The Oxford Handbook of Roman Britain 0191002526, 9780191002526

Citation preview

T h e Ox f o r d H a n d b o o k o f

ROM A N B R I TA I N

ii

The Oxford Handbook of

ROMAN BRITAIN Edited by

MARTIN MILLETT LOUISE REVELL and

ALISON MOORE

1

iv

3 Great Clarendon Street, Oxford, ox2 6dp, United Kingdom Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries © Oxford University Press 2016 The moral rights of the authors‌have been asserted First Edition published in 2016 Impression: 1 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by licence or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this work in any other form and you must impose this same condition on any acquirer Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America British Library Cataloguing in Publication Data Data available Library of Congress Control Number: 2016944843 ISBN 978–​0–​19–​969773–​1 Printed in Great Britain by Clays Ltd, St Ives plc Links to third party websites are provided by Oxford in good faith and for information only. Oxford disclaims any responsibility for the materials contained in any third party website referenced in this work.

For Richard Reece

vi

Contents

List of Figures  List of Tables  List of Contributors  Introduction 

xi xxi xxiii xxvii

PA RT I   NAT U R E OF T H E E V I DE N C E 1. Early Studies in Roman Britain: 1610 to 1906  Richard Hingley

3

2. Roman Britain since Haverfield  Martin Millett

22

3. Romano-​British Archaeology in the Early Twenty-First Century Pete Wilson

43

4. The Development of Artefact Studies  Ellen Swift

63

5. The Textual and Archaeological Evidence  Henry Hurst

95

6. The Early Roman Horizon  Lacey Wallace

117

7. Britain at the End of Empire  Simon Esmonde Cleary

134

8. Britain before the Romans  Timothy Champion

150

9. Beyond Hadrian’s Wall  Fraser Hunter

179

viii

viii   Contents

10. Mobility, Migration, and Diasporas in Roman Britain  Hella Eckardt and Gundula Müldner

203

11. Multiculturalism on Hadrian’s Wall  Claire Nesbitt

224

12. Britons on the Move: Mobility of British-​Born Emigrants in the Roman Empire Tatiana Ivleva 13. Britain, Gaul, and Germany: Cultural Interactions  Tom Moore

245 262

PA RT I I   S O C I E T Y A N D T H E I N DI V I DUA L 14. Inscriptions and Identity  Valerie M.  Hope 15. Ideas of Childhood in Roman Britain: The Bioarchaeological and Material Evidence  Rebecca Gowland

285

303

16. The Life Course  Alison Moore

321

17. Status and Burial  John Pearce

341

18. Gender in Roman Britain  Melanie Sherratt and Alison Moore

363

19. Deviancy in Late Romano-British Burial Belinda Crerar

381

20. Clothing and Identity  H. E. M. Cool

406

21. Cemeteries and Funerary Practice  Jake Weekes

425

22. Identity and the Military Community in Roman Britain  Ian Haynes

448

Contents   ix

23. Roman Military Culture  Lindsay Allason-​Jones

464

PA RT I I I   F OR M S OF K N OW L E D G E 24. Changing Materialities  Andrew Gardner 25. Forms of Knowledge: Changing Technologies of Romano-​British Pottery  Jeremy Evans

481

510

26. Metals and Metalworking  David Dungworth

532

27. Medicine  Patricia Baker

555

28. Sociolinguistics  Alex Mullen

573

29. Art in Roman Britain  Ben Croxford

599

30. Names of Gods  Amy Zoll

619

31. Ritual Deposition  Alex Smith

641

32. Christianity in Roman Britain  David Petts

660

33. Memories of the Past in Roman Britain  Zena Kamash

681

PA RT I V   L A N D S C A P E A N D E C ON OM Y 34. ‘By Small Things Revealed’: Rural Settlement and Society  Martin Millett

699

x

x   Contents

35. Rural Transformation in the Urbanized Landscape  Martin Pitts

720

36. The Development of Towns  Adam Rogers

741

37. Urban Monumentality in Roman Britain  Louise Revell

767

38. The Exploitation of Animals in Roman Britain  Mark Maltby

791

39. Arable Farming, Horticulture, and Food: Expansion, Innovation, and Diversity Marijke van der Veen

807

40. Coinage and the Economy  Philippa Walton and Sam Moorhead

834

41. Economy and Power in Late Roman Britain  James Gerrard

850

Index 

869

List of Figures

1.1 The Cogidubnus inscription from Chichester, Sussex. From Gale (1723). Reproduced by permission of Durham University Library.

7

1.2 Plan of a Roman villa excavated by Samuel Lysons at Bignor in Sussex. Source: Lysons (1815: plate 19). Reproduced by permission of Durham University Library.

10

1.3 Knook Castle and British villages, Wiltshire. Source: Hoare (1810). Reproduced by permission of Durham University Library.

12

1.4 The civil and military districts. Source: Haverfield (1912: figure 1). Reproduced by permission of Durham University Library.

15

1.5 The Roman site at Chesters (Cilurnum) showing the excavated areas and ‘forum’. Source: Budge (1903: opposite p. 98). Reproduced by permission of Durham University Library.

16

1.6 Sketch plan of the graves forming a ‘family circle’ at Aylesford, Kent. Source: Evans (1890: figure 4). Reproduced by permission of Durham University Library.

17

3.1 Excavation of a major Roman building by Oxford Archaeology South in advance of housing development just outside the walled town of Cirencester in 2009. Source: © Oxford Archaeology South, with thanks to Berkeley Homes for agreement to reproduce.

45

3.2 Volunteers excavating Building 16 at Piddington villa, Northamptonshire, in 2012. Source: © Pete Wilson.

46

3.3 Volunteers excavating the sequence of barracks in the north-​west quadrant of the fort at Vindolanda in 2011. Source: © The Vindolanda Trust. 53 3.4 Combined geophysical surveys undertaken by the Landscape Research Centre with the support of Historic England and the Aggregates Levy Sustainability Fund on the southern side of the Vale of Pickering, North Yorkshire; depicts more than 1.5 kilometres of late Iron Age and Roman trackside ribbon development or ‘ladder settlement’ in addition to a wealth of other prehistoric and post Roman features. Source: © Landscape Research Centre. 4.1 Gender profiles in artefact studies. (a) Percentage of men and women with published articles on small finds listed in the Society of Antiquaries

55

xii

xii   List of Figures subject catalogue from 1900 to 1988 (end date of catalogue). (b) Percentage of men and women publishing articles in Britannia 1970–​89 (excluding reviews). (c) Subject of articles by women in Britannia 1970–​89 (excluding reviews). Source: © Ellen Swift.

78

5.1 Memorandum on wooden leaf tablet from Vindolanda, referring to the British use of cavalry (Tab. Vindol. II. 164): dimensions 78 x 186 millimetres. After a punctuation mark in l. 4, the text reads: ‘nec residunt | Brittunculi ut iaculos | mittant’ (‘nor do the Brits mount in order to throw javelins’). Source: Full text and translation in Bowman and Thomas (1994: 107–108). © The Trustees of the British Museum.

