles. The structure was then enclosed by a low wall that marked a large area, 65x70 m. The third phase saw the structure expanded, with added rooms and polychrome wall-painting for decoration (Fig. 3.16). Excavations in the 1990s discovered a collapsed wall in the central courtyard that had two arched and stuccoed window frames (Fig. 3.17). The site became a complex of buildings surrounded by the enclosure wall, with a bathhouse and cistern added in the southern part of the enclosure, while at a later period, part of the enclosure wall was torn down for the addition of a horreum (Pitts 1987, 223–225; Turčan 1990, 40–41) (Fig. 3.15 and 3.21). The entire complex appears to have been abandoned some time in the later third century by the Romans, or whoever occupied it (Staník and Turčan 2000, 25–26).

The Germans along the Middle Danube became increasingly entangled in the politics of the Roman Empire. At the same time, the economy of the Germans also became increasingly tied to and dependent upon the Roman one. It can also be argued that the economy of the Middle Danube provinces, especially Pannonia, became more dependent on the Germans as well. The eastward progression from the first to third centuries AD of terra sigillata, other types of pottery, and bronze vessel production should be seen as a reflection not only of the prosperity along the Middle Danube in the provinces but also the prosperity and buying power of the Germans (Drummond and Nelson 1994, 164, 166). We see archaeologically a flood of Roman goods into Quadi territory, but these items need to be divorced from their “Roman” identity. The Roman items were obtained through looting, gift-exchange, or bartering by the Germans and were used within the context of German society. These import items were seen not only as functional but also as a means of increasing the social prestige of the owner. Even though there is a large volume of Roman material and elites do accumulate more expensive items, the total volume of material is spread out through all levels of society in small amounts. The average German, while having access to Roman material, did not in fact own a large amount and what they did own was largely common pottery, easily obtained from traveling merchants, frontier markets, or from each other (we should not discount trade and exchange among households). While this pottery was used in various household activities, it also played a role in projecting an image of prosperity for the user.

A great quantity of Roman tiles has been found, all dating to the second century AD. Many of the tiles have stamps made by Legions X Gemina PF, XIII Gemina, XIIII Gemina MV, XV Apollinaris, and some private tilers from Carnuntum, such as G.I.F., C.I. IUL, and C.VAL.CONST (Ondrouch 1945/46, 14–15; Bujna 1976, 510; Hečková 1986, 383 and 390, Tables 1 and 2; Pitts 1987, 225) (Figs. 3.25 and 3.26). Coins have also been found, primarily dating to the first third of the second century, with others dating to the third and fourth centuries, but one gold aureus of Vespasian was also found (Hečková 1986, 382 and 391; Rajtár 1986, 106; Table 4; Staník and Turčan 2000, 26). The two legions, 319

XIII and XV, which stamped most of the tiles at Stupava are attested by other records as having been stationed at Vindobona (Vienna) and Carnuntum before AD 115. Roman provincial pottery is in abundance at the site, ranging from various Pannonian wares to terra sigillata, especially Rheinzabern, Westendorf, and Pfaffenhofen Wares, with 83 pieces being of the Rheinzabern type, dating to Antoninus Pius and Marcus Aurelius (Kuzmová 1997, 49; Tabelle 12). The evidence for an actual Roman presence, along with ceramics and coins, includes the iron shoe-nails characteristic of Roman legionary footwear, iron ringlets from chain mail (dating to the middle of the second century), and a bronze lion head that probably came from a military parade helmet (Staník and Turčan 2000, 23) (Fig. 3.23).

huts fit his Type III, which have a rectangular shape, two posts along each long wall, and a post in the middle of the shorter ends (Figs. 7.8 and 7.24). The huts were dug into the ground to a depth of 10–15 cm with the entrance in the middle of the south wall, facing toward the southwest, an arrangement that is not unusual for German huts. A ridge pole would have gone down the middle, which would have held the superstructure of the roof. The walls were made of wattle-and-daub overlying a wooden lattice woven around the support poles. Pieces of burned wattle-and-daub were found in the fill, some of which have the impression of branches from the wooden lattice underneath (Figs. 7.15, 7.16, and 7.30). The roof would have been thatched. Hut 1’s overall size was 5.135 m x 3 m, while Hut 2 was 4.87 m x 3.54 m (Figs. 7.8 and 7.24). Usually for these huts, the living floor was either packed earth or a wooden floor. The two huts at Urbárske Sedliská did not have a living floor, or at least no thin layer of hard-packed earth was excavated, which would seem to imply that there may have been a wooden floor. Hut 1 was not well preserved: a modern trench was cut through it in 1982 for a drainage pipe, and plowing may have sheared away several upper centimeters (Figs. 7.7 and 7.8). Hut 2 is much better preserved, having escaped plough damage by being buried deeper (Figs. 7.7 and 7.24).

Turčan and Staník argue that the site went through different phases of use and that the different construction phases reflect these changes (Staník and Turčan 2000, 23–24, 26; Turčan 2002, 235). The first phase, with the presence of military equipment, make the site a probable Roman outpost, which by the third phase had taken on an economic role, since part of the Amber Route passed along the Morava River and Stupava may have acted as an emporium for Roman merchants, whom Tacitus records as having operated among the Germans (Tac. Ann. 2.62). The third phase of the complex shares similarities with the villae rusticae seen in Pannonia and so may have had a dual function as an emporium and elite residence.

Hut 1 yielded over 500 pieces of pottery, of which only 18 were Roman, along with animal bones and two Roman bronze pieces, one being a needle and the other a fibula (Figs. 9.7, 9.8, and 10.11). The Roman fibula dates to between AD 50 and 100 based on Almgren’s typology, being a type 68. The Roman pottery was too fragmentary for secure dating purposes, but primarily fit within the types seen during the second century AD. Using the chronology for the Roman Period in Slovakia, the German pottery had seven diagnostic sherds fitting within a range of AD 50–150 and another three dating to the first half of the second century, providing a probable date of the first half of the second century; this date was further corroborated by the radiocarbon sample that provided an AMS date of 1870 ± 40 BP, with an intercept date at Cal AD 130. The German pottery had various types of decorative elements and were all handmade.

Urbárske Sedliská lies 800 m to the northwest of the Stupava site (Fig. 3.22). Archaeological work was first done at Urbárske Sedliská in 1982 when a Grubenhaus was excavated along with a Bronze-Age burial. The 2002 project was designed to explore further the area in order to locate more ancient German habitation. Since work had not been done at the site for 20 years, there was a need to locate the best areas for archaeological exploration. A field survey was designed toward this goal, using shovel test pits (Figs. 5.3 and 5.4). The survey recovered various types of artifacts, of which the ceramics that could be identified came from primarily three periods: the first being the Neolithic/Eneolithic, represented primarily by the LBK, Lengyel, and some Bajč-Retz; the second was the Roman Period with mainly ancient German ceramics; and the third was the Modern Period.

Hut 2 yielded over 600 pieces of pottery, of which only 28 were Roman, and once again a large amount of animal bone was present, along with a Roman bronze hairpin, another needle, two bronze pieces that would have been decorative pieces on a belt and a glass-paste bead for a necklace (Figs. 9.7, 9.8, and 10.11). The Roman pieces primarily came from storage vessels and the decorative element on some of the body-sherds shows an origin from Pannonia. The German pieces again could be identified as dating to the second century AD, while the Roman pottery, although fragmentary, again indicated a second century date. The hairpin is similar to examples found in German graves, such as at Abrahám, dating to between AD 10 and 180. The radiocarbon date also indicates a late first century or early second century date, with the radiometric date being 1910 ± 60 BP, and an intercept date of Cal AD 90.

The survey was successful in that two shovel test pits had high concentrations of German ceramics mixed with animal bones in an organic-rich soil. These areas were the focus of the second phase of the project, when larger excavation units were opened that led to the discovery of two German huts (Figs. 6.1, 6.2, and 7.7). Their discovery was based on the soil composition of the huts in relation to their surroundings. The huts were filled with a rich dark grey to black soil, having charcoal flecking, and a large amount of material remains and animal bones (Figs. 7.11 and 7.25). The layout of the two huts corresponds to the classic Grubenhaus design and, according to the typology devised by Kolník, the 320

The bones showed a varied diet, with cow, deer, sheep/goats, and pig appearing, but the largest percentage were of cow and pig. Other bones also appeared, coming from horse, birds, wild ox, boar, and bear. Some of the bones show butchering marks, while others were burned. Among the deer remains were antler fragments that appear to have been worked. The bear is a surprising find, represented by one tooth coming from the species Ursus arctos, or Brown Bear commonly referred to as the Grizzly Bear. Along with animal bones, there were also shell fragments from freshwater mussels and one snail shell.

other sites in Quadi territory; it would thus appear that whatever was being made in the hut was intended for Germans (possibly further north at Zohor or Láb, where the rich burials were found, see Chapter 3.2 and 3.5) or was being sent directly to the frontier.3 The two huts only 150 m from the Roman-style structure were excavated in the mid-1990s by Turčan in an area labeled Morávkove Pole. Hut 1 (labeled as 1/96) had a size of 5.10 x 4.20 meters with eight post-holes placed in a more hap-hazard fashion in comparison to the two huts from Urbárske Sedliská (Turčan 2005, 293 and 295). When constructed, the hut was dug to a greater depth, reaching in some spots up to 50 cm (Turčan 2005, 293). Just as with other huts, this one too was dug into a sand stratum. The hut had two phases: the first phase was a normal living place and the second was a workshop, having a kiln or oven built in the hut, with metal slag being found in and around the oven. The pottery from the hut had a wide assortment of German pottery and also 89 Roman sherds, of which 23 were terra sigillata. Three pieces of terra sigillata that may be of the same vessel appear to be Pfaffenhofen ware, dating to the first half of the third century AD (Turčan 2005, 295). An almost complete Pannonian Stamped-Ware bowl was found in the hut, identified as a Ring Bowl in form with rouletting on the interior and exterior, dating to the early third century AD (Turčan 2005, 295). Other finds included Roman glass, a comb, and a spindle whorl (Turčan 2005, 298). Two coins were also found: one was a denarius of Elagabulus (218–222) and the other an antoninianus of Probus (276–282), which had a hole drilled into it for probable use as jewelry (Turčan 2005, 293 and 298). Hut 1 then dates to sometime in the second half of third century AD (Turčan 2005, 299).

The two huts, according to finds, date to some time in the first half of the second century, just before or during the construction of the first phase of the Roman-style structure. The huts, therefore, provide an opportunity to see German material culture before or just at the start of the greater exposure to Roman goods known at Stupava. The two huts discovered at Urbárske Sedliská were compared to the previously excavated hut at the site, along with two huts found just outside the Roman-style structure in Stupava and with the finds from that structure (Fig. 3.22). Other huts were also used for comparative purposes, including some from the site of BratislavaDúbravka located to the south of Stupava, closer to the Roman frontier, and huts from Moravia because these were located further from the frontier and north of Stupava. Below, I provide brief descriptions of the other hut from Urbárske Sedliská and the two from Stupava, because of their proximity to the two huts I discovered. The hut excavated in 1982 at Urbárske Sedliská has been mentioned already in Chapter One. The hut was a Grubenhaus of trapezoidal shape, the longitudinal axis of which was oriented northeast to southwest, having a size of 4.8 x 3.9 meters (Turčan 1985, 93) (Fig. 1.13). The entrance to the structure could not be determined, and only two post-holes were found, one in the middle of the northeast wall and the other in the middle of the structure; the shape of this Grubenhaus does not accord with the types set forth by Kolník (Turčan 1985, 95). The pottery was primarily German, but a total of 59 provincial and Roman pieces were found as well (Turčan 1985, 107) (Figs. 1.14 and 1.15). The German ceramics were mostly open shapes and one pot could be reconstructed (Turčan 1985, 107). Based on these and other pieces, Dr. Turčan dates the hut to the middle of the second century AD, which places the site roughly in the same time period as the first phase of the Roman-style structure.

Hut 2 in comparison to the other four huts already mentioned was extremely rich in material remains. The hut’s size was similar to the others being 4.2 x 3.8 meters in size, having six post-holes and dug to a depth of 10–20 cm (Turčan 1997). Over 1,500 German sherds were recovered, while Roman sherds numbered at 330 pieces, of which only one was terra sigillata. Thirty-one pieces of metal, including an iron ring and two fibulae were recovered. Turčan dates the hut to the third century (Turčan 1997). Chronologically then, the five huts roughly correspond to the life-span of the Roman-style structure. The two huts I excavated in 2002 date to just before and during the first phase of the Roman-style structure, while the hut excavated in 1982 dates to the later part of the first phase or even early second phase. The last two huts, located closest to the structure, date to the end of the third phase or just after the abandonment of the structure. The pottery findings from the five huts are presented in Table 11.1 below. As mentioned in Chapter 3, there are no published reports on the total amount and types of pottery excavated from the Roman-style structure. We do know, however, that as of 1997, there were 159 pieces of terra sigillata discovered and hundreds of examples of various Pannonian wares at the site. What can be clearly seen in

The most surprising discovery from this hut was a series of small crucibles, only one of which was whole. Crucibles are containers made to hold metals for melting and fusing or for calcining ores, metals, etc. The crucibles are rounded with pinching of the walls to create depressions, and are of a type known to have been used in the working of precious metals for jewelry during the La Tène period (Turčan 1985, 110). They have been found both in oppida and also in settlements. The Roman-style structure and associated complex at Stupava have few pieces of jewelry, or other metal objects in comparison to 321

the pottery totals of the various huts in the Stupava area is a small increase in the quantity of Roman pottery over time and they are never in the majority in the huts. The youngest huts do show a marked increase in Roman pottery in comparison with the older huts, but this difference is lessened by the fact that there is also a greater amount of German pottery as well in the second hut at Morávkove Pole (Table 11.1). It would seem then that the presence of the Roman-style structure had little to no effect on the ability of local Germans to have easier or greater access to Roman pottery. This finding is further strengthened by the small presence of various Roman bronze items, coins, glass wares, and other material in the huts that could have been obtained, one would think, in greater quantity from the Roman-style structure.

Table 11.2. Summary of pottery totals found in huts dating to AD 50–166 Pottery Type Hut German Roman Other US* Hut 1 493 18 3 US Hut 2 661 28 6 Dúbravka 13/88 503 3 6 Dúbravka 129/93 282 0 8 Blučina 4 211 17 0 Komořany Z-9 196 31 0 Křepice 2 308 141 0 Křepice 7 249 167 0 Rajhrad 2 98 5 0 * US Hut 1 and 2 refers to Urbárske Sedliská and the two huts excavated in 2002.