97

5.2 Sale of a slave girl, Fortunata, recorded on a stylus tablet from London, drawn by R. Tomlin: dimensions c. 140 x 114 millimetres, probably one of three tablets. The writing survived as scratches in the wooden tablet, becoming illegible where a triangular patch of the original wax still survived; [p]‌uellam Fortunatam are the first two words of l. 3. Source: Full text and translation in Tomlin (2003, 2011). © Roger Tomlin.

98

5.3 Curse tablet no. 43 from the sanctuary of Mercury at Uley, Gloucestershire, drawn by R. Tomlin: dimensions c. 95 x 83 millimetres. Docilinus asks Mercury to drive Varianus, Peregrina, and Sabinianus to death for injuring his farm animal unless they redeem their action. Source: Text and translation in Hassall and Tomlin (1989: 329–​331, no. 3) and Tomlin (2002: 172). © Roger Tomlin.

99

5.4 Tile with personal-​name graffito scratched before firing, Candid[us]. From a fourth-century deposit at Hayton, East Yorkshire. Source: Tomlin and Hassall (2001: 393). Photograph: M. Millett.

101

5.5 Tombstone of Classicianus from London: pieces as displayed in the British Museum with reconstruction of original monument drawn by Richard Grasby (Grasby and Tomlin 2002: figure 21), slightly adjusted. Reconstructed text (as in RIB 12): ‘Dis | [M]‌anibus | [G(ai) Iul(i) G(ai) f(ili) F]ab(ia tribu) Alpini Classiciani | ... | ... | proc(uratoris) provinc(iae) Brita[nniae] | Iulia Indi filia Pacata I[ndiana(?)] | uxor [f(ecit)]’ (‘To the spirits of the departed (and) of Gaius Julius Alpinus Classicianus, son of Gaius, of the Fabian voting-​tribe ... procurator of the province of Britain, Julia Pacata I[ndiana], daughter of Indus, his wife, had this built’). Source: Photograph © The Trustees of the British Museum.

103

5.6 Europa and the bull with two cupids in front of the presumed semicircular couch in the apse of the dining room at Lullingstone villa, Kent. The inscription (RIB 2448.6) reads: ‘Invida si ta[uri] vidisset Iuno natatus | Iustius Aeolias isset adusque domos’ (‘If jealous Juno had seen the swimming of the bull, more justly would she have gone to the palace of Aeolus’). Source: © Alamy.

106

List of Figures    xiii 9.1 Distribution of Roman military sites in Scotland. (a) Flavian (c. ad 78–86); (b) Antonine (c. ad 139–​165). Source: Reproduced by courtesy of Professor D. J. Breeze.

183

9.2 The Antonine military complex at Inveresk (East Lothian). Source: Drawing by Alan Braby.

187

9.3 Distribution of Roman finds from non-​Roman sites north of Hadrian’s Wall. (a) overall distribution; (b) middle Roman Iron Age c. ad 160–​250 (the sites of Birnie and Traprain Law, mentioned in the text, are marked: *represents find-spots of denarius hoards); (c) late Roman Iron Age c. ad 250–​400; *marks the major centres of Traprain Law, Edinburgh Castle, Eildon Hill, and Dumbarton Rock. Source: © Fraser Hunter.

191

10.1 Epigraphically recorded age of travellers to Rome 30–​600 ad as recorded by Noy (2000) and for the late antique west by Handley (2011). Source: after Handley (2011: graphs 1 and 2). © Department of Archaeology, University of Reading.

205

10.2 Strontium and oxygen isotope data of humans from Lankhills/​ Winchester, contrasting burials identified as ‘intrusive’ according to Clarke’s criteria (1979) with other burials from the site; the two boxes indicate the local strontium isotope range for Winchester and estimates for the range of oxygen isotope values consistent with a childhood in Britain based on human skeletal phosphate (δ18Op) data available in 2009 (thin line) and 2012 (thick line). Source: after Evans et al. (2006) and Eckardt et al. (2009). © Department of Archaeology, University of Reading. 212 10.3 Reconstruction of the so-​called ‘Ivory Bangle Lady’ from York. Source: © Aaron Watson.

215

11.1 Map showing areas of the Roman Empire with military units on Hadrian’s Wall. Source: © Rob Witcher, illustration by Christina Unwin.

229

11.2 Regina’s tombstone. Source: © Claire Nesbitt. 

239

12.1 Distribution of British brooches. Source: Brooches distribution partly after Morris (2010: 86, figure 4.35 and appendix 6); map by author. © Tatiana Ivleva.

250

12.2 Britons abroad: profession and status. Source: © Tatiana Ivleva. 

252

12.3 Distribution of the military diplomas (star), funerary (circle), and votive (diamond-​shape) inscriptions mentioning British emigrants. Source: Map by author. © Tatiana Ivleva.

253

13.1 Distribution of cremation burials in late Iron Age Britain. Source: after Fitzpatrick (1997). © T. Moore.

263

13.2 Comparison of late Iron Age cremation burials from (a) Welwyn Garden City, Hertfordshire, and (b) Clemency, Luxembourg. Source: © British Museum and Musée National d’Histoire et d’Art, Luxembourg.

265

xiv

xiv   List of Figures 13.3 Similarities in villa layouts from northern Gaul and Britain. Source: © T. Moore.

269

13.4 Approximate distribution of Romano-​Celtic temples. Source: after Derks (1998). © T. Moore.

271

16.1 Percentage of burials with grave goods with each age category at Colchester, Poundbury, and Lankhills. Source: © Alison Moore.

327

16.2 Percentage of total burials with grave goods at East London, Gloucester, and York. Source: © Alison Moore.

332

16.3 Percentage of aged burials with grave goods at rural sites in western and eastern regions of southern Roman Britain. Source: © Alison Moore. 

333

17.1 Bronze and ceramic vessels in grave 6260, the cremation burial of an adult (ad 5–​80), from a rural settlement at Tollgate, Springhead. The bronze jug, pan, and mixing bowl as well as two ceramic flagons were placed beneath a table. On this were piled terra nigra and terra rubra platters, beakers, cups, and local copies. The skull and forelimb of a pig were found close by, as was the cremated bone. One or more boards with glass gaming pieces and bone dice lay closer to the grave’s centre. Source: courtesy of Oxford Archaeology.

350

17.2 A bronze drinking vessel found at Llantilio Pertholey near Abergavenny with a handle in the form of a snarling leopard with inlaid silver and amber (PAS NMGW-​9A9D16). Fieldwork subsequent to its discovery by metal-​detecting revealed the original funerary context. Source: courtesy of National Museums and Galleries of Wales.

352

19.1 Comparative age of decapitated and non-​decapitated inhumations along the Fen Edge. Source: © B. Crerar. 

391

19.2 Comparative sex of decapitated and non-​decapitated inhumations along the Fen Edge. Source: © B. Crerar. 

392

19.3 Comparative grave-good allocation for decapitated and non-​decapitated inhumations along the Fen Edge. Source: © B. Crerar.

393

19.4 Comparative coffin allocation for decapitated and non-​decapitated inhumations along the Fen Edge. Source: © B. Crerar. 

394

19.5 Comparative posture of decapitated and non-​decapitated inhumations along the Fen Edge. Source: © B. Crerar. 

395

19.6 Guilden Morden cemetery, showing distribution of decapitated burials, adapted from Fox and Lethbridge (1926: figure 1) and Lethbridge (1934: site plan). Source: © B. Crerar.

398

20.1 Chronological and regional distributions of brooches. Note: The area and type labels are those given in Table 20.1 when not otherwise specified. Figures 20.1 (a-c) are Correspondence Analysis plots that show which regions have higher proportions of which types. Source: Data derived from Mackreth (2011). © Hilary Cool.

414

List of Figures    xv 25.1 Map showing quantitative evidence for the distribution of Severn Valley wares, in comparison with the core distribution of the western Iron Age coinage group. Note: Columns represent the percentage at a site; dashes represent sites without Severn Valley wares. Known kiln sites producing Severn Valley wares are shown. Both the contour map and the bar plots show that there is a very sharp fall-​off at the edge of the distribution of the fabrics. Source: © J. Evans.

512

25.2 Map showing the core areas of the seven regional coin series of Iron Age Britain, after Creighton (2000: figure A.1), compared with the distribution of Dressel 1 amphorae finds, after Fitzpatrick (1985), and the distribution of pre-​conquest Romanized toilet instruments (Hill 1997). Source: © J. Evans.

516

26.1 Map of Roman Britain showing selected sites with evidence for goldworking. Source: © David Dungworth.

533

26.2 Principal lead-​silver orefields and sites of primary silver extraction. Source: © David Dungworth. 

536

26.3 Zinc and lead content of Colchester-​type brooches (single piece construction and made of brass) and Colchester derivative brooches (multiple component construction and made of leaded bronze). Source: © David Dungworth. 

540

26.4 Silver content of denarii and zinc content of contemporary sestertii and dupondii (data from Walker 1976–​8; Dungworth 1996). Source: © David Dungworth.

541

26.5 Principal iron production sites. Source: © David Dungworth.

544

26.6 Dated Roman iron production sites in the Weald, after Cleere (1974). Source: © David Dungworth.

545

28.1 Kernel density plot of the distributions of ‘broad period Roman’ styli recorded by the Portable Antiquities Scheme as of 17 January 2014, plotted against selected constraints. Source: Robbins 2014: figure 4 for constraints, with kind permission of the author; background map data from Ordnance Survey/​EDINA supplied service, Crown Copyright/​database right 2013. Produced by L. Wallace.

580

28.2 Tomlin’s line drawing of Tab. Sulis 18. Source: Reproduced with the kind permission of the author. 

581

28.3 Tomlin’s line drawing of Tab. Sulis 30. Source: Reproduced with the kind permission of the author. 

583

28.4 Tomlin’s line drawing of the lead curse tablet at Red Hill, Ratcliffe-​on-​ Soar. Source: Hassall and Tomlin (1993: 311), reproduced with the kind permission of the author.

586

28.5 Latin and Ogam-​inscribed stone, St Dogmaels/​Llandudoch, Pembrokeshire, Wales, (a) photograph, (b) drawing and transcription.

xvi

xvi   List of Figures Source: © Crown copyright: Royal Commission on the Ancient and Historical Monuments of Wales.

589

29.1 Finds per site of anthropomorphic sculpture.

601

29.2 Zoning of sculpture quality.

605

30.1 Distribution of inscribed votive altars from Roman Britain. Source: after Millett (1995a: 110). © Amy Zoll.

623

30.2 Main concentrations of votive altars. Source: after Millett (1995a: 110), temples after Millett (1995a: 112); TOT rings after Daubney (2010: 112). © Amy Zoll.

625

30.3 Distribution of votive inscriptions dedicated to the gods Cocidius, Belatucadrus, and Vitiris/​Veteres, including double-​named variants. Source: after Zoll (1995a). © Amy Zoll.

631

32.1 Silver ring with early Christian symbol (anchor and fish) from the Roman fort at Binchester, County Durham. Source: © Durham University.

665

32.2 Potential Roman church from Silchester, Hampshire. Based on Frere (1975: figure 1). Source: © David Petts.

667

32.3 Painted wall plaster from Lullingstone Roman villa (Kent) showing chi-​rho symbol flanked by an alpha and omega. Source: © The Trustees of the British Museum. 

669

34.1 Magnetometer survey of part of the Roman landscape in the Vale of Pickering, North Yorkshire. Across the top of the image a trackway runs east-​west, flanked on either side by settlement enclosures and small fields. Beyond these to the south are the fainter traces of a series of fields bounded by ditches. Superimposed on these, in the northern zone, the back dots represent grubenhaüser (sunken-​featured buildings) of Early Medieval date. To the south the land rises towards the Yorkshire Wolds and the ancient landscape is obscured by deposits of wind-​blown sand. Source: Courtesy of Dominic Powlesland, Landscape Research Centre.

702

34.2 Plan of the Roman villa at Fishbourne, Sussex, in the later first century ad showing the approach from the east and associated structures. Source: Drawn by Lacey Wallace, after Cunliffe (1971); Cunliffe et al. (1996); Manley and Rudkin (2003).

705

34.3 Plans showing the development of the Roman villa and associated structures at Frocester Court, Gloucestershire. Source: Drawn by Lacey Wallace, based on Price (2000–​10).

707

34.4 Plan of the nucleated rural settlement at Higham Ferrers, Northamptonshire, in the late third–​fourth centuries ad. A series of buildings set within enclosures flank the Roman road, which runs along

List of Figures    xvii the valley of the river Nene just above the flood plain. Source: Illustration from Lawrence and Smith (2009), reproduced courtesy of Oxford Archaeology.

709

34.5 Photograph of a domestic building (Figure 34.4, No. 10810) within the nucleated rural settlement at Higham Ferrers, Northamptonshire. The Roman road and the river Nene are visible in the background. Source: Illustration from Lawrence and Smith (2009), reproduced courtesy of Oxford Archaeology.

710

34.6 Cut-​away reconstruction drawing of the third-​century aisled hall excavated at Shiptonthorpe, East Yorkshire. Source: Drawing by Mark Faulkner, from Millett (2006).

713

34.7 Aerial photograph of part of the Roman landscape at Burnby Lane, Hayton, East Yorkshire, looking north-​east. The valley of the Burnby Beck runs down the centre of the photograph, with its relict course showing as a light crop-​mark against the darker green of the flood plain. On either side of this damp ground can be seen a series of settlement enclosures. On the left a length of boundary ditch is also visible. This runs parallel with the stream, and formed one side of a droveway that linked the lowlands to the south-​west with the Wolds to the north-​east, allowing animals to be moved for seasonal grazing. Source: Photograph by Peter Halkon, courtesy of the Hayton Project.