Table 11.1. Summary of pottery totals from the five huts in the Stupava area Pottery Type German Roman Other Hut Hut 1 493 18 3 Hut 2 661 28 6 US82 753 59 0 Morávkove Pole 1 205 89 0 Morávkove Pole 2 1458 330 0

Table 11.3. Summary of pottery totals found in huts dating to AD 166–180 Pottery Type Hut German Roman Other US 82 753 59 0 Blučina 1 400 63 0 Blučina 5 336 51 0 Komořany Z-2 89 30 0 Křepice 5 232 7 0 Křepice 17 111 19 0 Rajhrad 2 98 5 0

When we take the data from the five Stupava huts and add in data from the Dúbravka huts and the group of huts from Moravia, the absence of an impact by the presence of the Roman-style structure is made even clearer. Tables 11.2–11.4 groups the data from the huts into three periods based on the date of each hut: 1) AD 50–166, 2) AD 166–180 and 3) AD 180–250. The quantity of German and Roman pottery in the Stupava huts is in fact not all that different from the huts found at Dúbravka and further north in Moravia. This finding is somewhat surprising since we would have expected those huts closer not only to the frontier but also to a structure filled with Roman pottery and other imports from the provinces to have had a greater impact on local Germans. Imported pottery constitutes no more than 12% of the total pottery from the huts dating to AD 50–166 (12%) and 10% for those dating to AD 166–180, while huts dating to AD 180–250 have a modest increase to 19%. When we include all other artifacts, such as bones, metal items, etc., we see that Roman imports are an even smaller part of the total amount of artifacts recovered. Huts 1 and 2 from Urbárske Sedliská, for example, have Roman imports of only 1.4% in Hut 1 and 2% in Hut 2 (see Charts 11.1 and 11.2 at the end of this chapter).

Table 11.4. Summary of pottery totals found in huts dating to AD 180–250 Pottery Type Hut German Roman Other Morávkove 205 89 0 Pole 1 Morávkove 1458 330 0 Pole 2 Mušov 1 117 21 0 Mušov 2 236 51 0 Mušov 3 241 25 0 The data discussed above fit in with the statement made earlier that the average German did not own a great amount of Roman material. These small amounts and other findings (see below) show that the traditional view of Romanization, in which Roman material is considered superior by Germans and that they acquired them to become more Roman, can be clearly seen as out-dated. If Roman goods were considered superior, then we should have seen a substantial increase over time in the amount of Roman goods in German huts, but in fact the data (both at Urbárske Sedliská and further away in Morava, see Chapter 9.3) do not support this conclusion. We need to look toward German society itself and try to find a “German” context for the role of Roman goods, even in the limited quantities that we do find archaeologically.

322

11.3. How Does the Project at Urbárske Sedliská Enable a Better Understanding of Ancient German Identity?

Roman material coming into the German huts at all levels of society had some type of meaning for the user, whether the object was seen as merely more functional than the German counterpart or signifying prosperity in comparison to the hut next door that did not have the same material. In either case, the German was applying a new meaning to the object and not necessarily re-using the “Roman” identity attached to the object. Indeed, the German user may have been totally unaware of the object’s original meaning, use, or identity.

The challenge for archaeologists is to tease out an accurate picture of what was happening to the Quadi culturally once they settled in southwest Slovakia and became increasingly oriented toward the Roman Empire. An easy term to latch onto is “Romanization,” which in the past has been used by some scholars to describe what was happening to the Germans along the frontier.4 However, in the last decade there has been a reassessment of Romanization, with more attention given to the role of the non-Roman as an active participant in cultural interaction, leading to the rejection of the idea that locals were merely empty receptacles waiting to be filled with elements of a superior culture (Hingley 1996, 40; 1997, 84–85; Barrett 1997, 51; Mattingly 1997, 8–9; Wells 1999, 126–127; 2005, 50 and 95). We need to take into consideration how the native society used Roman material culture to fit their needs, while maintaining a German identity (Hingley 1996, 41; Jones 1997, 134). Identity is not static and is not imposed on one group by another group (sometimes termed the “other”), but instead is self-created, influenced by how the “other” views that group and also how that group maintains difference from the “other” (Shennan 1994, 16; Ucko 1994, xviii; Jones 1997, 28, 52–53). Material culture plays an important role in the creation and maintaining of identity and can be considered one of the critical markers in identity (Shennan 1994, 16; Jones 1997, 28). This type of viewpoint has not been explored adequately in the past in reference to the ancient Germans, and even when Romanization was rejected there was little effort to determine how material culture was creating a “German identity.” There has also been too much emphasis on the elites, primarily caused by the spectacular elite burials in the lands beyond the frontier, but in fact all levels of German society were concerned with enhancing their identity and it is clear that easily obtained Roman goods, even if in small quantities, played a role. The enhancement of identity through the use of goods was of course an important part of elite behavior in Italy during the Republic and Imperial Period and, as the empire grew, the incorporated elites from other places, such as Gaul, also participated in similar behavior (Woolf 1998, 169); of course we see this too among the Quadi elites. This action also took place among the lower classes, sometimes in emulation of the elites, but also independently within the context of their own class (Woolf 1998, 170).

We must keep in mind from where Germans were getting Roman material. Did Germans always obtain their items from Roman merchants or looting Roman houses in the provinces? I believe we need to keep in mind the preexisting groups in the region. When the Romans eventually created the provinces along the Danube, they did not eradicate the local population and supplant them with people from the city of Rome or the Italian peninsula. Instead, the local population was left largely intact, sometimes with their elites still in place, while a thin “Roman” veneer was put over the territory. This veneer deepened over time to become a thick lacquer, but we have to remember that the local population, especially in the countryside was sometimes untouched by the Roman presence (Saddington 1991, 417). Peter Wells, in a recent paper, cites D. Gabler in Hungary where research shows for the Kapos Valley and Zala County that some local Pannonian populations, in the imperial period, were living unchanged from the earlier Iron Age, with little to no Roman goods found in their material culture (Gabler 1991, 425; Wells 2005, 71).5 This example helps to show that even though a population may have been surrounded by new material culture, traditional identity could be maintained. These are groups that the Germans may have had contact with as well and so would have also played a role in how Germans reacted to Roman goods appearing in markets in Pannonia and peddled by traveling merchants. When the Germans settled into the region of southwest Slovakia, Moravia and Bohemia, it was a gradual process, whereby they absorbed the trade-routes and contacts already established in the area, so that, when they were trading, they were doing so with the Celtic, German, Thracian, Illyrian and Dacian populations absorbed into the Roman Empire in the provinces along the Danube. Of course, in some instances Roman merchants were there as well, but they too were probably following the established routes of contact. Southwest Slovakia has an abundance of Roman material found amongst the Quadi, but does this necessarily mean “Romanization” of a culture or should we see this instead as merely the “Romanization” of the trade-goods available along the established routes of contact in the region? The Amber Route, for example, that went up the March River, onward into Poland to the Vistula, and there to the Baltic coast, was already in place and not created by the Romans. Therefore, the people in the region were already in contact with each other and influencing each other before the arrival of the Romans. We see this in Slovakia during the first half of the first century AD, with

In relation to the behavior of the various classes in maintaining or enhancing identity, an artifact does not need to support only one type of identity (i.e., Roman or Celtic; or elite). In a new context, or divorced for a long period from its origins, an artifact may acquire a new meaning or identity (Jones 1997, 116, 126; Shanks and Tilley 1987; Wells 2001, 25). Just because an object was made in the Roman Empire does not mean that it maintained a “Roman” identity or that it imparted a “Roman” identity on a user in another cultural group. 323

the earliest of German settlements. The huts found are in fact Celtic Late La Tène in shape and mixed in with the German pottery we also find Celtic pottery. We also find trade goods from the Przeworsk culture of southern Poland in various German huts in Slovakia and Moravia. As one moved further east, material started to come in from the Celto-Dacian Púchov culture (possibly Tacitus’ Cotini) of northern Slovakia and the Sarmatian Iazyges and Dacian tribes located in eastern Hungary and Romania.

If one day the traders are selling purely Celtic pots and the next day pots with Roman decorations, that is what the German buyer would barter for or purchase, having no choice. Overtime, various Roman items gained their own popularity and were sought after by Germans, either through the actions of elites first and then emulation by the lower class or through personal choices of the lower class. I would argue then that if we are going to use the term Romanization at the frontier, then we should see the Roman goods amongst the Germans as partly symbolic of the Romanization of the Pannonian population over time.

When we read the ancient sources, such as Tacitus, Strabo, or Pliny, we find mention of the movement of groups, such as the Celtic Boii and the Germanic Marcomanni in the region of the former Czechoslovakia. The impression we might gain is that they moved quickly, totally abandoned the place they left, and supplanted the population in some other place, thereby creating the impression of disruption in local trade routes and contacts with other groups. Archaeologically, this is not apparent. While movement of peoples did take place, the evidence shows steady infiltration and slow ascendancy of one group of material culture over another, but the original material culture also continues in much smaller amounts in isolated pockets. For example, there is always talk about how the Boii abandoned Bohemia sometime between the 60s and the 40s BC and that the Germanic Marcomanni and Quadi moved into the region sometime around 10 BC. Archaeologically, however, German material starts to appear in Bohemia already at the beginning of the first century BC. In Moravia there is a similar situation: Celtic settlements decline in number starting around 150 BC, and German material dominates the archaeological record by the end of the first century BC. This evidence helps to show the gradual spread of Germans into the region, thereby allowing the continuation of the established trade contacts and routes, such as the various parts of the Amber Route. At the same time, Celtic material is still found in the region, sometimes even within German huts, showing some kind of continued presence and contact (but also possible evidence of continued circulation of popular Celtic material among German huts) (Kolník 1991, 432).

We must reject the idea of Romanization for the Germans, or find Roman-oriented views of how the Germans were using these imports, or create a RomanoGerman identity for Germans beyond the frontier. The Roman goods we see in southwest Slovakia need to be seen within the context of German culture. By having even mundane cooking pots from Pannonia obtained through trade, the German who owned them may have been seen as prosperous since he could afford to obtain them, or as being part of a successful retinue, or that through the largess of the dux that person had been singled out for recognition. We cannot talk about any type of Romanization, since the various pots, terra sigillata, and bronze buckets were all seen within a purely German context. It is not that the Germans thought Roman items were inherently superior or that they wanted to live as Romans did; instead the acquisition of these goods was used to show their prosperity in a German context, much as in having a large herd of cattle. We have to imagine that this took place not only among elites, but also among the lower classes. Competition among households was probably a very important component rolled into the prestige and honor seen in the warrior lifestyle. In conjunction with the view expressed above, I would argue that the discovery of Roman winesets made of bronze and other materials in German context does not necessarily mean these items were being used to serve wine. We need to consider German uses for these items, such as the serving of beer or mead. The structure of German society helped to create a context for the Roman goods being found in their territory. By the first and second centuries AD, as already noted, German society had abandoned its traditional tribal structure in favor of powerful leaders, either termed kings (the rex) or war-leaders (the dux). The ownership, use, and redistribution of Roman and other high-value items were used to reinforce the control exercised by the elites. We see this in the dux who maintained his power through the distribution of loot and gifts, along with hosting banquets. The goods used by the rex and dux, especially Roman material, were important visual manifestations of power. The dux, and the elite in general, maintained their power through the creation of an image as successful people. Because the Germans were in a complex relationship with the Romans, who were seen as a powerful force by the Germans, one would expect the German elites to try to emulate within a German context certain Roman elements they saw as symbolic of power in order to re-enforce their

When the Romans do arrive in the last quarter of the first century BC in Pannonia and solidify their hold on the territory in 11 BC (first as part of the province of Illyria, then in AD 9 as the province of Pannonia), they do not wipe out the local Celtic, Illyrian, and Thracian groups. These people continued on, mostly in the way they had before with little impact by the Romans. In fact, as I mentioned above regarding the various Roman pottery imports, many of them, such as the Pannonian- Stamped Wares, Pannonian-Striped Wares, and even the Raetian Wares, have been shown to have strong local La Tène roots influencing their designs. It would seem the case then that when Germans were trading and acquiring goods from the provincial markets of the frontier, they were usually not dealing with Romans directly from Rome, but instead with the locals with whom they had been trading and interacting before. 324

own image of being successful and powerful. This circumstance can be seen in the elite burials with Roman goods or the Roman-style structures, such as at Stupava, Cífer-Pác, or Milanovce. I argue that they were not trying to be Roman, but merely using specific elements of what they thought necessary to project power in a way that would be understood by other Germans. In effect, they created a “Germanization” of Roman materials and ideas.

apparent in the literature. Personally, I am more inclined to believe in the late first or early second century date, but final say on this has to wait on completion of excavations at the site and full publication of the findings of all the various excavation seasons. 3. This conclusion may change once a complete analysis and inventory are published for all the finds at Stupava.

Connected to this view of the frontier and rejection of “Romanization” is the concept of identity. Much as the frontier may now be seen as a multi-level interaction zone; so too identity now may be seen as something created through the interplay of elements both within a group and from outside the group, whether it is how other groups perceived the group or the creation of boundaries to emphasize the difference between one group and another. Identity can be maintained through the use of traditional material culture within the group or through the manipulation of outside material within the mechanisms of the group (Hingley 1997, 86–87).

4. There has also been the polar opposite view, wherein the Germans were viewed as so different that whatever Roman material they had played no role whatsoever in their society, implying the Germans were in too primitive a state to ever truly appreciate those wonderful objects they had acquired from the Romans. This viewpoint is best seen in a comment made by Mortimer Wheeler in Rome Beyond the Imperial Frontiers (1954, 72), where he states, “…Save in so far as Roman weapons and armour influenced the German war-gear (and vice versa), the impact of the one culture upon the other was, at any rate at first, of no great significance. The two societies were basically far too disparate for fruitful interaction, and Roman sherds or coins on the trampled floor of a German hut meant no more than did the Arretine dishes which strewed the squalid wigwams of Cunobelin at Colchester… The alluring taste for half-understood exotic things has often enough helped to bend the course of history.” This mindset of how to view the Germans stops the archaeologist from being able to create a German viewpoint on how these Roman items were being used by the Germans.

The Quadi bordering the Roman Empire maintained their identity as a separate people through the reinforcement of various aspects of their culture that they thought were representative of being German, such as traditional housing and pottery, and also in continuing traditional behavior, such as feasting and the worship of their gods through sacred groves. However, the Germans also saw the province of Pannonia, its frontier markets, and the goods from there as opportunities to improve their social status at home in southwest Slovakia. The behavior connected to the attempt to improve social status was manifested in a variety of ways. The German might participate in raids into the empire to acquire loot and a “warrior” identity, thereby collecting an array of Roman goods for image enhancement. He might engage in trade/barter, or serve in the Roman armies: either would have served as a way to acquire material wealth so that when he returned he could either purchase greater status through cattle and gift-giving or “show-off” his wealth. Also, coming back to the point made earlier, many of these Roman goods would have been acquired either through non-Roman merchants or looting, thereby divorcing the original “Roman” significance from the item. Both of these instances would have enabled Germans to provide these items with new meaning and significance within the context of their own society.

5. Okun’s view of the situation in Germania Superior further elucidates the mitigated impact of the Roman presence. In Germania superior, it appears that the Roman presence had the greatest effect in highly visible areas, such as Augst or Vindonissa, while smaller sites further away showed little change (Okun 1991, 437–438). She goes even further to say that the evidence supports the view that in many instances the locals did not adopt Roman culture but created something new, combining local, traditional habits with Roman ones (Okun 1991, 438).

Endnotes: 1. The terms “peace-chief” and “war-chief” I first introduced in footnote 20 of Chapter 3, but for purposes of clarity I will continue to use the traditional terms: king (rex) and war-leader (dux). 2. This debate about the starting dates for the complex was made clear to me through personal discussion with various archaeologists in Slovakia and is not as readily

325

Charts Chart 11.1. Totals and Percentages of Artifacts from Hut 1, Urbárske Sedliská

Slag, 7, 0.5%

Daub, 49, 3.5% Other, 3, 0.2%

Stones, 4, 0.3%

German Items, 495, 35.2%

Bones, 830, 59.0% Roman Imports, 19, 1.4%

Chart 11.2. Totals and Percentages of Artifacts from Hut 2, Urbárske Sedliská Daub, 105, 7%

Other, 12, 1%

Stones, 9, 1%

German Items, 666, 41%

Bones, 781, 48%

Roman Imports, 31, 2%

326

Appendix A: Data Collected from Shovel Test Pits

327

328

329

330

331

332

forms that have such a line close to the rim and then below, on the body of the pot, could be various types of LBK decorations. Figure B.1:1 provides an example of such a bowl from Dvorníky and dates to the Late LBK phase (Točík 1970, Tab. ix, 6).

Appendix B Catalog of Finds from the Survey Phase and Units: Non-Roman Period Artifacts

3. Ear/handle. Find Spot: STP 4-E-50 Fr. Un. H. 4 cm. Th. 2 cm. O. and I. very pale brown (10YR 7/3) and C. dark gray (10YR 2/1).

The catalog below follows a similar format as seen in the pottery and artifact sections of Chapters 8–10, where artifacts were discussed by group instead of by STP, Unit, or Feature (see Table 8.4 for abbreviations used for pottery descriptions). Only a representative sample of artifacts is presented in the appendix, with a preference for those of interest (i.e., diagnostic, highly decorated, or unusual). Artifacts from strata 1 and 2 of Units 3 and 4 are listed in this appendix that do not deal directly with the two huts, while material found around the huts clearly modern or intrusions from other periods are listed here in this appendix as well. There is no discussion of the various periods represented, only a listing of artifacts or pottery belonging to them.

A small ear/handle from an amphora or globular pot. The clay was poorly levigated, having rock inclusions (1 – 4 mm in size) along with sand, while grass or chaff had also been added, leaving the surface pot-marked and rough. These ears could be present at the top of the body where it met the neck of the pot or on the body itself, placed horizontally, as seen on a Late LBK pot from Bratislava (Fig. B.1:2). Figure B.1:3 provides an example of ear placement on a pot from Bajč, also dating to the Late LBK. 4. Knob. Inv. No. 1183. Find Spot: Area 6, Figure: 5.21. Fr. Un. D. 5.9 cm. Th. 1.1 cm. W. 45 g. O. mottled red (2.5YR 5/8) to dark reddish brown (2.5YR 3/4), I. dark reddish gray (2.5YR 3/1), and C. is reddish black (2.5YR 2.5/1).

B.1. Pottery B.1.1. Early Neolithic: LBK

Knob from a closed vessel such as an amphora or globular pot. The knob was pressed in so that the center is concave. The sherd is rough to the feel and the texture is hackly. Inclusions include mica and rocks, while grass or chaff had been present as well, with the rocks ranging up to 4 mm in size.

The LBK pottery from Urbárske Sedliská is mostly body sherds in a very fragmentary state. A few rims were found and some decorated pieces, but all within the standard repertoire of LBK form and design. The most interesting pieces came from Unit 7, which also yielded the largest pieces.1 It would seem the LBK component at Urbárske Sedliská is largely of the early phase, based on preliminary identification done with the help of Dr. Farkaš at AM-SNM. In regards to the decorative elements present, only two pieces had an actual line decoration, while some knobs were also found. Basic identification of LBK pottery was usually done by the fabric itself, because of the fragmentary state of finds and the paucity of decorated sherds.

5. Knob. Inv. No.: 1184, Find Spot: Area 6, Figure 5.21. Fr. Un. D. 5.7 cm. Th. 6.5 mm. W. 65 g. O. mottled dark gray (10YR 4/1) to pale brown (10YR 6/3), I. light brown (7.5YR 6/3), and C. black (10YR 2/1). Another knob from a closed vessel such as an amphora or globular pot. The knob is also pressed in but in a better defined fashion, creating a more pronounced hollow center. The sherd is rough to the feel and fractures in a hackly fashion. Inclusions include mica and sand, while grass or chaff had been added as well. The clay was better levigated having no rocks, while the mica and sand were very small.

1. Knob. Find Spot: STP 4-B-40 Fr. Un. D. 2 cm. P.H. 1 cm. O. and I. very pale brown (10YR 8/3), C. very dark gray (10YR 3/1). Knob from a closed vessel. The fabric has sand and rock inclusions. Firing temperatures were not constant, with the sherd having a very dark gray core.

6. Handle. Inv. No.: 1120, Find Spot: Unit 2, stratum 1. Fr. Un. O. and I. mottled light yellowish brown (10YR 6/4) to gray (10YR 5/1).

2. Rim. Find Spot: STP 3-E-25 Fr. Un. O., I., and C. dark gray (10YR 4/1).

Handle from most likely a closed vessel for carrying liquids. The clay was poorly levigated, having a rough surface with holes indicating where straw or grass may have been, sand is also present in the clay. Texture is hackly. The handle is seen on bottle shaped vessels common in Early LBK, having three horizontal bands of upturned handles lined vertically, with holes in them for

Rim from a thin walled bowl with one incised line decoration present on the outer surface. The clay was better levigated, having only fine sand inclusions, than the previous sherd. Firing temperatures were uniform with the sherd being one color throughout. The band decoration is close to the rim, which is seen on bowl 333

rope to be threaded through for ease of transporting (Točík 1970, 28).

(up to 1 mm in size) inclusions, with added grass or chaff.

7. Knob. Inv. No.: 1125, Find Spot: Unit 2, stratum 1. Fr. Un. O. and I. light yellowish brown (10YR 6/4) and C. black (10YR 2/1).

12. Body sherds. Inv. No.: 1009, 1014 and 1210, Find spot: Unit 7, strata 2 and 3, Figure: B.3. Fr. Un. P.L. 10.6 cm. W. 160 g. O. mottled weak red (2.5YR 4/2) to dark reddish gray (2.5YR 4/1), I. reddish brown (5YR 5/3), and C. black (5YR 2.5/1).

Knob from a vessel, most likely closed for carrying liquids. The clay has sand and rock inclusions. Surface is rough to the feel and the texture is hackly.

Body sherds with preserved incised line decoration. The clay has sand, mica, and rock/quartz inclusions, with grass or chaff having been added. The inclusions range in size from 1/10 to 4 mm in size. The incised line is 5.3 cm long, 2 mm wide, and approximately 1 mm deep.

8. Body sherd. Inv. No.: 861, Find Spot: Unit 4, stratum 2. Fr. Un. D. 1 cm. P.L. 6.5 cm. P.H. 1 cm. Th. 5 mm. O. and C. brown (10YR 5/3) and I. very pale brown (10YR 7/3).

13. Handle and portion of vessel wall. Inv. No.: 1015, 1016 and 1211, Find spot: Unit 7, strata 2 and 3, Figure: B.4. Fr. Un. Vessel wall: P.H. 8.8 cm. Th. 1.5 cm. Handle: 3.3 cm wide, Th. 1.75 cm. W. 115 g. O. and I. red (2.5YR 5/6) and C. very dark gray (5YR 3/1).

Body sherd with knob and fracture along incised line decoration. Fractures are laminated, possibly indicating the clay was over-fired or used as a cooking pot. The position of the incised line decorations, where two lines meet just below a protrusion can be seen in a small pot from Milanovce, providing an idea of how the pot for sherd 861 may have looked (Fig. B.1:4).

Three sherds making up a handle from a vessel, with lab conservation. The clay has sand, mica, quartz and rock inclusions, along with grass or chaff having been added. Inclusions range in size from 1/10 – 2 mm. Handles like this could appear near the rim on small pots, where the body and neck meet (see Fig. B.1:3) or on the bellies of globular pots.

9. Knob and portion of vessel wall. Inv. No.: 989, Find Spot: Unit 5, stratum 2. Fr. Un. Vessel wall: D. 5.6 cm. Th. 9 mm.; Knob: D. 2.2 cm. W. 30 g. O. and I. mottled reddish brown (5YR 5/4) to red (2.5YR 5/6), C. black (5YR 2.5/1). Knob from a vessel, most likely closed for carrying liquids. The fabric is rough to the feel and texture is hackly. Inclusions include sand and mica, while scarring is indicative of grass or chaff having been added as well.

14. Knob. Inv. No.: 1212, Find spot: 7/STP 5-F-20, stratum 2. Fr. Un. D. 6.2 cm. Th. 9 mm. W. 50 g. O. reddish yellow (5YR 6/6), I. red (2.5YR 5/6), I.M. reddish yellow (5YR 6/6), and C. black (5YR 2.5/1).

10. Knob. Inv. No.: 1008, Find spot: Unit 7, stratum 2. Fr. Un. D. 2.3 cm. P.L. 2.6 cm. W. 35 g. O. reddish yellow (7.5YR 6/6) and C. black (10YR 2/1).

Knob with center depression. The clay has sand, quartz, mica and rock inclusions, while grass or chaff had been added as well. Inclusion sizes range from 1/16–3mm.

Knob with three lines cutting through vertically. The interior surface color is unknown because the knob is broken off in this area. There are sand and mica inclusions along with impressions of grass or chaff having been added.

B.1.2. Middle to Late Neolithic: Lengyel 15. Body sherd. Find Spot: STP 3-A-30. Fr. Un. P.L. 8.7 cm. O., I., and C. very dark gray (10YR 3/1).

11. Body sherds. Inv. No.: 799, 1001 and 1209, Find spot: Unit 7, stratum 3, Figure: B.2. Fr. Globular amphora. B.D. 18 cm. Base Th. 1.7 cm. P.H. 17 cm. Th. 1.1 cm.

Large body sherd to unidentifiable vessel type. The clay has mica and sand inclusions, with the sand in large amounts and size ranging up to 3 mm. 16. Rim. Find Spot: STP 3-D-25. Fr. Un. D. 12 cm. P.L. 3 cm. Th. 5 mm. W. 2 g. O. brown (7.5YR 5/3), I. and C. black (7.5YR 2.5/1).

Knob: P.L. 4 cm. 3.2 cm wide. O. reddish yellow (7.5YR 6/6), I. yellowish red (5YR 5/6), and C. black (5YR 2.5/1). There are three sherds making up approximately 15% of an entire pot, most likely a globular amphora. A portion of the base is preserved along with one knob, which is pressed in at the center to allow rope to be thread through. The surface has a smooth or soapy feel, while texture appears to be hackly. There are sand and quartz

Rim piece from a small thin-walled vessel, possibly a globular pot. The surface is rough to the touch and the clay has sand and mica inclusions.

334

17. Knob. Inv. No. 1129, Find spot: Unit 2, stratum 1. Fr. Un. D. 1 cm. P.H. 5 mm. O., I., and C. very dark gray (10YR 3/1). Small knob off a vessel. There are sand and mica inclusions.

B.1.3. Late Neolithic to Early Eneolithic: Bajč-Retz

18. Base. Inv. No. 1150, Find spot: Unit 2, stratum 2. Fr. Un. B.D. 12 cm. P.H. 6.5 cm. Th. 7.5 mm. W. 50 g. O. light brown (7.5YR 6/4), I. brown (7.5YR 5/2), C. gray (10YR 5/1).

Body sherd with the distinctive decorative technique of Furchenstich (Farkaš 1990, 123). There are fine sand, mica, and quartz inclusions, with sand abundant and the other two uncommon.

Fragment of the base to a vessel itself or a fragment of the foot of a vessel. The surface is rough and the clay has sand, mica and quartz inclusions, ranging in size up to 3 mm.

2. Body sherd. Find Spot: 3-C-25. Fr. Un. P.L. 1.9 cm. Th. 3.9 mm. W. 1 g. O. brown (7.5YR 5/3), I. brown (7.5YR 5/3), and C. black (7.5YR 2.5/1).

19. Knob. Find spot: Unit 3, stratum 1. Small knob from the body of a vessel. O. and I. reddish yellow (5YR 6/6), C. black (5YR 2.5/1). Sand is the primary inclusion.

Body sherd with the Furchenstich technique of decoration. This sherd has mica for inclusions.

1. Body sherd. Find spot: STP 3-C-20. Figure: B.6. Fr. Un. P.L. 3.6 cm. Th. 7 mm. W. 2 g. O. and I. light brown (7.5YR 6/3) and C. black (7.5YR 2.5/1).

3. Body sherds. Inv. No. 1117, Find spot: Unit 2, stratum 1. Fr. Un. P.L. 4.5 cm. Th. 4 mm. O., I., and C. brown (7.5YR 5/3).

20. Knob. Find spot: Unit 3, stratum 1. Small knob from the body of a vessel. O. and I. mottled dark reddish brown (5YR 3/2) to reddish brown (5YR 4/4), C. black (5YR 2.5/1). The inclusions include sand and mica.

Two decorated sherds off the body, where there is a marked curvature.

21. Knob. Inv. No.: 297, Find spot: Unit 3, stratum 2. Fr. Un. D. 4.5 cm. P.H. 1.5 cm. O. and I. pale brown (10YR 6/3) and C. dark gray (10YR 4/1).

4. Body sherds. Inv. No. 1164 and 1165, Find spot: Unit 2, stratum 2. Fr. Un. P.L. 6 cm. Th. 6.5 mm. W. 7 g. O. and I. dark gray (7.5YR 4/10) and C. black 7.5YR 2.5/1).

Large knob off a vessel. The clay has sand and mica inclusions.

Two decorated sherds off the body of a vessel. There are sand and mica inclusions in the clay.

22. Body sherd and knob. Find spot: Unit 3, stratum 2. Fr. Un. Sherd: P.L. 7.7 cm. Th. 1 cm. Knob: D. 1.8 cm. L. 1.2 cm. O. reddish yellow (5YR 7/6), I. and C. light brown (7.5YR 6/4).