716

35.1 Correspondence analysis highlighting differences in the composition of pottery assemblages (top) by vessel form (bottom) in the hinterland of Roman London and Colchester, c. 50 bc–​ad 250. Note: Details of abbreviations are outlined in Table 35.1. © Martin Pitts.

727

35.2 Pottery assemblages from Roman London, Essex, and Cambridgeshire plotted according to the percentage prevalence of lids versus jars, c. 50 bc–​ ad 250 Source: after Perring and Pitts (2013). © Martin Pitts.

729

35.3 Multidimensional scaling analysis of the prevalence of non-​work-related health conditions in late Roman Britain. Source: after Pitts and Griffin (2012). © Martin Pitts.

733

35.4 Comparison of the Gini coefficient of inequality with the mean number of furnishings per grave for selected Romano-​British cemeteries Source: after Pitts and Griffin (2012), with additions. © Martin Pitts.

736

36.1 Map of southern Britain with the main settlements mentioned in the chapter. Source: © A. C. Rogers. 

744

36.2 Plan of pits, shafts, and timber-fenced enclosures excavated within the centre of Roman Dorchester (Durnovaria) at the Greyhound Yard site, 1981–​4. Source: adapted from Woodward and Woodward (2004: figure 1). © A. C. Rogers.

753

xviii

xviii   List of Figures 36.3 Plan of the relationship between the fortress and later colonia at Colchester; the walls around the town were constructed in the early second century ad; the theatre and temple were constructed within the fortress annexe. Source: adapted from Crummy (1993: figure 2.9). © A. C. Rogers.

754

36.4 Plan of the relationship between the town wall circuit and enigmatic earthworks at Silchester (Calleva Atrebatum) possibly associated with the earlier oppidum here. Source: adapted from Fulford (1984: figure 85).

757

37.1 Plan of Caerwent forum, with location of probable curia marked. Source: Adapted from Brewer (2006: 39).

770

37.2 Plan of Silchester forum with known and approximate locations of inscriptions and statues indicated. Source: P. Copeland; adapted from Isserlin (1998: figure 9.1) with additions.

771

37.3 Plan of Canterbury. Source: (Millett 2007: figure 5.15). Reproduced by kind permission of the Kent History Project, Kent County Council, from plans provided by the Canterbury Archaeological Trust.

775

37.4 Plan of Wroxeter. Source: (White et al. 2013: figure 4.21).

777

39.1 (a) Charred spelt wheat grain, showing infestation by a grain weevil from a third/​fourth century corn drier at Grateley South, Hampshire; Photo: Gill Campbell; © English Heritage; reproduced by kind permission of English Heritage; (b) Charred remains of the granary weevil (Sitophilus granarius L.) from a deposit filled with charred ‘rubbings’ of malted spelt in an early/​mid Roman ditch at Northfleet villa, Kent ([D.] Smith 2011); Photo: David Smith; (c) Charred germinated spelt grains from a late Roman corn drier at Northfleet villa, Kent ([W.] Smith 2011: Plate 9); © High Speed 1 Ltd; image reproduced with the kind permission of High Speed 1 Ltd; (d) Charred detached spelt grain sprouts from a late Roman ditch at Northfleet villa, Kent ([W.] Smith 2011: Plate 10); © High Speed 1 Ltd; image reproduced with the kind permission of High Speed 1 Ltd; (e) Charred coriander; one of a cache of more than 1,000 coriander fruits found on the floor of a shop in Colchester that was burnt down during the Boudiccan Revolt of ad 6/61. The material was recovered as part of the excavations carried out at 45-​6 High Street (Murphy 1977); Photo: Gill Campbell; © English Heritage; image reproduced by kind permission of English Heritage; (f) Waterlogged olives from a latrine block at the rear of a first century tavern at 1 Poultry, London (Davis 2011: figure 39. 275); Photo: Andy Chopping; image reproduced by kind permission of Museum of London Archaeology.

811

39.2 Predominant mode of preservation for the three main categories of food and fibre crops, based on the total number of occurrences of each of these plants in archaeobotanical assemblages from Roman Britain. Source: after Van der Veen (2008); Van der Veen et al. (2008).

815

List of Figures    xix 39.3 Frequency patterns of selected foods, waterlogged records only. a) foods that initially increase but then decline; b) foods that increase over time; c) foods that decline over time. N = number of records with waterlogged remains. Source: after Van der Veen (2008); Van der Veen et al. (2008).

818

39.4 A near complete waterlogged imported pine cone (Pinus pinea) from a late third/​early fourth century ditch fill containing possible other votive objects at Clatterford Roman villa (McPhillips 2001). Source: © English Heritage; image reproduced by kind permission of English Heritage. 

824

40.1 Roman coins recorded by the Portable Antiquities Scheme. Source: © Philippa Walton. 

835

40.2 BERK-​65D307—a Roman Republican denarius issued in c. 207 bc found in Berkshire. Source: © Portable Antiquities Scheme.

837

40.3 The Frome hoard being excavated. Source: © Somerset County Museums Service.

842

40.4 The Coleraine Hoard (County Antrim) deposited in the first half of the fifth century ad. Source: © The Trustees of the British Museum.

845

41.1 Plan of the roadside settlement at Higham Ferrers. Source: (after Lawrence and Smith 2009: figure 2.18). © James Gerrard/​Andrew Agate.

854

41.2 Plan of the Roman villa at Turkdean. Source: (after Holbrook 2004: figure 4). © James Gerrard/​Andrew Agate.

857

41.3 The distribution of villas near Ilchester: 1) Dinnington, 2) Lopen, 3) Seavington St Mary, 4) Ilchester Mead, 5) Batemoor Barn, 6) Lufton, 7) West Coker, 8) East Coker, 9) Westlands. Source: © James Gerrard/​ Andrew Agate. 

858

xx

List of Tables

2.1 The educational networks of the contributors to this volume. 

34

4.1 Finds categories in Roman site reports up to the 1960s. © Ellen Swift.

66

11.1 Foreign units serving on Hadrian’s Wall and their provincial origins. Source: © Claire Nesbitt.

230

16.1 Distribution of grave goods with young/​prime adults and percentage of items related to personal appearance with female burials at four major urban cemeteries. Source: © Alison Moore.

328

16.2 Percentage of grave goods with mature/​older male and female burials at five urban cemeteries. Source: © Alison Moore.

330

19.1 Sites consulted. 

390

20.1 Brooches from England and Wales by region and broad date band Source: Mackreth (2011). © Hilary Cool.

412

21.1 Some significant and published case studies. 

426

31.1 Occurrence of selective votive objects in Southern Romano-​British cult sites. Source: After Smith (2001: 155, figure 5.13).