5. Body sherd. Find spot: Unit 3, stratum 1, Figure: B.7. Fr. Un. P.L. 3.8 cm. Th. 3 mm. O. mottled brown (7.5YR 5/3) to dark brown (7.5YR 3/2), I. and C. black (7.5YR 2.5/1).

Body sherd with knob. The curvature of the piece indicates the piece came from the belly of the pot at the apex of the curve, where it is common to find knobs. Two examples can be seen in figure B.5. The clay has a high number of inclusions, mostly sand and mica, with some quartz. Inclusions range in size up to 3 mm.

Body fragment with very ornate Furchenstich decoration, which has white encrustation preserved (Farkaš 1990, 123). There are sand and mica inclusions. There is a distinct curvature to the sherd indicating it most likely came from the body of the vessel.

23. Knob. Inv. No.: 301, Find spot: Unit 3, stratum 2. Fr. Un. P.H. 1 cm. D. 1.5 cm tapers to 1 cm. O. and I. light red (2.5YR 6/6) and C. dark reddish gray (2.5YR 4/1).

6. Body sherd. Inv. No. 982, Find spot: Unit 5, stratum 1, Figure: B.8. Fr. Un. P.L. 2.8 cm. Th. 5 mm. W. 2 g. O. and I. dark reddish gray (5YR 4/2) and C. black (5YR 2.5/1).

Small knob with a shallow depression in the center. The clay has sand and mica inclusions.

Body fragment with Furchenstich decoration. There sand and mica inclusions in the clay.

24. Knob. Find spot: Area 6. Fr. Un. D. 3 cm. P.H. 2.5 cm. O., I., and C. light red (2.5YR 6/6).

B.1.4. Middle Eneolithic to the Final Late Bronze Age As seen in the survey, the upper strata of the units did not yield much material from the Late Eneolithic to the Final Late Bronze Age (3800–800 BC). The absence of these later cultures is either caused by the fragmentary nature

Large knob for the body of a vessel. There is a high amount of sand and quartz inclusions, ranging up to 3 mm in size.

335

of the pottery and therefore these groups cannot be identified or later groups preferred different areas for settlement. The two sherds presented below have been identified as belonging to the Bronze Age (for Slovakia, ca. 2300–800 BC; Novotná 1993, 82; Furmánek et. al. 1991, 127 and 359).

B.2. Metal The types of metal finds uncovered in the units mirrored those found in the STPs, in that there was a large group of heavily corroded modern items, such as hoe blades, screws, nails, and machine parts, mixed with items that could be older, but were so corroded they could not be identified. Fifty-one metal items were recovered from the units and when added to the STP number a total of seventy-seven pieces were found. Table B.1 provides a summary of what could be identified in broad terms for items recovered from all units and STPs. Nails/Spikes were the largest group after unknown, which were pieces so corroded their original function could not be determined. The presence of nails and spikes is not surprising because of the presence of the house nearby along with a chain-link fence, using cement posts. Two items of interest are presented below.

7. Rim. Find spot: STP 3-D-45. Fr. Un. D. 23 cm. P.L. 4.8 cm. Th. 7.9 mm. W. 15 g. O. mottled very dark brown (10YR 2/2) to dark brown (10YR 3/3). Rim piece from a closed vessel, with the lip being turned outward at a 45 degree angle. The clay has sand, mica and quartz inclusions, ranging up to 1 mm in size. 8. Body sherds. Find spot: 5-F-45. Fr. Un. P.L. 11 cm. Th. up to 3 cm. O. and I. brownish yellow (10YR 7/6) and C. very pale brown (10YR 8/3).

Table B.1. Metal Items from Units and STPs Type Count Nails/Spikes 21 Screws 3 Tools 5 Machine 3 Parts Pins 1 Other 3 Unknown 41 Total: 77

Three sherds that go together to from a portion of the body from a closed vessel (broken during excavation). The clay has an abundance of sand along with some quartz and mica, ranging up to 3 mm in size. B.1.5. Great Migration Period – Present Day As seen with the STPs, so too with the units: there is no pottery that could be identified of the period from AD 375 to 1000. A few sherds could be dated to the Middle Ages, after which there is again a brief lull before the Modern Period at which point there is a steady increase in pottery. When the data are combined from the units with the counts from the STPs, then a total of 13 sherds could be identified belonging to the Middle Ages, one to the Renaissance, and 230 sherds for the Modern Period. It would seem that after the German occupation in the early second century, the site does not see cultural activity of a measurable nature until the present.

9. Bird-Caller. Find spot: STP 2-1-55. Figures: B.9 and B.10. Home-made bird caller used by a hunter.2 There are two parts to the caller: the metal mouthpiece (8 cm long with a 3 cm diameter at the wide end) and a wooden construction inserted at the end that creates the bird noise (3 cm long with a 1 cm diameter at its widest).

Among the Middle Age sherds at Urbárske Sedliská were five rims, while among the modern sherds from the STPs there were eight rims, and from the units there were twenty rims and three bases. When looking at the modern pottery as a group, glazing was a common form of surface treatment, with seventy sherds from STPs and Units being glazed. There were also twelve sherds of porcelain, of which there were three rims and two bases. A majority of the porcelain was without decoration, with only three sherds having some type of decoration. One sherd was from the bottom of a vessel and had a portion of the manufacturer’s mark. The second sherd of porcelain had a traditional blue glaze decoration, while a large rim fragment of a small vase was more recent, having a transfer pattern on the body. Much of the modern pottery probably came from the house that once stood nearby to the northeast of where excavations took place.

The metal portion is bronze and appears originally to have been the mouthpiece of a brass instrument, possibly a cornet, euphonium, or even a soprano saxhorn. The piece is funnel shaped, with a wide rim, and examples are seen paired with a bass flugelhorn (ca. 1848) or soprano cornet (ca. 1843) (Heyde 1987, 121 and 123).3 The preservation of the bronze is very good with a fine patina. There are no manufacturer’s marks, numbers, or symbols on the mouth-piece. A similar mouthpiece in the Edinburgh University collection of historic musical instruments is identified as one for a soprano saxhorn possibly made in Gautrot, France, in the late nineteenth century (Edinburgh 2004). The wooden piece is also well preserved. It was made up of a wooden dowel (called the stopper or insert) cut in half and hollowed in the center, so that a second thinner filament (technically called the reed) could sit inside (Luckey and Lewis 2003, 236). When a person blew into the caller the air would cause the reed to vibrate inside 336

the insert, mimicking the call of a bird, possibly duck or goose.

come from the LBK culture or later Celtic period, being hand made). The STP had been dug to a depth of 90 cm, with the artifacts coming from a depth of 55 cm, which would put them below the plough zone. The combination of depth and being found amidst ancient pottery would indicate the stone is ancient. The shape and size indicate that the stone was most likely a bullet for a sling. According to M. Korfmann, easily worked stones would have been used, such as limestone, and would be eggshaped or ovoid, while in size, sling missiles could be up to five cm in diameter (Korfmann 1973, 38–39). Even though, by the Classical Period, slings were usually made of clay or lead, stone missles were still used, such as those used by the famous slingers from the Balearic islands (Korfmann 1973, 39). Clay missiles date back at least 7,000 years, so just by the material used, one cannot assign a date to the stone missile under discussion here (Korfmann 1973, 39). It can be said, however, that the sling was in use by the Neolithic period in the Balkans and southeastern Europe and so would have made an appearance also in Central Europe (Korfmann 1973, 42). The stone therefore cannot be precisely dated, possibly belonging to any culture from the Neolithic to the Roman Period.

The age of the duck-caller still needs to be determined, but based on the design of the mouthpiece (as mentioned above) and the insert, it may date to the last half of the 19th century. The caller may in fact be contemporary with the two coins found on the site (dating to 1885/6), but this is conjecture. The dating of the insert is based on its design, which mimics store-bought callers of the late nineteenth century (at least in comparison to American examples). In that period, there were some designs with a metal mouthpiece vaguely similar to a brass instrument’s mouthpiece that then had inserted into it the stopper with reed (Luckey and Lewis 2003, 241 and 244). One wonders about the caller’s effectiveness, since the reed appears now to be free-floating. The manufactured callers used an insert with a sounding board on which the reed sat, wedged at one end to make the other end free to vibrate. The insert on the caller found onsite has no clear sounding board or wedge, thereby probably making it very hard to make the reed vibrate. It is not surprising to have found hunters’ paraphernalia, since even today, the area still teems with ducks, quail, deer, and rabbit, making Urbárske Sedliská a popular hunting area for the locals (a hunting station was located 500 meters to the north of the project area and the named wildlife was seen by the crew during the project).

40. Whetstone (?). Find spot: Unit 3, stratum 2. Figure: B.13.

10. Pin. Find spot: STP 3-B-20, Figure: B.11. Medieval pin used to hold the wheel in place on an axle.4 The pin has a preserved length of 8.5 cm, with the head being 3 cm wide and the diameter of the pin being on average 1 cm. The material is bronze and the pin was not machine-made, but hammered.

A ground stone with a cylindrical shape, broken off at each end. The width at the wider end is 2.4 cm, which tapers to a width of 2 cm. The length of the stone is 3.7 cm. It appears too small to have been an axe or adze, but may have been simply a whetstone. With both ends broken off, there is no way of seeing percussion hits to indicate the stone’s use as a hammerstone. See chapter 10.2.5 for descriptions of similar stones found in Hut 2, particularly stones 13–15 (Figures 10.15:13, 10.16:13, 10.14, 10.15:15, and 10.16:15).

B.3. Rocks/Stones

B.4. Coins

Twenty-five rocks were found during excavations of the units. Natural rocks were also found during the survey, but only a few were kept. Two stones showed sign of possibly being worked.

As mentioned in Chapter 5 under the heading ‘Other’ one coin was found. Another coin was found in Unit 3 next to Hut 1. Both are copper, dating to the time of the AustroHungarian Empire.

39. Sling stone (?). Find spot: STP 5-H-55, Figure: B.12.

41. Inv. No.: 946, Find spot: STP 4-B-30. The coin is an Austrian Kreuzer, minted in 1885, during the rule of Franz Joseph (Krause and Mishler 1975, 54). The obverse has the double headed eagle of the empire, while the reverse has the denomination of 1 with the year underneath, encircled by laurel leaves. The coin is in good condition.

The stone is egg shaped having the long axis at 5 cm and the shorter axis at 4 cm, while the surface of the stone is relatively smooth, but still rough, making it seem unlikely to originate from a river. In regards to the nature of the stone, it appears weak and would have been inappropriate for use as a hammer-stone, but would be as a sling-stone. The stone’s size also seems to preclude use as a hammerstone. There were seven sherds found with the rock but they could only be identified as Ancient Undetermined. The sherds have no decoration and none are diagnostic, while their composition is not unique enough to provide a cultural identity (except to say that they do not appear to

42. Inv. No.: 947, Find spot: Unit 3, stratum 2. The coin is a Hungarian Krajczar, minted in 1886 (Krause and Mishler 1975, 629). The obverse has the crown and shield of the Hungarian kingdom, while the reverse has the denomination of 1 with the year 337

underneath, encircled by laurel leaves. Below the year is the mint mark KB, indicating the town of Kremnica, which is now in the Slovak Republic. The coin is in good condition. Endnotes 1. Because of these large sherds, the area around Unit 7 should be explored further to see if there is any cultural feature present dating to the Neolithic. 2. Identification of this artifact as a bird caller was by Mr. Slovak (curator of metals at AM-SNM). 3. Mouthpieces are of course removable and so it becomes difficult to date them since they could have been reused on various instruments. The instruments mentioned in the text are not for dating purposes, but more to identify the brass instrument most suitable for the mouthpiece. 4. Artifact identified by Dr. Turčan and Mr. Slovak, who also cleaned the piece.

338

Figures

Figure B.1. Various LBK vessels (Not to scale). 1. Early LBK bowl from Dvorníky, 2. , 3. LBK vessel from Bajč, 4. Early LBK vessel from Milanovce (Točík 1970, Tab. IX:6, VIII:1, X respectively).

339

Figure B.2. LBK base, body, and handle.

Figure B.3. LBK body sherds with incised line decoration at bottom.

340

Figure B.4. LBK handle (conserved in lab with some plaster addition).

Figure B.5. Lengyel vessels (early phase) from Nitriansky Hrádok-Zámeček (Točík 1970, Tab. xxxviii).

341

Figure B.6. Bajč-Retz sherd.

Figure B.7. Bajč-Retz sherd.

Figure B.8. Bajč-Retz sherd.

342

Figure B.9. Bird Caller.

Figure B.10. Bird Caller with the dowel removed and opened to show filament.

343

Figure B.11. Metal pin (shown after cleaning in the lab).

Figure B.12. Sling-stone.

Figure B.13. Whetstone. 344

List of Journal Abbreviations AA Acta Arch. Acad. Scient. Hungaricae AJP Arch. Ért Archeol. J. Arch. Rozhledy AVANS Carn. Jahrb. Zbor. SNM Hist. Stud. Comit Štud. Zvesti AÚ SAV TAPA Zbor. SNM Arch. Diss. Pann. JFA JRA JRS OCD Pam. Arch. Pam. Muz. RLÖ Saal. Jahrb. Slov. Arch. Slov. Num. Zeitschr. Ethn.

American Anthropologist Acta Archeologica Academiae Scientarium Hungaricae The American Journal of Philology Archaeologiae Értesítő The Archaeological Journal Archeologické rozhledy Archeologické výskumy a nálezy na Slovensku Carnuntum Jahrbuch Zborník Slovenského národného múzea, História Studia Comitatensia Študijné zvesti Archeologického ústavu Slovenskej akadémie vied. Transactions of the American PhilologicalAssociation Zborník Slovenského národného múzea, Archeológia Dissertationes Pannonicae Journal of Field Archaeology Journal of Roman Archaeology Journal of Roman Studies Oxford Classical Dictionary (3rd edition) Památky archeologické Pamiatky a múzeá Der Römische Limes in Österreich Saalburg Jahrbuch Slovenská archeológia Slovenská Numizmatika Zeitschrift für Ethnologie

345

Bibliography Primary Sources Appian. Roman History. Loeb edition. Translated by H. White. Cambridge: Harvard University Press. 1955, 1982, 1990. Julius Caesar. The Gallic War. Loeb edition. Translated by H. J. Edwards. Cambridge: Harvard University Press. 1963. Cicero. De re publica; De legibus. Loeb edition. Translated by C. W. Keyes. Cambridge: Havard University Press. 1928. Dio Cassius. Roman History. Loeb edition. Translated by E. Cary. Cambridge: Harvard University Press. 1954, 1955, 1961, 1968, 1982. Eutropius. Breviarium. Translated by H. Bird. Liverpool: Liverpool University Press.1993. Rufus Festus. The Brevarium of Festus; a critical edition with historical commentary. J. W. Eadie. University of London Classical Studies, 5. London: Athlone Press. 1967. Florus. Epitome of Roman History. Loeb edition. Translated by E. Forster. Cambridge: Harvard University Press. 1995. Martial. Epigrams. Loeb edition. Translated by W. Ker. Cambridge: Harvard University Press. 1968. Ovid. Fasti. Loeb edition. Translated by Sir J. G. Frazer. London: W. Heinemann Ltd. 1931. Pliny the Elder. Natural History. Loeb edition. Translated by D. E. Eichholz and H. Rachkham. Cambridge: Harvard University Press. 1989. Claudius Ptolemy. Geographia. Edited by C. Nobbe. Books I–III. Hildesheim: Georg Olms Verlagsbuchhandlung. 1966. ⎯⎯ The Geography. Translated by E. Stevenson. Reprint. New York: Dover Publications, Inc. 1991. ⎯⎯ Ptolemy’s Geography. Books 1, 2, 7, and 8. Translated by J. Berggren and A. Jones. Princeton: Princeton University Press. 2000. Silius Italicus. Punica. Loeb edition. Translated by J. D. Duff. Cambridge: Harvard University Press. 1983, 1989. Statius. Silvae. Loeb edition. Translated by D. R. S. Bailey. Cambridge: Harvard University Press. 2003. Res Gestae Divi Augusti. Loeb edition with Velleius Paterculus. Translated by F. Shipley. Cambridge: Harvard University Press. 1998. Scriptores Historiae Augustae. Loeb edition. Translated by D. Magie. Harvard University Press. 1991. Strabo. The Geography. Loeb edition. Translated by H. Jones. Cambridge: Harvard University Press. 1983, 1988, 1989. Suetonius. De Vita Caesarum. Loeb edition. Translated by J. C. Rolfe. Cambridge: Harvard University Press. 1989. Tacitus. Agricola, Germania, Dialogus. Loeb edition. Translated by M. Hutton and W. Peterson. Cambridge: Harvard University Press. 1992. ⎯⎯ Germany. Translated by H. Benario. Warinster: Aris & Phillips LTD. 1999. ⎯⎯ The Histories, The Annals. Loeb edition. Translated by C. Moore and J. Jackson. Cambridge: Harvard University Press. 1962, 1963. Aurelius Victor. De Caesaribus. Translated by H. Bird. Liverpool: Liverpool University Press. 1994. Velleius Paterculus. Compendium of Roman History. Loeb Edition with Res Gestae Divi Augusti. Translated by F. Shipley. Cambridge: Harvard University Press. 1998. Vergil. Vergil’s Aeneid: Books I–VI. Revised Edition. Ed. Clyde Pharr. Lexington: D. C. Heath and Company. 1964. ⎯⎯ Aeneid. Translated by L. Lind. Bloomington: Indiana University Press. 1962. Secondary Sources Almgren, O. 1897. Studien über Nordeuropäische Fibelformen. Stockholm: Druck von Ivar Hæggström. Anderson-Stojanović, V. 1992. Stobi: The Hellenistic and Roman Pottery. Princeton: Princeton University Press. Audouze, F., and O. Büchsenschütz. 1992. Towns, Villages, and Countryside of Celtic Europe. London: B.T. Batsford Ltd. Bachmann, H. 1982. The Identification of Slags from Archaeological Sites. Occasional Publication No. 6. London: Institute of Archaeology. Barrett, J. C. 1997. “Romanization: a critical comment.” In Dialogues in Roman Imperialism, ed. D. J. Mattingly, 51– 66. JRA Supplementary Series, 23. Barrett, J. C., A. P. Fitzpatrick, and L. Macinnes. 1989. Barbarians and Romans in North-West Europe. Oxford: BAR International Series 417. Barth, F. 1998. Ethnic Groups and Boundaries. Reissue of the 1969 edition. Long Grove: Waveland Press Inc. Bartík, J., and Turčan, V. 1990. “Germánska Chata v Báhoni.” Zbor. SNM 84, Hist. 30: 73–83. Bartík, J., and V. Varsik. 1994. “Výnimočný import kameového skla z Abrahámu.” Zbor. SNM 88, Arch. 4: 63–74. Beckmann, B. 1966. “Studien über die Metallnadeln der römischen Kaiserzeit im freien Germanien.” Saal. Jahrb. 23: 5–100. Benadik, B. 1971. “Obraz Doby Laténskej na Slovensku.” Slov. Arch. 19: 465–491. 346

Benario, H. 1975. An Introduction to Tacitus. Athens: University of Georgia Press. ⎯⎯ 1999. Tacitus Germania. Warminster: Aris & Phillips LTD. Beninger, E. 1937. Die germanischen Bodenfunde in der Slowakei. Reichenberg-Leipzig. Bentley, G. 1987. “Ethnicity and practice.” In Comparative Studies in Society and History 29: 24–55. Berggren, J., and A. Jones. 2000. Ptolemy’s Geography. Princeton: Princeton University Press. Berlekamp, H. 1988. “Rhein-Weser-Germanen.” In Die Germanen, Band I, ed. Bruno Krüger, 408–414. Berlin: Akademie-Verlag Berlin. Birley, A. 1987. Marcus Aurelius. Revised Edition. New Haven: Yale University Press. Blagg, T., and M. Millett. 1990. The Early Roman Empire in the West. London: Oxbow Books. Börner, W. 1997. “Wien-Vindobona.” In Der Römische Limes in Österreich, eds. Herwig Friesinger and Fritz Krinzinger, 241–252. Wien: Verlag der Österreichischen Akademie der Wissenschaften Wien. Bónis, É. 1942. Die kaiserzeitliche Keramik von Pannonien. Diss. Pann. 20, ser. 2. Budapest. ⎯⎯ 1970. “A brigetioi sávos kerámia.” Folia Archaeologica 21. Budapest. ⎯⎯ 1980. “Pottery.” In The Archaeology of Roman Pannonia, eds. A. Lengyel and G. T. B. Radan, 357–380. Budapest: Akadémiai Kiadó. Bosworth, B. 1998. “Augustus, the Res Gestae and Hellenistic Theories of Apotheosis.” JRS 89: 1–18. Bouzek, J. and I. Ondřejová. 1990. “‘Třetí zóna’ Mezi Římem a Barbarikem Při Noricko-Pannonském Limitu.” Arch. Rozhledy 42: 22–35. Braund, D. 1984. Rome and the Friendly King. London: Croom Helm. Breeze, D. 1982. The Northern Frontiers of Roman Britain. New York: St. Martin’s Press. Brukner, O. 1980. Rimska Keramika U Jugoslovenskom Delu Provincije Donje Panonije. Dissertationes et Monographiae, Tome 24. Beograd: Pokrajinksi Zavod za Zaštitu Spomenika Kulture SAP Vojvodine Savez Arheoloških Društava Jugosalvije. Bryce, J. 1914. The Ancient Roman Empire and The British Empire in India/The Diffusion of Roman and English Law throughout the World. London: Oxford University Press. Bucher, G. 2000. “The Origins, Program, and Composition of Appian’s Roman History.” TAPA 130: 411–458. Bujna, J. 1976. “Revízny výskum Rímskej stanice v Stupave (okr. Bratislava-Vidiek) v roku 1974.” Arch. Rozhledy 28: 494–549. Burns, T. 2003. Rome and the Barbarians, 100 B.C.– A.D. 400. Baltimore: The Johns Hopkins University Press. Cahnman, W. 1949. “Frontiers between East and West in Europe.” Geographical Review 39 (no. 4, October): 605–624. Carniero, R. 1970. “A theory of the origin of the state.” Science 169: 733–738. ⎯⎯ 1981. “The chiefdoms as precursor of the state.” In The Transition to Statehood in the New World, eds. G. Jones and R. Krautz, 37–79. Cambridge: Cambridge University Press. Cary, M. 1965. A History of Rome. 2nd edition. London: Macmillan & Co. LTD. Chamberlain, M. 1972. “Lord Cromer's 'Ancient and Modern Imperialism': A Proconsular View of Empire,” in The Journal of British Studies 12, no. 1: 61–85. Chropovský, B., ed. 1981. Archaeological Research in Slovakia. Nitra: AÚ-SAV. Churchill, Sir W. 1933. The river war: an account of the reconquest of the Soudan. (first publ. 1899). London. Clarke, K. 1997. “In Search of the Author of Strabo’s Geography.” JRS 87: 92–110. Cleary A. S. E. 1990. The Ending of Roman Britain. Savage: Barnes and Noble Books. Cohen, R. 1978. “Ethnicity: problem and focus in anthropology.” Annual Review of Anthropology 7: 379–403. Collingwood, R. and R. Wright. 1992. The Roman Inscriptions of Britain. Vol. 2, fascicule 4. Wolfeforo Falls: Alan Sutton Publishing Inc. Collins, J., and B. Molyneaux. 2003. Archaeological Survey. Walnut Creek: Altamira Press. Creighton, J. D. and R. J. A. Wilson. 1999. Roman Germany. JRA Supplementary Series 32. Portsmouth. Cromer. 1910. Ancient and Modern Imperialism. London: John Murray. Cunliffe, B. 1988. Greeks, Romans and Barbarians. New York: Methuen. ⎯⎯ 1994. “The Impact of Rome on Barbarian Society, 140 BC – AD 300.” In The Oxford Illutrated History of Prehistoric Europe, ed. Barry Cunliffe, 411–446. Oxford: Oxford University Press. Dabas, M., and A. Tabbagh. 2000. “Magnetic Prospecting.” In Archaeological Method and Theory: An Encyclopedia, ed. Linda Ellis, 335–339. New York: Garland Publishing Inc. Diáz-Andreu, M. 1996. “Constructing identities through culture: the past in the forging of Europe.” In Cultural Identity and Archaeology, eds. Paul Graves-Brown, Sian Jones, and Clive Gamble, 48–61. Routledge: London. Dobiáš, J. 1964. Dějiny ČeskoSlovenského Území Před Vystoupením Slovanů. Praha: Československé Akademie Věd. Drexel, F. 1911. Das Kastell Faimingen. Heidelberg. ⎯⎯ 1918. “Die germanischen Hütten auf der Markussäule.” Germania 11: 114–118. Droberjar, E. 1995 “Zur Frage der ältesten germanischen und römischen Siedlungsfunde in Mähren.” In Kelten, Germanen, Römer im Mitteldonaugebiet, eds. Karol Pieta, Ján Rajtár, and Jaroslav Tejral, 21–38. Brno: Archeologický ÚSTAV Akademie Věd České Republiky. ⎯⎯ 1997. Studien zu den Germanischen Siedlungen. Prague: Nationalmuseum Prag. ⎯⎯ 2002. Encyklopedie Římské a Germánské Archeologie v Čechách a na Moravě. Praha: Libri. 347

Drummond, S. and L. Nelson. 1994. The Western Frontiers of Imperial Rome. London: M. E. Sharpe. Dyson, S. 1985. Creation of the Roman Frontier. Princeton: Princeton University Press. Earle, T. 1987. “Chiefdoms in Archaeological and Ethnohistorical Perspective.” AnnualReview of Anthropology 16: 279–308. Edinburgh University Visited March 20, 2005. Eggers, H. 1951. Der römische Import im freien Germanien. Bds. 1 and 2. Hamburg. Eisner, J. 1933. Slovensko v Pravěku. Bratislava: Učené Společnosti Šafaříkovy. Elschek, K. 1995. “Die germanische Besiedlung von Bratislava-Dúbravka während der älteren römischen Kaiserzeit.“ In Kelten, Germanen, Römer im Mitteldonaugebiet, eds. Karol Pieta, Ján Rajtár and Jaroslav Tejral, 39–52. Brno: Archäologisches Institut der Akademie der Wissenschaften der Tschechischen Republik Brno. ⎯⎯ 2000. “Rímsko-germánska vidiecka usadlosť s kúpeľom v Bratislave-Dúbravke.” Pam. Múz. 3: 27–29. ⎯⎯ 2005 “Bratislava-Devínska Nová Ves, objekt 3/1996 a Láb (okr. Malacky), objekty 1–2/1997.” Zbor. SNM 99, Arch. 15: 255–274. Elton, H. 1996. Frontiers of the Roman Empire. Bloomington: Indiana University Press. Fagan, B. 1995. Ancient North America. Revised edition. London: Thames and Hudson. Farkaš, Z. 1984. “Nálezová Správa z výskumu v Stupave, okr. Bratislava vidiek v polohe Urbárske Sedliská, časť pod dvorom Štátnych majetkov za r. 1982.” Unpublished field-notes from the excavation, stored at AM-SNM, Bratislava. ⎯⎯ 1990 Ich Ohniská Kryje Zem. Bratislava: Obzor. Farkaš, Z., and B. Novotný. 1993 “Mladšia a neskorá doba kamená.” In Najstaršie Dejiny Bratislavy, ed. Tatiana Štefanovičová, 39–79. Bratislava: ELÁN. Fassbinder, J. W. E., and W. E. Irlinger. 2000. “Combining Magnetometry and Archaeological Interpretation: A Square Enclosure in Bavaria.” In Magnetic Prospecting in Archaeological Sites, eds. Helmut Becker and Jorg W. W. Fassbinder, 14–19. Monuments and Sites 4. Munchen: Lipp GmbH. Ferguson, R., and N. Whitehead, eds. 1992. War in the Tribal Zone: Expanding States and Indigenous Warfare. Sante Fe: School of American Research. Fischer, T. 1983/84. “Ein Keller mit Brandschutt aus der Zeit der Markomannenkriege 170/175 aus dem Lagerdorf des Kastells Regensburg – Kumpfmühl.“ Bericht der Bayerischen Bodendenkmalpflege 24/25: 24–64. Fitz, J. 1980. “Administration and army.” In The Archaeology of Roman Pannonia. Eds. A. Lengyel and G. T. B, 125– 140. Radan. Budapest: Akadémiai Kiadó. Fitzpatrick, A. P. 1996. “‘Celtic’ Iron Age Europe: the theoretical basis.” In Cultural Identity and Archaeology, eds. Paul Graves-Brown, Sian Jones, and Clive Gamble, 238–257. London: Routledge. Franke, A. 1980. Rom und die Germanen. Tübingen: Grabert-Verlag. Freeman, P. 1996. “British Imperialism and the Roman Empire.” In Roman Imperialism: Post-Colonial Perspective, eds. J. Webster and N. Cooper. Leicester Archaeology Monographs No. 3: 19–34. Leicester: University of Leicester. ⎯⎯ 1997. “Mommsen through to Haverfield: the origins of Romanization studies in late 19th-c. Britain.” In Dialogues in Roman Imperialism, ed. D. J. Mattingly, 27–50. JRA Supplementary Series 23. Frey, O.-H. 1995. “The Celts in Italy.” In The Celtic World, ed. Miranda Green, 515–532. New York: Routledge. Friesinger, H., J. Tejral, and A. Stuppner, eds. 1994. Markomannenkriege: Ursachen und Wirkungen. Brno: Archäologisches Institut der Akademie der Wissenschaften. Friesinger, H., and A. Stuppner. 1997 “Der Oberleiserberg be Ernstbrunn.” In Der Römische Limes in Österreich, eds. Herwig Friesinger and Fritz Krinzinger, 282–287. Wien: Verlag der Österreichischen Akademie der Wissenschaften Wien. Furmánek, V. and K. Pieta. 1985 Počiatky odievania na Slovensku. Bratislava: Tatran. Furmánek, V., L. Veliačik, and J. Vladár. 1991. Slovensko v dobe bronzovej. Bratislava: Slovenskej Akadémie Vied. Gabler, D. 1978 “Die Keramik von Vindobona.” In Vindobona – die Römer in Wiener Raum, 118–136. Wien. ⎯⎯ 1980. “The structure of the Pannonian frontier on the Danube and its development in the Antonine period: some problems.” In Roman Frontier Studies 1979, eds. Hanson, W. and L. J. F. Keppie, 637–654, vol. 2. Oxford: BAR International Series, 71 (2 vols.). ⎯⎯ 1986. “Differences Between Imported Pottery in the Western and Danubian Provinces of the Roman Empire.” Acta Arch. Acad. Scient. Hungaricae 38: 93–104. ⎯⎯ 1989. The Roman Fort at Ács-Vaspuszta (Hungary) on the Danubian Limes. Vols. 1 and 2. BAR International Series 531(i and ii). Oxford: B.A.R. ⎯⎯ 1991. “The survival of late La Tène settlements in the Roman period.” In Roman Frontier Studies 1989, eds. V. Maxfield and M. J. Dobson, 424–431. Exeter: University of Exeter Press. Gassner, V. 1988. “Gelbtonige Keramik aus datierten Fundkoplexen in Carnuntum.” Carn. Jahrb.: 133–161. ⎯⎯ 1990 “Feinware aus Carnuntum.” Carn. Jahrb., 253–292. ⎯⎯ 1991 “Mittelkaiserzeitliche glasierte Kermaik aus Pannonien.” Carn. Jahrb., 9–55. ⎯⎯ 1993 “Pannonische Glantonware mit Stempelverzierung aus Carnuntum.“Ptujski arheološki Zborník, 359– 383. Ptuj. 348