645

35.1 Codes used in Figures 35.1 and 35.2 

728

xxii

List of Contributors

Lindsay Allason-​Jones is a Visiting Fellow at University of Newcastle upon Tyne, Chair of the Marc Fitch Fund and the Hadrian Arts Trust Patricia Baker is a Senior Lecturer in Classical and Archaeological Studies at University of Kent Timothy Champion is Emeritus Professor of Archaeology at University of Southampton H. E. M. Cool is a Director of Barbican Research Associates Ltd Belinda Crerar is a Curator at the British Museum Ben Croxford is Historic Environment Officer for Merseyside David Dungworth is the Head of Archaeological Conservation and Technology at Historic England Hella Eckardt is Associate Professor in Roman archaeology at University of Reading Simon Esmonde Cleary is Professor of Roman Archaeology at University of Birmingham and has been a Visiting Professor at the University of Toulouse-​le Mirail Jeremy Evans is a freelance Roman pottery consultant and a Director of Barbican Research Associates Ltd Andrew Gardner is a Senior Lecturer in the archaeology of the Roman Empire at University College London and a Co-​Director of the Caerleon Excavations and of the West Dean Archaeological Project James Gerrard is a Lecturer in Roman Archaeology at University of Newcastle upon Tyne Rebecca Gowland is a Senior Lecturer in human bioarchaeology at the Department of Archaeology at University of Durham Ian Haynes is Professor of Archaeology at University of Newcastle upon Tyne Richard Hingley is Professor of Roman Archaeology at University of Durham Valerie M. Hope is a Senior Lecturer in the Department of Classical Studies at Open University Fraser Hunter is the Iron Age and Roman Curator at National Museums Scotland

xxiv

xxiv   List of contributors Henry Hurst is a Reader in Classical Archaeology (Emeritus) at University of Cambridge Tatiana Ivleva is the Postdoctoral Marie Curie Fellow at University of Newcastle upon Tyne Zena Kamash is Lecturer in Roman Archaeology and Art at Royal Holloway, University of London Mark Maltby is a Reader in Archaeology at University of Bournemouth Martin Millett is the Laurence Professor of Classical Archaeology at University of Cambridge Alison Moore is a Sessional Lecturer in Roman Archaeology and an Academic Editor Tom Moore is a Senior Lecturer in Archaeology at University of Durham Sam Moorhead is National Finds Adviser for Iron Age and Roman Coins, Department of Portable Antiquities and Treasure at the British Museum Gundula Müldner is a bioarchaeologist at University of Reading Alex Mullen is Assistant Professor in Classical Studies at University of Nottingham Claire Nesbitt is a Postdoctoral Research Associate in the Department of Archaeology at University of Durham John Pearce teaches Roman archaeology at King's College London David Petts is a Lecturer in Archaeology in the Department of Archaeology, University of Durham Martin Pitts is Senior Lecturer in the Department of Classics and Ancient History at University of Exeter Louise Revell is Associate Professor in Roman Studies at University of Southampton Adam Rogers is a Teaching Fellow at University of Leicester Melanie Sherratt is a Ph.D. candidate at University of Durham Alex Smith is a Senior Research Fellow at University of Reading Ellen Swift is a Senior Lecturer in Archaeology at University of Kent Marijke van der Veen is Professor of Archaeology at University of Leicester Lacey Wallace is a Postdoctoral Research Associate in Roman Archaeology in the Faculty of Classics, University of Cambridge Philippa Walton is a Research Fellow at Wolfson College, University of Oxford Jake Weekes is a Research Officer at the Canterbury Archaeological Trust

List of contributors    xxv Pete Wilson is an Archaeological Consultant and formerly Head of Research Policy (Roman Archaeology) for English Heritage Amy Zoll is a Social Sciences Information Technology Specialist and Ph.D. candidate in Anthropology at the University of Pennsylvania

xxvi

Introduction martin millett, louise revell, and alison moore

The potential reader of this book might be forgiven for an initial response along the lines of ‘why yet another book on Roman Britain?’ After all, as Sir Mortimer Wheeler wrote in discussing the subject: [W]e are slowly approaching a little more nearly to the mind of Roman Britain, for what that mind be worth-​and all human mentality is presumably worth something. We must not expect much of great consequence. Hadrian’s Wall, for example, has long ceased to matter as a major historical problem. … But for the major attainments of mankind the young scholar, thus trained, must now look to other periods, other lands. Haverfield’s work was not merely a monument, it was a tombstone. (Antiquity no. 138, 1961: 157–​8)

We would assert that there are two sound answers to this question. First, whilst there are many books on Roman Britain, most follow an approach that was first used in the 1930s with R.G. Collingwood’s contribution to the Oxford History of England, namely using the archaeological evidence within a chronologically organized narrative of historical events, complemented by chapters on topics such as art and religion. It is remarkable how enduring this framework has proved, not only in describing the subject but also in setting the agenda for how the subject should be approached right down to the radical reappraisal provided in David Mattingley’s An Imperial Possession (2006). In his review of Peter Salway’s volume in the Oxford History of England published in 1981 and replacing Collingwood’s 1937 volume, Richard Reece attacked this approach, memorably saying: Textbooks on Roman Britain to date make the subject appear like a nice sand-​pit in which toddlers can safely be left to play. I am thankful that it is a wild overgrown garden in which anything may happen. And I shall continue to try to prove it. (Archaeological Journal Vol. 139, 1982: 456).

This book takes its lead from Richard Reece’s critique, and seeks to provide a comprehensive review of Roman Britain from an entirely new perspective, that represented by younger scholars who have embraced new perspectives on the subject since the establishment of the Theoretical Roman Archaeology Conference (TRAC) in 1991. In this sense although about Roman Britain, the book offers a new approach that is very different from the many other volumes already on the shelf.

xxviii

xxviii   Introduction Second, for all students of the archaeology of the Roman Empire, it must be acknowledged that although Roman Britain itself was arguably marginal to the mainstream, it remains amongst the most fully explored and best documented of Rome’s provinces. There are very fully publications of primary data: inscriptions in the Roman Inscriptions of Britain (1965–​); sculpture in the Corpus Signorum Imperii Romani (1977–​); mosaics in D. Neal and S. Cosh’s Roman Mosaics of Britain (2002–​10); and place-​names in A.L.F. Rivet and C. Smith’s Place-​names of Roman Britain (1979). In addition there is a wealth of synthetic works, finds studies, maps, and site reports. These mean that the province is worthy of study because with its evidence we can explore ideas and interpretations more thoroughly than in other less well documented provinces. In that sense, by offering a Handbook which contains new approaches this book seeks to capitalize on such past work whilst offering thoughtful signposts for the future. In summary, given the dramatic increase in new scholarship over the last thirty years in the field of Roman Britain, we feel the time is ripe for a comprehensive review of the topic. This Handbook is organized around sections that reflect recent theoretical approaches to the material. In particular, we hope that it reflects the new questions and interpretations brought to the subject since the inauguration of TRAC. The sections are structured around the following themes: (i) the nature of the evidence; (ii) Britain in the Roman Empire; (iii) society and the individual; (iv) forms of knowledge; (v) landscape, settlement, and the economy; and (vi) materiality. The individual articles range from surveys to more focused discussions, and represent the methodological and theoretical breadth of the field. We are deliberately eschewing a chronological structure that has dominated past work on Roman Britain but it is of course important to bear in mind how the province originated and developed through time. The table below provides a summary of the textual evidence for the province, thus providing the basic evidence on which historical accounts are usually based. In many works on Roman Britain there is a tendency to embellish this primary record with historical inference derived from the study of finds, principally pottery and coins. This may of course be entirely reasonable in the context of a particular site narrative or regional study, but we are concerned that generalizing from such evidence to provide a broader chronological narrative provides a false sense of secure knowledge rather than acknowledging the fragmentary and mutable nature of the evidence. For this reason the framework provided below simply uses texts.