Genser, K. 1990. Der Donaulimes in Österreich. Stuttgart: Limesmuseum Aalen. Giesen, J. www.jgiesen.de/planets/solutions.html. Visited April 17, 2005. Gluckman, M. 1965. Politics, Law and Ritual in Tribal Society. Oxford: Basil Blackwell. Gose, E. 1950. Gefäßtypen der römischen Keramik im Rheinland. Kevelaer. Gowing, A. 1990. “Tacitus and the Client Kings.” TAPA 120: 315–331. Graves-Brown, P., S. Jones, and C. Gamble, eds. 1996. Cultural Identity and Archaeology. London: Routledge. Green, M. 1992. Animals in Celtic Life and Myth. London: Routledge. ⎯⎯ 1995a. The Celtic World. London: Routledge. ⎯⎯ 1995b. “Introduction: who were the Celts?” In The Celtic World, ed. Miranda Green, 3–7. London: Routledge. Grünewald, M. 1979. “Die Gefäßkeramik des Legionslagers von Carnuntum (Grabungen 1968–1974).” RLÖ 29, Wien. Gudeman, A. 1900. “The Sources of the Germania of Tacitus.” TAPA, vol. 31: 93–111. Halouzka, R. 1993. “Prírodné prostredie Bratislavy.” In Najstaršie Dejiny Bratislavy, ed. Tatiana Štefanovičová, 9–27. Bratislava: ELÁN. Hanson, W. 1989. “The Nature and Function of Roman Frontiers.” In Barbarians and Romans in North-West Europe, eds. J. C. Barrett, A. P. Fitzpatrick, and L. Macinnes, 55–63. Oxford: BAR International Series 471. ⎯⎯ 1997. “Forces of change and methods of control.” In Dialogues of Roman Imperialism, ed. D. Mattingly, 67– 80. JRA Supplementary Series, no. 23. Hanson, W. and L. J. F. Keppie. 1980. Roman Frontier Studies 1979. Oxford: BAR International Series, 71 (2 vols.). Harden, D. 1987. Glass of the Caesars. Milan: Olivetti. Härke, H., ed. 2002. Archaeology, Ideology and Society. 2nd edition. Frankfurt: Peter Lang. ⎯⎯ 2002. “The German experience.” In Archaeology, Ideology and Society. 2nd edition, ed. H. Härke, 13–40. Frankfurt: Peter Lang. Haßmann, H. 2002. “Archaeology in the ‘Third Reich.’” In Archaeology, Ideology and Society. 2nd edition, ed. H. Härke, 67–142. Frankfurt: Peter Lang. Haverfield, F. 1905. “The Romanization of Roman Britain.” Proceedings of the British Academy, 2: 185–217. ⎯⎯ 1912. The Romanization of Roman Britain. London: Oxford Clarendon Press. Hayes, J. 1997. Handbook of Mediterranean Roman Pottery. Norman: University of Oklahoma Press. Hečková, J. 1976. “Druhý rok revízneho výskumu rímskej stanice v Stupave.” AVANS v Roku 1975: 104–105. Nitra: AÚ-SAV. ⎯⎯ 1986 “Römischer Baukomplex in Stupava.” Arch. Rozhledy, vol. 38, no. 4: 378–410. ⎯⎯ 1991 “Graficko-numerický prehľad keramiky z obytných objektov doby rímskej v Chotíne.” Štud. Zvesti AÚSAV 27: 129–241. Heyde, H. 1987. Das Ventil-blasinstrument. Wiesbaden: Breitkopf and Hartel. Hides, S. 1996. “The genealogy of material culture and cultural identity.” In Cultural Identity and Archaeology, eds. Paul Graves-Brown, Sian Jones, and Clive Gamble, 25–47. Routledge: London. Hines, J. 1996. “Britain after Rome: between multiculturalism and monoculturalism.” In Cultural Identity and Archaeology, eds. Paul Graves-Brown, Sian Jones, and Clive Gamble, 256–270. Routledge: London. Hingley, R. 1996. “The ‘legacy’ of Rome: the rise, decline, and fall of the theory of Romanization.” In Roman Imperialism: Post-Colonial Perspectives, eds. Jane Webster and Nicholas Cooper, 35–48. Leicester Archaeology Monograph, No. 3. Leicester: University of Leicester. ⎯⎯ 1997. “Resistance and domination: social change in Roman Britain.” In Dialogues in Roman Imperialism, ed. D. J. Mattingly, 81–102. JRA Supplementary Series, no. 23. ⎯⎯ 2000. Roman Officers and English Gentlemen. London: Routledge. Hodder, I. 1986. Reading the Past: current approaches to interpretation in archaeology. Cambridge: Cambridge University Press. Holčík, Š., and V. Turčan. 1998 “Rukoväť lanxu z Krakovian-Stráží.” Zbor. SNM 92, Arch. 8: 75–84. Honoré, T. 1987. “Scriptor Historiae Augustae.” JRS, vol. 77: 156–176. Hošek, R. 1985. Tituli Latini Pannoniae Superioris: Annis 1967–1982 in Slovacia Reperti. Praha: Univerzita Karlova. How cultivation affects soil. 2002 < http://www.ricecrc.org/reader/soil-types-structure/ss693-cultivation.htm> Visited March 2007. Isaac, B. 1988. “The meaning of limes and limitanei in ancient sources.” JRS 17: 125–74. ⎯⎯ 1993. The limits of empire. The Roman Army in the East. Revised edition. Oxford: Clarendon Press. James, S. 1993. The World of the Celts. London: Thames and Hudson. ⎯⎯ 2002. “Writing the Legions: The Development and Future of Roman Military Studies in Britain.” Archaeol. J. 159: 1–58. Jerem, E. 1993. “Celts of Eastern Europe.” In The Celtic World, ed. M. J. Green, 581–602. London: Routledge. Jones, A. 1970. Augustus. New York: W. W. Norton & Company. Jones, S. 1996. “Discourses of identity in the interpretation of the past.” In Cultural Identity and Archaeology, eds. Paul Graves-Brown, Sian Jones, and Clive Gamble, 62–80. London: Routledge. ⎯⎯ 1997. The Archaeology of Ethnicity. London: Routledge. 349

Jones, S. and P. Graves-Brown. 1996. “Introduction: archaeology and cultural identity in Europe.” In Cultural Identity and Archaeology, eds. Paul Graves-Brown, Sian Jones, and Clive Gamble, 1–24. London: Routledge. Kandler, M. 1997. “Carnuntum.” In Der Römische Limes in Österreich, eds. Herwig Friesinger and Fritz Krinzinger, 258–271. Wien: Verlag der Österreichischen Akademie der Wissenschaften Wien. Kelly, R. L., L. Poyer, and B. Tucker. 2005. “An Ethnoarchaeological Study of Mobility, Architectural Investment, and Food Sharing among Madagascar’s Mikea.” AA 107, no. 3: 403–416. Keyes, C. 1976. “Towards a new formation of the concept of ethnic group.” Ethnicity 3: 202–213. ⎯⎯ 1981. “The dialectics of ethnic change.” In Ethnic Change, ed. C. F. Keyes, 3–31. Seattle: University of Washington Press. Kilian, L. 1983. Zum Ursprung der Indogermanen. Bonn. ⎯⎯ 1988. Zum Ursprung der Germanen. Bonn. Kirschbaum, S. 1995. A History of Slovakia. New York: St. Martin’s Griffin. Kolník, T. 1959. “Ausgrabungen auf der römischen Station in Milanovce in den Jahren 1956–1957.” In Limes Romanus Konferenz, 27–48. Nitra: AÚ-SAV ⎯⎯ 1961 “Pohrebisko v Bešeňove.” Slov. Arch. 9: 219–300. ⎯⎯ 1962 “Nové sídliskové nálezy z doby rímskej na Slovensku. Arch. Rozhledy 14: 244–368, 371–380, 385–397. ⎯⎯ 1964 “Stav a bádanie o dobe rímskej na Slovensku.” Referáty Liblice 1: 123–127. ⎯⎯ 1971 “Prehľad a Stav Bádania o Dobe Rímskej a Sťahovani Národov.” Slov. Arch. 19: 499–548. ⎯⎯ 1977a “Anfänge der germanischen Besiedlung in der Südwestlowakei und das Regnum Vannianum.“ In Symposium Ausklang der Latène-Zivilisation und Anfänge der germanischen Besiedlung im mittleren Donaugebiet, 143–171. Bratislava: Slovenska akadémie vied. ⎯⎯ 1977b “Rímsky Nápis z Boldogu.” Slov. Arch. 25: 481–500. ⎯⎯ 1979 Skvosty Antiky na Slovensku. Bratislava: Tatran. ⎯⎯ 1980 Romerzeitliche Graberfelder in der Slowakei. Teil 1. Nitra: AÚ-SAV ⎯⎯ 1981 “The Roman and Great Migration Periods.” In Archaeological Research in Slovakia, ed. Bohuslav Chropovsky, 113–132. Nitra: AÚ-SAV. ⎯⎯ 1984 Rímske a germánske umenie na Slovensku. Bratislava: Tatran. ⎯⎯ 1991 “Römer und Barbaren im Nördlichen Mitteldonaubebiet.” In Roman Frontier Studies 1989, eds. V. Maxfield and M. Dobson, 432–434. Exeter: University of Exter Press. ⎯⎯ 1997 “Zum Anteil der Militäreinheiten beim Aufbau der Sogenannten Römischen Stationen im MittelDanubischen Barbaricum.” In Roman Frontier Studies 1995, eds. W. Groenman-van Waateringe, B. L. van Beck, W. J. H. Willems, and S. L. Wynia, 417–423. Oxford: Oxbow Monographs, 91. ⎯⎯ 1998a “II. Tschechische und Slowakische Republik Haus und Hof im quadischen Limesvorland.” In Haus und Hof im Östlichen Germanien. Universitätsforschungen zur prähistorischen Archäologie, 144–159. Bonn: Dr. Rudolf Hablet GmbH. ⎯⎯ 1998b “Archeologické svedectvo.” In Územie Slovenska Pred Príchodom Slovanov, ed. Richard Marsina, 53– 87. Vol. 1 of Pramenie k Dejinám Slovenska a Slovákov. Bratislava: ArtPress. ⎯⎯ 2000a “Stéla z Boldogu – najstarší náhrobný nápis na Slovensku.” Pam. Múz. 3: 20–21. ⎯⎯ 2000b “Cífer-Pác – záhada na pokračovanie Germánska rezidencia alebo aj rímska vojenská stanica?” Pam. Múz. 3: 41–44. Kolník, T; E. Krekovič; L. Snopko; I. Geržová; V. Ferus; I. Hečková. 1993. “Doba rímska.” In Najstaršie Dejiny Bratislavy, ed. Tatiana Štefanovičová, 210–262. Bratislava: ELÁN. Kolník, T., V. Varsik, and J. Vladár. 2005. “Branč, Objekt 60.” Zbor. SNM 99, Arch. 15: 313–323. Kolníková, E. 1998. “Numizmaticke pramené.” In Územie Slovenska Pred Príchodom Slovanov, ed. Richard Marsina, 34–39. Vol. 1 of Pramenie k Dejinám Slovenska a Slovákov. Bratislava: ArtPress. Korfmann, M. 1973. “The Sling as a Weapon.” Scientific American 229(4): 35–42. Kossack, G. 1981. “The Germans.” In The Roman Empire and its Neighbors. 2nd edition, Fergus Millar, 294–320. New York: Holmes & Meier Publishers, Inc. Kováč D. et. al. 1998. Kronika Slovenska. Bratislava: Fortuna Print & Adox. Kováč, M., J. Michalík, D. Plašienka, and M. Putiš. 1990. Malé Karpaty Mts.: Geology of the Alpine-Carpathian Junction. Guide to Excursions. Bratislava: Dionýz Štúr Institute of Geology. Kraskovská, Ľ. 1965. “Popolnicové Pohrebisko v Ivanke Pri Dunaji.” Slov. Arch. 13: 163–182. ⎯⎯ 1974a. Gerulata-Rusovce. Bratislava ⎯⎯ 1974b. “Rímske pohrebisko na parcele JRD v Rusoviciach.” Zbor. SNM 68, Hist. 14: 81–111. ⎯⎯ 1977a. “Rímske Žiarové Hroby v Bratislave-Devínskej Novej Vsi.” In Arch. Rozhledy 39: 487–491. ⎯⎯ 1977b. “Vyskum rimskej stanice v Stupava v roku 1976.” AVANS v Roku 1976: 178–179. Nitra: AÚ-SAV. ⎯⎯ 1978. Roman Bronze Vessels from Slovakia. Trans. Hana Schuck. Oxford: BAR International Series [44]. ⎯⎯ 1980. “Revízny Výskum Rímskej Stanice v Stupava.” AVANS v Roku 1979: 116–117. Nitra: AÚ-SAV. ⎯⎯ 1982. “Revízny Výskum Rímskej Stanice v Stupava v Rokoch 1978 a 1979.” Zbor. SNM 76, Hist. 22: 51–73. Krause, C. L. and C. Mishler. 1975. Standard Catalog of World Coins. Iola: Krause Publications. 350

Krekovič, E. 1979. “Hrobové Nálezy z Doby Rímskej v Šoproni, Okr. Galanta.” Arch. Rozhledy 31: 413–419. ⎯⎯ 1981. “Rímska Importovaná Keramika na Slovensku.” Slov. Arch. 29: 341–374. ⎯⎯ 1983. “Römische Lampenfunde im Slowakischen Barbaricum.” Arch. Rozhledy 35: 510–516, 595. ⎯⎯ 1987. “Rímske Importy na Slovensku.” Pam. Arch. 78: 231–282. ⎯⎯ 1997a. “The Structure of Roman Imports in Slovakia.” In Roman Frontier Studies 1995, eds. W. Groenmanvan Waateringe, B. L. van Beck, W. J. H. Willems, and S. L. Wynia, 233–236. Oxford: Oxbow Monographs, 91. ⎯⎯ 1997b, “Marbled Ware in Pannonia and the Roman Army.” Rei Cretariae Romanae Fautorum Acta 35: 41–44. ⎯⎯ 1998. Römische Keramik aus Gerulata. Studia Archaeologia Et Mediaevalia, Tomus IV. Bratislava: Academic Electronic Press. ⎯⎯ 2000a. “Kniežacie hroby.” Pam. Múz., 3: 48–51. ⎯⎯ 2000b. “Roman Pottery Beyond the Frontier: The Middle Danubian Region.” Rei Cretariae Romanae Fautorum Acta 36: 259–262. ⎯⎯ 2001a. “Roman fibulas in Slovakia.” In International Connections of the Barbarians of the Carpathian Basin in the 1st – 5th centuries A.D, 95–100. Aszód –Nyíregyháza. ⎯⎯ 2001b. “Some Comparisons Between Germanic and Sarmatian Limes in Pannonia.” In Acta Arch. Acad. Scient. Hungaricae 52: 103–107. Budapest: Akadémiai Kiadó. Křížek, F. 1939. Terra Sigillata in der Slowakei. Brno: J. Martinek. ⎯⎯ 1961. “Nové nálezy terry sigillaty na Slovensku. Slov. Arch. 9: 301–324. ⎯⎯ 1966. “Nové nálezy terry sigillaty na Slovensku II. Slov. Arch. 14: 97–145. Kulakov, V. I. 2005. The Amber Lands in the Time of the Roman Empire. BAR International Series 1356. Oxford: Arcaeopress. Kuzmová, K. 1980. “Nížinné Sidliská z Neskorej Doby Laténskej v Strednom Podunajsku.” Slov. Arch. 28: 313–340. ⎯⎯ 1997. Terra Sigillata im Vorfeld des Nordpannonischen Limes. Nitra: AÚ-SAV. Kuzmová, K., and Rajtár, J. 1996. Gerulata I. Nitra: AÚ-SAV. Kuzmová, K., and Roth, P. 1988. Terra Sigillata v Barbariku. Nitra: AÚ-SAV. Kuzma, I., E. Hanzelyová, J. Rajtár, and J. Tirpák. 1996. “New Results in Aerial Archaeology in Slovakia: Experience with Reconnaissance, Geophysical Measurement and Follow-up Excavations.” Archaeological Prospection 3: 71– 79. Kreuz, A. 1999. “How to become a Roman farmer: a preliminary report on the environmental evidence from the Romanization project.” In Roman Germany, eds. J. D. Creighton and R. J. A. Wilson, 71–98. JRS Supplementary Series, 32. Portsmouth. Lamiová-Schmiedlová, M. 1969. “Römerzeitliche Siedlungskeramik in der Südostslowakei.” Slov. Arch. 17: 403–501. ⎯⎯ 1977. “Ältere römische Kaiserzeit in der Ostslowakei.“ In Symposium Ausklang der Latène-Zivilisation und Anfänge der germanischen Besiedlung im mittleren Donaugebiet, 201–208. Bratislava: Slovenska akadémie vied. Lengyel, A. and G. Radan. 1980. The Archaeology of Roman Pannonia. Budapest: Akadémiai Kiadó. Lightfoot, K. and A. Martinez. 1995. “Frontiers and Boundaries in Archaeological Perspective.” Annual Review of Anthropology 24: 471–492. Lind, L. 1962. See Vergil above in primary sources. Lintott, A. 1981. “What Was the ‘Imperium Romanum’?” Greece & Rome. 2nd Ser., vol. 28, no. 1, Jubilee Year: 53–67. Luckey, C. and R. Lewis. 2003. Collecting Antique Bird Decoys and Duck Calls. 3rd edition. Iola: Krause Publications. Luttwak, E. 1976. The Grand Strategy of the Roman Empire. Baltimore: The Johns Hopkins University Press. MacKinnon, M. 2004. Production and Consumption of Animals in Roman Italy. JRA Supplementary Series, 54. Portsmouth. MacMullen, R. 2000. Romanization in the Time of Augustus. New Haven: Yale University Press. Macphail, R. I., and Linderholm, J. In Press. “Interpreting fills of grubenhäuser: Examples from England and Sweden.” In Archaeological Soil Micromorphology. ed. L. Trombino. Milan: Quademi di Geodinamica Alpina e Quaternaria. Mączyńska, M. 1976. “Paciorki z okresu rzymskiego i wczesnej fazy okresu wędrówek ludów na obszarze środkowoeuropiejskiego barbaricum.“ Archeologia 28: 61–96. Mann, J. 1979. “Power, force and the frontiers of empire.” JRS. Volume 69: 175–183. ⎯⎯ 1991. “The Notitia Dignitatum – Dating and Survival.” Britannia, vol. 22: 215–219. Maróti, E. 1991. “A római kori pecsételt kerámia és e Resatus kérdés.“ Stud. Comit. 21: 365–427. Martin, R. 1981. Tacitus. Berkeley: University of California Press. Marsina, R., ed. 1998. Územie Slovenska Pred Príchodom Slovanov. Vol. 1 of Pramenie k Dejinám Slovenska a Slovákov. Bratislava: ArtPress. Mattern, S. 1999. Rome and the Enemy. Berkeley: University of California Press. Mattingly D. 1997. Dialogues in Roman Imperialism. JRA, Supplementary Series, no. 23. Portsmouth. ⎯⎯ 1997. “Dialogues of power and experience in the Roman Empire.” In Dialogues in Roman Imperialism, ed. D. Mattingly. JRA Supplementary Series 23: 7–26. ⎯⎯ 2004. “Being Roman: expressing identity in a provincial setting.” JRA. Volume 17, Part I: 5–26. Maxfield, V. and M. J. Dobson. 1991. Roman Frontier Studies 1989. Exeter: University of Exeter Press. McManamon, F. 2000. “Surveying and Site Examination, Manual Methods.” In Archaeological Method and Theory: 351

An Encyclopedia, ed. Linda Ellis, 605–610. New York: Garland Publishing Inc. Mellor, R. 1993. Tacitus. New York: Routledge. Mielcke, R. 1915. “Die angeblich germanischen Rundbauten an der Markussäule in Rom.” Zeitschr. Ethn. 47: 75–91. Millar, F. 1964. A Study of Cassius Dio. Oxford: Clarendon Press. ⎯⎯ 1981. The Roman Empire and its Neighbors. 2nd edition. New York: Holmes & Meier Publishers, Inc. ⎯⎯ 1982. “Emperors, frontiers and foreign relations.” Britannia 13: 1–23. Millett, M. 1990. The Romanization of Britain. Cambridge: Cambridge University Press. Ministr, Z. 1999. “Astronomical orientation and calendar function of prehistoric structures.” In Primeval socio-ritual architecture in Moravia, ed. V. Podborský, 248–259. Brno: Archaeology and Museology Department, FA-MU. ⎯⎯ www.webproster.cz/veda_a_vyzkum/ministr/mini5a.html. Visited on April 17, 2005. Mócsy, A. 1961. Pannonia. From Pauly-Wissowa, Realencyclopädie der classischen Altertumswissenschaft Supplementband IX. Stuttgart: Alfred Druckenmüller Verlag. ⎯⎯ 1974. Pannonia and Upper Moesia. London: Routledge & Kegan Paul. ⎯⎯ 1980. “Der pannonische Limes: Probleme der neueren Forschungen.” In Roman Frontier Studies 1979, eds. Hanson, W. and L. J. F. Keppie, 627–636, vol. 2. Oxford: BAR International Series, 71 (2 vols.). Mommsen, T. 1874–1886. Römische Geschichte. Berlin: Weidmann. Nagy, L. 1928. “Egy pannóniai terra sigillata gyár. Arch. Ért. 47: 96–113. Nance, J. and B. Ball 1986. “No Surprises? The Reliability and Validity of Test Pit Sampling.” American Antiquity 51(3): 457–483. Neustupný, E. and J. Neustupný. 1961. Czechoslovakia Before the Slavs. London: Thames and Hudson. Novotná, M. 1993. “Doba bronzová.” In Najstaršie Dejiny Bratislavy, ed. Tatiana Štefanovičová, 80–115. Bratislava: ELÁN. Nowak, M. 2004. “Is ‘Pot History’ Real Prehistory? The Case of the Early LBK.” In LBK Dialogues, eds. Alena Lukes and Marek Zvelebil, 7–16. Oxford: BAR International Series, 1304. O’Nions, R. K. and C. J. Gallentine 1990. “Rare Gas Studies of Basin Scale Fluid Movement.” Philospophical Transactions: Physical Sciences and Engineering, vol. 344, no. 1670: 141–156. Okun, M. L. 1991. “Pluralism in Germania Superior.” In Roman Frontier Studies 1989, eds. V. Maxfield and M. J. Dobson, 435–438. Exeter: University of Exeter Press. Ondrouch, V. 1938. Limes Romanus na Slovensku. Bratislava: Učená Společnost Šafaříkova. ⎯⎯ 1945–1946. “Rímske Stanice v Stupave a Pajštúne.” In Historica Slovaca, vols. 3–4: 1–58. Bratislava: Slovenská Akadémia Vied a Umeni. ⎯⎯ 1957. Bohaté Hroby z Doby Rímskej na Slovensku. Bratislava: Slovenskej Akadémie Vied. Orton, C; Tyers, P. and Vince, A. 1993. Pottery in Archaeology. Cambridge: Cambridge University Press. Owen, F. 1988. The Germanic People. Reprint of the 1960 edition. New York: Dorset Press. Pavlů, J. 2004. “The Origins of the Early Linear Pottery Culture in Bohemia.” In LBK Dialogues, eds. Alena Lukes and Marek Zvelebil, 83–90. Oxford: BAR International Series, 1304. Pavuk, J. and Siska, S. 1981. “The Neolithic and Eneolithic.” In Archaeological Research in Slovakia. Ed. B. Chropovsky, 31–60. Nitra: The Institute of Archaeology of the Slovak Academy of Sciences. Pelham, H. 1897. “The Roman frontier system.” Transactions of the Cumberland and Westmorland Antiquarian and Archaeological Society 14: 170–184. ⎯⎯ 1906. “The Roman frontier in southern Germany.” Transactions of the Royal Historical Society 20: 17–47. Pelicán, O. 1960. Slovensko a Rímske Impérium. Bratislava: Krásney Literatúry. Pichlerová, M. 1981. Gerualta Rusovce: Rímske Pohrebisko II. Bratislava: N. P. Martin. Pichlerová, M. and A. Neumann. 1979. “Rímsky náhrobný kameň v Boldogu (Interrex v canabae légie z Carnunta?). Zbor. SNM 73, Hist. 19: 51–60. Pichlerová, M. and K. Tomčíková. 1994. “Archeologická zbierka zo Želiezoviec II.” Zbor. SNM 88, Arch. 4: 85–128. ⎯⎯ 1995. “Archeologická zbierka zo Želiezoviec III.” Zbor. SNM 89, Arch. 5: 91–128. Pieta, K. 1981. “The La Tène Period.” In Archaeological Research in Slovakia, ed. Bohuslav Chropovsky, 97–112. Nitra: AÚ-SAV. ⎯⎯ 1982. Die Púchov – Kultur. Nitra: AÚ-SAV. ⎯⎯ 1996. Liptovská Mara: Včasnohistorické centrum severného Slovenksa. Archeologické pamätníky Slovenska, zv. 5. Bratislava: Slovenská Akadémia Vied. Pieta, K. and V. Plachá. 2000. “Nové objavy na rímskom Devíne.” Pam. Múz. 3: 6–9. Pieta, K., J. Rajtár, and J. Tejral, eds. 1995. Kelten, Germanen, Römer im Mitteldonaugebiet. Brno: Archeologický USTAV Akademie Věd České Republiky. Pieta, K. and L. Zachar. 1993. “Mladiša doba železná (laténska).” In Najstaršie Dejiny Bratislavy, ed. Tatiana Štefanovičová, 143–209. Bratislava: ELÁN. Pitts, L. 1987. “Roman Style Buildings in the Barbaricum (Moravia and SW Slovakia).” Oxford Journal of Archaeology. Vol. 6, issue 2. ⎯⎯ 1989. “Relations between Rome and the German ‘Kings’ on the Middle Danube in the First to Fourth Centuries A.D.” JRS 79: 45–58. 352