Timeline of Roman Britain 55–​54 bc Julius Caesar, engaged on his conquest of Gaul, made two military expeditions to Britain. The first in 55 bc was of limited success although it generated a useful propaganda advantage in Rome where Britain was known as a mysterious place on the other

Introduction   xxix side of the ‘Ocean’. The larger-​scale campaign in 54 bc was more successful and resulted in the imposition of treaties on the peoples of the south-​east, which marks the first stage of a projected Roman annexation. (Caesar, Gallic Wars, V)

54 bc–​ad 43 Events, first in Gaul then at Rome, prevented any capitalization on Caesar’s success. Diplomatic contacts between British tribes and Rome continued with the arrival in Rome of tribal leaders deposed in Britain under Augustus (27 bc–​ad 14) and Caligula (ad 37–​41). It is plausible that Rome considered some of the peoples of the south and east to have been client states. (Res Gestae, 32; Tacitus, Annals, II). Both Augustus and Caligula (in ad 40) are said to have considered invading Britain, although neither acted. (Dio Cassius, LIII; Tacitus, Annals, II)

54 bc–​ad 43 There was also commercial contact, with the export of products from Britain and imports from the empire, which seems to have been worthwhile for Rome (Strabo, IV). Rome influenced political evolution in Britain in the period down to AD 43 and this engendered internal conflicts. Although outside the territory of the empire, some of the states remained as clients of successive emperors and had close contacts with Rome. This was a period of rapid development and the growth of one dominant indigenous people, the Catuvellauni, whose capital lay at Camulodunon (Colchester). Their expansion threatened the adjacent areas, and deposed leaders appealed to Rome to help re-​ establish themselves.

ad 43 The last such leader was Berikos (Verica) who fled to the emperor Claudius leading the latter to intervene in Britain. Claudius, anxious to obtain prestige through military success, sent an invasion force of about 40,000 men. The force moved through southern England, and took the principal indigenous centre, Camulodunon (Colchester), under Claudius’ personal command. Claudius accepted the surrender of eleven British kings, and was honoured at Rome where he returned immediately. (Dio Cassius, LX; Suetonius, Claudius, 17; Inscription, CIL V no. 920; Dio Cassius, LX)

ad 43–​60 Following the defeat of the south-​eastern states, campaigning continued in the southwest under the command of the future emperor Vespasian, who won a series of victories.

xxx

xxx   Introduction A revolt occurred in East Anglia (in ad 47), but the expansion of the Roman occupation continued into Wales where it was slowed by the guerrilla tactics of the inhabitants, and by a revolt in the Pennines. Early in the reign of Nero (ad 54–​68) consideration was given to withdrawal, presumably because of slow military progress. (Suetonius, Vespasian, 4; Tacitus, Annals, XII; Suetonius, Nero, 18)

ad 58–​60 Campaigns were undertaken against the Druids on Anglesey. (Tacitus, Annals, XIV)

ad 60 East Anglia rose against the Romans under the leadership of Queen Boudica, with the destruction of the Roman towns of Colchester, London, and Verulamium. (Tacitus, Annals, XIV)

ad 61–​71 The crushing of the revolt was followed by a period of consolidation. (Tacitus, Agricola, XVI)

ad 71–​4 The allied kingdom of the Brigantes in the Pennines became divided internally, so Rome intervened and conquered the area. (Tacitus, Histories, III, 45 and Agricola, XVII)

ad 74–​7 Expansion of the Roman province continued with the subjugation of south Wales. (Tacitus, Agricola, XVII)

ad 77–​83 The new governor, Agricola, led a sustained series of campaigns which enlarged the province substantially. (Tacitus, Agricola, XVIII–​XXXVIII)

Introduction   xxxi

ad 80 Agricola completed the conquest of Wales and Anglesey, then he turned to the consolidation of northern Britain, before moving into Scotland where he reached the Tay in ad 80. He garrisoned the area up to the Forth-​Clyde isthmus, and dealt with south-​west Scotland, at the same time suggesting to Rome that Ireland would be easy to conquer.

ad 82 Agricola again advanced northwards up the eastern coastal plain, leading to the establishment of a garrison as far as the Tay.

ad 83 The great set-​piece battle of Mons Graupius, somewhere in north-​east Scotland, marked the successful completion of Agricola’s campaigns. At this stage Britain was also circumnavigated by Agricola’s fleet.

c. ad 87 Following Agricola’s departure, Roman military attention was turned to the Danube, and as a result troops from Britain were removed, with resultant withdrawal from Scotland. (Inscription, ILS no. 2719) The historical sources for the next thirty years are very poor but it is suggested, on the basis of the distribution of forts and dated pottery, that the army was slowly withdrawn from most of Scotland and the occupied zone contracted to the area based on a line from the Tyne to the Solway.

ad 122 Hadrian visited Britain following a military disturbance. The construction of Hadrian’s Wall, built ‘to separate the Romans from the barbarians’, followed shortly afterwards. (Scriptores Historiae Augustae (SHA), Hadrian, 5; 11)

ad 139–​42 The new emperor, Antoninus Pius, abandoned Hadrian’s Wall and advanced northwards, driving back the barbarians, before constructing a new wall between the Forth and the Clyde. (SHA, Antoninus Pius, 5)

xxxii

xxxii   Introduction

ad 138–​61 Antoninus Pius may have removed territory from the Brigantes and fought a war in the Genounian region although this reference may not concern Britain at all. (Pausanius, Description of Greece, VIII, 43) The evidence for the following period is ephemeral and ambiguous. The conventional interpretation that the Antonine Wall was abandoned in the mid-150s and then briefly reoccupied again until 163–​4 has been revised. It is now thought that the withdrawal back to Hadrian's Wall was piecemeal and took place through the period between 155 and 163–​4.

ad 181–​4 Tribes invaded across Hadrian’s Wall, but were repulsed. (Dio Cassius, LXXII)

ad 184–​5 The emperor Commodus took the title Britannicus, and issued coins to commemorate his victory in Britain. (Coins, RIC 437, 440, 451)

ad 185 A mutiny among the troops in Britain led to the downfall of Perennis, the praetorian prefect in Rome. As a result, Commodus appointed the future emperor Pertinax as Governor of Britain. (Dio Cassius, LXXII; SHA, Pertinax, 3)

ad 192–​96/​7 Commodus and then his successor Pertinax were assassinated, and in the struggle for the succession Clodius Albinus, the governor of Britain, was one of three contenders. For an initial period, Septimius Severus recognized him as his deputy (i.e. Caesar) while fighting against the third pretender, Pescennius Niger. Once he was defeated, Severus and Albinus engaged in a civil war, with Albinus taking troops from Britain with him. Severus won after a major battle outside Lyon. Following his victory Severus divided Britain into two provinces, probably to ensure that no single governor again had command of so large an army. (Dio Cassius, LXXIII; Herodian, II, 15; Dio Cassius, LXXV; Herodian, III, 8)

Introduction   xxxiii

ad 197–​205 The tribes of the Maeatae and Caledonians in Scotland broke their treaty with Rome and waged war in the North. Since Roman military concerns lay elsewhere, the Maeatae were bought off. (Dio Cassius, LXXV)

ad 208–​11 Renewed troubles in Britain provided the opportunity for the emperor Severus to come in person with his sons to campaign in Scotland. He defeated the Caledonians but this was followed by a revolt of the Maeatae, and a renewed guerrilla war by the Caledonians. While preparing for the campaign of 211 Severus died in York. He was succeeded by his sons, Geta and Caracalla, who terminated the expedition and withdrew from the territory gained. (Dio Cassius, LXXVI; Herodian, III, 14–​15; Dio Cassius, LXXVII; Herodian, III, 15) After 244 the structure of authority within the empire failed, and there was a period of about forty years during which emperors neither maintained power for long, nor ensured a peaceful succession. In the period from 259 to 274 Britain formed part of the Gallic Empire, which had effectively seceded from the remainder of the empire. Political stability was restored with the accession of Aurelian (27–​5) and consolidated by Diocletian (from 284), who shared power with Maximian from 286. Within Britain the events of the mid-​third century up to this period are little known.

ad 286–​9 Carausius was appointed to patrol the English Channel which was infiltrated by ‘barbarians’. His success enabled him to profit from the campaigns, and for exploiting this he was sentenced to death by the emperor. He responded by declaring himself emperor. His control of Britain was unchallenged until Maximian launched a naval attack which seems to have been repulsed. Carausius then occupied parts of northern Gaul. (Aurelius Victor, 39; Eutropius, IX. 21; Panegyric on Maximian, 11–​12)

ad 293 Constantius, Maximian’s deputy (i.e. Caesar), attacked and took Boulogne. Carausius was assassinated by one of his ministers, Allectus, who was declared emperor in Britain and northern Gaul. (Panegyric on Constantius Caesar, 6; Eutropius, IX, 22)

xxxiv

xxxiv   Introduction

ad 296 Constantius’ army attacked, landing on the south coast, and Allectus was killed. Constantius was accorded a triumphal welcome in London. (Panegyric on Constantius Caesar, 13–​20)

ad 305–​6 Constantius, now co-​emperor with Maximian, came to northern Britain to conduct a military campaign. He was joined by his son Constantine, and reached the far north of Scotland. While staying at York the emperor died, and Constantine was declared emperor. (Aurelius Victor, 40; Eutropius, X, 1–​2)

ad 312–​14 Coins show that the emperor Constantine was in Britain. In 315 he took the title Britannicus indicating that he had won a victory here. (Coins, RIC 133–​41, 142–​3, 144–​5)

ad 342–​3 A winter visit to Britain by the emperor Constans perhaps suggests a crisis here. (Libianus, Orationes, LIX)

ad 350–​3 A revolt in Gaul toppled Constans, who was replaced by Magnentius. The defeat of the rebellion was followed by reprisals in Britain. (Ammianus Marcellinus, XIV. 5)

ad 360 Troops were sent to Britain to deal with the Picts and Scots (although the motive may have been simply to remove the troops from Gaul). (Ammianus Marcellinus, XX. 1)

ad 364 Barbarian raids by Picts, Scots, and Saxons are recorded. (Ammianus Marcellinus, XXVI, 4)

Introduction   xxxv

ad 367–​8 A concerted barbarian attack occurred, with the loss of a senior military commander. Count Theodosius was sent to recover the situation and we are told that he restored the province after a major campaign. (Ammianus Marcellinus, XXVII, 8; XXVIII.3)

ad 383–​8 Magnus Maximus, a British army commander, led another revolt, killing the emperor Gratian. Maximus gained control of much of the western empire and ruled it from Trier. He moved against Italy and was defeated by Theodosius. (Zosimus. IV, 35; 37; Orosius, VII, 35)

ad 396–​8 The general Stilicho ordered an expedition against the barbarians in Britain. (Claudian, Stilicho, II, 247–​55)

ad 399 Peace was restored by Stilicho’s expedition. (Claudian, Eutropius, I, 391–​3)

ad 401–​2 Troops were withdrawn by Stilicho to defend Italy. (Claudian, Gothic War, 416–​18)

ad 406–​7 A usurper, Marcus, took power in Britain. Marcus was replaced first by Gratian then by Constantine III. Constantine was left to defend Gaul against the barbarians. (Zosimus, VI, 2; Orosius, VII, 40)

ad 408–​9 Britain is attacked by the Saxons. ‘The Britons freed themselves, expelling their Roman governors and setting up their own administration.’ (Zosimus, VI, 5)

xxxvi

xxxvi   Introduction

ad 410 The emperor Honorius replied to an appeal for help from Britain by telling the cities to look after their own defence. (There are serious doubts about whether this reference does refer to Britain: it is more likely to relate to Bruttium in southern Italy.) (Zosimus, VI, 10)

ad 429 St Germanus visited Britain to counter heresy in the church and defeated a Saxon raiding party. (Constantius, Vita)

ad 435–​7 St Germanus may have visited Britain for a second time. (Constantius, Vita) The accounts of events after the first decade of the fifth century are difficult to interpret. There were clearly continuing Saxon raids, which resulted in the loss of territory. By this stage events at Rome prevented any help being sent from outside Britain. It is likely that from shortly after 400 the government of the province had degenerated, with effective power moving into the hands of the landed aristocracy who acted as local barons. From the 440s land was lost to the Saxons piecemeal, as a result of the limited power of the local lords. Territory in Kent seems to have been lost first, but there remained a strong sub-​Roman enclave in the west and north which certainly lasted down to the early sixth century.