Plachá, V.; J. Hlavicová; I. Keller. 1990. Slovanský Devín. Bratislava: Obzor. Poczy, K. 1988. Keramik, Intercisa II. AH n.s., XXXVI. Budapest. Potter, D. 1992. “Empty Areas and Roman Frontier Policy.” AJP 113, no. 2, Summer: 269–274. Purcell, N. 1996a. “itineraries,” in the OCD, eds. Simon Hornblower and Antony Spawforth. 3rd edition: 775. ⎯⎯ 1996b. “Peutinger Table,” in the OCD, eds. Simon Hornblower and Antony Spawforth. 3rd edition: 1151. Radnoti, A. 1938. Bronzegefässe aus Pannonien. Diss. Pann. 2–6. Budapest. Rajtár. J. 1986. “Výskyt Mincí v Rímskych Stavebných Objektoch na Slovensku.” Slov. Num. 9: 99–121. ⎯⎯ 1996. “Limes romanus a rímske opevnenia na Slovensku.” Pam. Múz. 3: 18–23. ⎯⎯ 2000. “Kastel v Iži – hraničná pevnosť na Dunaji.” Pam. Múz., no. 3: 34–38. Renfrew, C. 1996. “Prehistory and the identity of Europe, or, don’t let’s be beastly to the Hungarians.” In Cultural Identity and Archaeology, eds. Paul Graves-Brown, Sian Jones, and Clive Gamble, 125–137. London: Routledge. Renfrew, C. and Bahn, P. 2000. Archaeology: Theories, Methods and Practice. 3rd edition. New York: Thames and Hudson. Rich, J. 1996. “Cassius Dio,” in the OCD, eds. Simon Hornblower and Antony Spawforth. 3rd edition: 299–300. Riha, E. 1988. Der römische Schmuck aus Augst und Kaiseraugst. Augst: Römermuseum Augst. Riley, M. 1995. “Ptolemy’s Use of His Predecessor’s Data.” TAPA 125: 221–250. Russell, L. 2001. “Introduction.” In Colonial Frontiers: Indigenous-European Encounters in Settler Societies, ed. L. Russell, 1–16. Manchester: Manchester University Press. Ruttkay, A. T. 2000. “Slavs in the Central Danubian region from the 8th till the 11th century and the beginnings of the history of the Slovak nation.” In Slovaks in the Central Danubian Region in the 6th to 11th Century, eds. M. Kucera and B. Egyhazy-Jurovska, 67–92. Bratislava: Slovenské národné múzeum. Saddington, D. B. 1991. “The Parameters of Romanization.” In Roman Frontier Studies 1989, eds. V. Maxfield and M. J. Dobson, 413–418. Exeter: University of Exeter Press. Salač, V. 1995. “Zur ältesten germanischen Besiedlung Böhmens.” In Kelten, Germanen, Römer im Mitteldonaugebiet, eds. Karol Pieta, Ján Rajtár, and Jaroslav Tejral, 145–176. Brno: Archeologický USTAV Akademie Věd České Republiky. Seager, R. 2005. Tiberius. 2nd edition. Malden: Blackwell Publishing. Service. E. 1962. Primitive Social Organization. New York: Random House. Shchukin, M. 1989. Rome and the Barbarians in Central and Eastern Europe. Parts 1 and 2. Oxford: BAR International Series 542. Shaffer, G. D. 1993. “An Archaeomagnetic Study of Wattle and Daub Building Collapse.” JFA 20, no. 1 (Spring, 1993): 59–75. Shanks, M. and C. Tilley. 1987. Social Theory and Archaeology. Oxford: Polity Press. Shennan, S. 1994. “Introduction: archaeological approaches to cultural identity.” In Archaeological Approaches to Cultural Identity, ed. S. Shennan, 1–32. London: Routledge. Shipley, F. See Res Gestae Divi Augusti under Primary Authors. Sihler, E. 1923. “Strabo of Amaseia: His Personality and His Works.” AJP 44, no. 2: 134–144. Sklenář, K. 1983. Archaeology in Central Europe: The First 500 Years. New York: St. Martin’s Press. Soproni, S. 1980 “Limes.” In The Archaeology of Roman Pannonia, eds. A. Lengyel and G. T. B. Radan, 219–238. Budapest: Akadémiai Kiadó. Southern, P. 1997 Domitian. Bloomington: Indiana University Press. Staník, I. and V. Turčan. 2000. “Rímska Stanica v Stupave.” Pam. Múz. 3: 22–26. Starr, R. 1980. “The Scope and Genre of Velleius’ History.” The Classical Quarterly, New Series, vol. 31, no. 1: 162– 174. Steinhubel, J. 2000. “The migration of nations and the arrival of the Slavs in Slovakia.” In Slovaks in the Central Danubian Region in the 6th to 11th Century, eds. M. Kucera and B. Egyhazy-Jurovska, 31–44. Bratislava: Slovenské národné múzeum. Stuiver, M., et. al. 1998. “INTCAL 98 Radiocarbon Age Calibration.” Radiocarbon 40 (3): 1041–1083. Stuiver, M. and H. van der Plicht. 1998. “Editorial Comment.” Radiocarbon 40 (3): xii– xiii. Sumner, G. 1970. “The Truth about Velleius Paterculus: Prolegomena.” Harvard Studies in Classical Philology 74: 257–297. Swan, O. E. and S. Gálová-Lorinc. 1990. Beginning Slovak. Columbus: Slavica Publishers, Inc. Syme, R. 1968. Ammianus and the Historia Augusta. Oxford: Clarendon Press. Szőnyi, E. 1972. “Észak-Itáliából importált császárkori agyagcsészék a győri Xántu János múzeumban.” Arrabona 14: 5–26. ⎯⎯ 1973. “Zur Verbreitung und Herstellung der sog. Rätischen Keramik in Pannonien. Acta Arch. Hung. 25: 87– 108. Talbert, R. J. A. 1985. Atlas of Classical History. London: Routledge. Talma, A. S. and J. C. Vogel. 1993. “A Simplified Approach to Calibrating C14 Dates.” Radiocarbon 35 (2): 317–322. Tempelmann-Mączyńska, M. 1985. Die Perlen der römischen Kaiserzeit und der frühen Phase der Völkerwanderungszeit im mitteleuropäischen Barbaricum. Mainz. 353

Thompson, E. 1965. The Early Germans. Oxford: Clarendon Press. ⎯⎯ 1982. Romans and Barbarians. Madison: The University of Wisconsin Press. Thompson, L. and H. Lamar. 1981. “Comparative Frontier History.” In The Frontier in History, eds. Howard Lamar and Leonard Thompson, 3–13. New Haven: Yale University Press. Točik, A. 1980. “Pokračovanie záchranného výskumu v polohe Kňazova jama v Komjaticiach. AVANS 1979: 215–229. Todd, Malcolm. 1975. The Northern Barbarians: 100 BC–AD 300. London: Hutchinson University Library. ⎯⎯ 2004. The Early Germans. 2nd edition. Oxford: Blackwell. Tomlin, R. 1996. “Notitia Dignitatum,” in the OCD, eds. Simon Hornblower and Antony Spawforth. 3rd edition: 1049. Trigger, B. G. 1989. A History of Archaeological Thought. Cambridge: Cambridge University Press. Turčan, V. 1984. “Sídliskové Nálezy z Doby Rímskej zo Senca a Kráľovej Pri Senci.” In Zborník Prác Ľudmile Kraskovskej,k Zivotnémi jubilee, 119–130. Bratislava: SNM. ⎯⎯ 1985. “Germánsky Výrobný Objekt zo Stupavy.” Zbor. SNM 79, Hist. 25: 93–116. ⎯⎯ 1988. “Germánske Sídliskové Nálezy z Kuchyne.” Zbor. SNM 82, Hist. 28: 107–125. ⎯⎯ 1990a. “Pamiatky doby rímskej v Stupave.” Prax 1: 40–41. ⎯⎯ 1990b. “Rimania v Stupave.” Príroda a spoločnosť 5: 56–59. ⎯⎯ 1991. “Stupava.” In Excursion Guide for the International Union of Prehistoric and Protohistoric Sciences, XIIth Congress: 66–67. ⎯⎯ 1996. “Germánske Sídliskové Objekty Medzi Trnavou a Bohdanovcami.” Zbor. SNM 90, Arch., 6: 107–122. ⎯⎯ 1997. “Nalezova Sprava z vyskumu germanskeho objektu v Stupave (Okr. Malacky) poloha Moravkove pole.” Unpublished field-notes from the excavation, stored at AM-SNM, Bratislava. ⎯⎯ 2000. “Rukoväť k antickému lanxu z Krakovian-Stráží.” Pam. Múz. 3: 52. ⎯⎯ 2002. “Sucasny Stav Archeologickeho Vyskumu Antiky na Slovensku.” Res Historica 14: 233–251. ⎯⎯ 2005. “Objekt č. 1/96 zo Stupavy, poloha Morávkove pole.” Zbor. SNM 99, Arch. 15: 293–300. Turner, F. 1920. The Frontier in American History. New York: Henry Holt and Company. Ucko, P. 1994. “Forward.” In Archaeological Approaches to Cultural Identity.” Ed. S. Shennan, ix–xx. Routledge: New York. Urban, O. 1997. “Stillfried – Kirchberg.” In Der Römische Limes in Österreich, eds. Herwig Friesinger and Fritz Krinzinger, 293–297. Wien: Verlag der Österreichischen Akademie der Wissenschaften Wien. Valachovič, P. 1998. “Správy antických autorov o Slovensku.” In Územie Slovenska Pre Príchodom Slovanov, ed. Richard Marsina, 21–33. Vol. 1 of Pramena k Dejinám Slovenska a Slovákov. Bratislava: ArtPress. Varsik, V. 1996. “Das Römische Lager von Rusovce-Gerulata.“ In Jahrbuch des Römisch-Germanischen Zentralmuseums Mainz 43: 531–600. 1998. “Štruktúra germánskych sídlisk z. 2.–3. storičia na juhozápadnom Slovensku (Branč a Veľký Meder).” ⎯⎯ Summary of dissertation work done by the author. Bratislava: Slovenksá Academia vied. ⎯⎯ 2002. “Besiedlung in der Älteren Römischen Kaiserzeit am Östlichen Rand von Bratislava.” Slov. Arch. 50: 127–152. ⎯⎯ 2005a. “Bratislava-Trnávka “Zadné”, objekt 44.” Zbor. SNM 99, Arch. 15: 275–292. ⎯⎯ 2005b. “Veľký Meder, objekt 301/03.” Zbor. SNM 99, Arch. 15: 301–312. Vaškovský, I. 1977. Kvartér Slovenska. Bratislava: Geologický Ústav Dionýza Štúra. Veit, U. 2002. “Gustaf Kossinna and his concept of a national archaeology.” In Archaeology, Ideology and Society. 2nd edition, ed. Heinrich Härke, 41–66. Frankfurt: Peter Lang. Vrba, E. 1997. “An Archaeological Survey of the Roman Presence in Slovakia.” Master’s Thesis, The University of Arizona. Unpublished. ⎯⎯ 2004. “Archaeological Survey at Urbárske Sedliská, Stupava, Slovakia: Preliminary Report.” Zbor. SNM 98, Arch. 14: 129–156. Wallace-Hadrill, J. 1971. Early Germanic Kingship in England and on the Continent. Oxford: Clarendon Press. Waters, M. 1992. Principles of Geoarchaeology. Tucson: The University of Arizona Press. Webb, W. 1931. The Great Plains. Boston: Ginn & Co. Webster, J. 1996. “Roman imperialism and the ‘post imperial age.’ In Roman Imperialism: Post-Colonial Perspective. Eds. J. Webster and N. Cooper. Leicester Archaeology Monographs No. 3: 1–17. Leicester: University of Leicester. Wells, C. 1972. The German Policy of Augustus. Oxford: Clarendon Press. Wells, P. 1980. Culture contact and culture change. Cambridge: Cambridge University Press. ⎯⎯ 1992. “Tradition, Identity, and Change Beyond the Roman Frontier.” In Resources, Power, and Interregion Interaction. Eds. Edward M. Schortman and Patricia A. Urban, 175–192. New York: Plenum Press. ⎯⎯ 1999. The Barbarians Speak. Princeton: Princeton University Press. ⎯⎯ 2001. Beyond Celts, Germans and Scythians. London: Duckworth. ⎯⎯ 2005. “Creating an Imperial Frontier: Archaeology of the Formation of Rome’s Danube Borderland.” Journal of Archaeological Research 13, no. 1, March: 49–88. Wessely, G. 1988. “Structure and Development of the Vienna Basin in Austria.” In The Pannonian Basin: A Study in Basin Evolution, eds. L. H. Royden and F. Horvath. Budapest: The Hungarian Geological Society. AAPG, 354

Memoir 45. Wheeler, E. 1993a. “Methodological Limits and the Mirage of Roman Strategy: Part I.”The Journal of Military History 57, no. 1, January: 7–41. ⎯⎯ 1993b. “Methodological Limits and the Mirage of Roman Strategy: Part II.” The Journal of Military History 57, no. 2, April: 215–240. White, P. 1967. “The Authorship of the Historia Augusta.” JRS 57, no. 1/2: 115– 133. Whittaker, C. R. 1989. “Supplying the System: Frontiers and Beyond.” In Barbarians and Romans in North-West Europe, eds. J. C. Barrett, A. P. Fitzpatrick, and L. Macinnes, 64–89. Oxford: BAR International Series 471. ⎯⎯ 1994. Frontiers of the Roman Empire. Baltimore: Johns Hopkins University Press. ⎯⎯ 2004. Rome and Its Frontiers: The Dynamics of Empire. London: Routledge. Whittle, A. 1996. Europe in the Neolithic. Cambridge: Cambridge University Press. Wielowiejski, J. 1970. Kontakty Noricum i Pannonii z ludami północnymi. Wrocław: Zakład Narodowy im. Ossolińskich Wydawnictwo ⎯⎯ 1980. Glowny Szlak Bursztynowy w Czasach Cesarstwa Rzymskeiego. Wrocław: Zakład Narodowy im. Ossolińskich Wydawnictwo. Wilkes, J. 1979. “Importation and Manufacture of Stamped Bricks and Tiles in the Roman Province of Dalmatia.” In Roman Brick and Tile: Studies in Manufacture, Distribution and Use in the Western Empire, ed. A. McWhirr, 65– 72. BAR International Series, 68. Oxford. Williams D. 1996. The Reach of Rome. New York: St. Martin’s Press. Wolfram, H. 1988. History of the Goths. Berkeley: University of California Press. —— 1997. The Roman Empire and Its Germanic Peoples. Berkeley: University of California Press. Wolfram, S. 2002. “Vorsprung durch Technik or ‘Kossinna Syndrome’? Arcaheological theory and social context in post-war West Germany.” In Archaeology, Ideology and Society. 2nd edition, ed. Heinrich Härke, 183–204. Frankfurt: Peter Lang. Wolfskill, G. and S. Palmer. 1981. “Introduction.” In Essays on Frontiers in World History. Eds. G. Wolfskill and S. Palmer, ix–xvi. Austin: University of Texas Press. Wood, M. and Queiroga, F. 1992. Current Research on the Romanization of the Western Provinces. Woolf, G. 1998. Becoming Roman. Cambridge: Cambridge University Press. Wolters, R. 2002. Římané v Germánii. Praha: Vyšehrad. Czech translation of German orignial, Die Römer in Germanien (2000), Munich: C. H. Beck. Wulf, D. 1999. “Did the Locker Plant Steal Some of My Meat?” Meat Science Extension and Research Visited in February 2007. Zvelebil, M. 2004. “Conclusion: The Many Origins of LBK.” In LBK Dialogues, eds. Alena Lukes and Marek Zvelebil, 183–205. Oxford: BAR International Series, 1304. Žitňan, A. Visited in April 2005.

355