PA RT  I

NAT U R E OF T H E E V I DE N C E

2

Chapter 1

Early Studi e s i n Rom an Bri ta i n 1610 to 1906 Richard Hingley

Introduction: Images of Civilization and Barbarity For great part of four hundred years, the Romans occupied this island in a state of peace and tranquillity: and a colony so fertile, and abounding in beautiful situations, must have been inhabited by many Roman adventurers, who migrated hither with their families, and built villas or country seats, where they lived in some degree of opulence and elegance. Even the Britons of rank might have built houses in the Roman taste. Whenever we talk of the Romans in Britain, we think of nothing but rapine and hostility. (Thomas Warton 1783: 59)

This chapter explores the origins of the study of Roman Britain, addressing the period from the late sixteenth century to the early twentieth. It compares and contrasts the accounts of Britannia created by William Camden (1610) and Francis Haverfield (1906). It also reviews some of the variety of other ideas about Roman Britain that developed between these times, exploring the ways that discoveries of archaeological objects and sites were used to support and/​or transform a number of semi-​contradictory ideas about British origins (cf. Hingley 2008a, 2011, 2012a). This chapter commences with the publication of the first English edition of William Camden’s seminal work, Britannia (1610). Camden was the first author to provide a detailed account of the evidence for the Roman province of Britannia, information that was used to provide the background for his history of Britain up to his own time. It ends

4

4   Richard Hingley with the publication of Haverfield’s seminal lecture, The Romanization of Roman Britain, in 1906. This lecture was later published as a small book (Haverfield 1912) that was republished on a number of occasions (Freeman 2007). In 1907 Haverfield was elected to the Camden Professorship at Oxford, a post that commemorated William Camden (Freeman 2007: 164). Haverfield and Camden followed broadly comparable agendas but in very different historical contexts. Both authors pursued a common theme: addressing the introduction of ‘civility’ (Camden) or ‘civilization’ (Haverfield) to the native population of the British Isles conquered by the Romans. Despite the considerable difference in the details of the tales told by Camden and by Haverfield, the common element in the thematic structure of these two highly influential accounts drew directly upon the Roman writings that had emphasized the introduction of Roman ways to indigenous Britons under imperial rule. Of particular significance to both authors was the section of text in Tacitus’ Agricola (21) that described the training of the sons of British chiefs in Roman ways and their consequential enslavement (Hingley 2008a: 10). This theme has provided a complex myth that played a significant role in ideas about the origins of English civilization from the late sixteenth century to the present day, although the terms in which this debate have been conceived have by no means remained constant (cf. Hingley 2008b). This origin myth communicated directly with the classically educated landed elite living to the south of the Antonine Wall from the late sixteenth century and, as Norman Vance has observed, came to act as the foundation of Victorian British pride (Vance 1997: 265; cf. Hingley 2010). It helped to communicate the humble origins of contemporary British greatness and, from the late sixteenth century to the twentieth, provided ideological support for the conquest and control of foreign territories incorporated into the British Empire (Hingley 2000, 2008a). This was not by any means the only myth of origin that was drawn from the Roman past by the British (for myths of origin, see Samuel and Thompson 1990; Broklehurst and Phillips 2004; Hingley 2008a: 4). Another powerful image of Roman Britain that dominated in the nineteenth century was soundly dismissed by Haverfield when he observed that, in ‘Britain, as it has been described by the majority of writers, we have a province in which Romans and natives were as distinct as modern Englishman and Indian, and the “departure of the Romans” in the fifth century left the Britons almost as Celtic as their coming had found them’ (Haverfield 1906: 190; cf. Haverfield 1896: 428–​429). Haverfield noted that this inaccurate image had arisen as a result of both an over-​reliance by the Victorians on the writings of Caesar and Tacitus and also ‘the analogies of English rule in India’. I have defined this image elsewhere as that of the ‘Celtic subaltern and Roman officer’ (Hingley 2000: 10). In this context, the term ‘subaltern’ refers to representatives of a perceived inferior race subject to the hegemony of a ruling class (cf. Spivak 1994). For much of the period that separated Camden from Haverfield, this idea dominated the perception of the character of Roman Britain, leaving little room for antiquaries to explore the potential civilizing influence of the Romans on the ancient Britons. This image of the Celtic subaltern suggested that ancient Britons retained their barbarian, or semi-​barbarian, manners throughout the period of Roman rule in Britain;

Early Studies in Roman Britain: 1610 to 1906    5 it also suggested that Roman officers lived in their Roman stations (forts, towns, and villas) alongside, but at some remove from, native peoples. Many nineteenth-​century accounts of Roman Britain had drawn deeply upon the cultural analogy provided by British rule in India, just as British officers had drawn upon Roman parallels to inform their actions and policies (Hingley 2008a: 240–​241). Haverfield’s account of Roman Britain was intended to point out the bias in the British–​India analogy by documenting the progressive influence of Roman civilization on the peoples in the south of the province. Haverfield stressed the common factors that linked the Romanized people of Britain to populations across the Roman Empire, including urbanism, villas, forts, and Roman material culture. Haverfield’s synthetic account of the Roman province mapped ancient British civilization and subservience onto two different geographical areas of the province, the ‘military district’ and the ‘civil district’ (see Hingley 2000; Webster 2001). It elevated the importance of the civilized Romano-​British populations; stressing the ancestral introduction of civilization to the people of the south and east of the British Isles and emphasizing the ancient barbarity of the people of the north and west (cf. Hingley 2008b: 319–​321).

William Camden: Chorography and British Civility During the late sixteenth century, a growing appreciation of the value of surviving classical texts transformed earlier more directly mythical accounts of the early history of Britain. Central to this new thinking was William Camden’s fundamental contribution, the first synthetic account of Britain’s Roman past that drew deeply on recently rediscovered classical texts and material objects. The first edition of his influential volume, Britannia, was published in Latin in 1586. Subsequent editions published over the following two and a half decades updated and expanded this review of the surviving Roman relics across England, Wales, and southern Scotland. Britannia communicated information derived from local informants who recorded and illustrated ancient sites and objects. The first English edition of Britannia, published in 1610, contained a detailed account of the Roman province, including the artefacts that had been identified, Latin inscriptions, and a handful of pre-​Roman coins (Hingley 2008a: 26–​40). The 1610 edition, subtitled a Chorographicall Description, brought the evidence for the Roman province into a direct engagement with Camden’s Britain (Hingley 2012a: 8–​9). Chorography is an analytical concept originating in the ancient Mediterranean world and used by early modern scholars in their accounts of the landscapes of England. Howard Marchitello observed that chorography delineates ‘topography not exclusively as it exists in the present moment, but also as it has existed historically’, since the concept is based on the idea that the character of the land described in

6

6   Richard Hingley particular places persists through time (Marchitello 1997: 78, 55). By the early seventeenth century, chorography had a close relationship with growing notions of landed property. As a method, it drew upon the history of the past of selected locations to help to justify the local aristocracy’s lineages and rights to estates (Swann 2001: 101–​ 107). By connecting the modern kingdom of England with the civil zone of the Roman province, Camden constructed a longer ancestry for the civilization and religion of the Elizabethan and Jacobean English—​an idea that placed the Welsh, Scots, and Irish in a subservient position. For Camden and his peers, the Roman history of Britain had a particular relevance as an ancient context for the introduction of civility and Christianity to Britain. The concept of civility in turn derived from the Latin civilitas, meaning the art of government or the qualities of citizenship (Bryson 1998: 43–​58). I have discussed the concepts of civility and civilization/​Romanization in the writings of Camden and Haverfield elsewhere (Hingley 2008b). It also contributed to a developing Jacobean fixation with exploring the unity and disunity of the new Great Britain (Hingley 2008a: 53). British policies in Ireland and North America at this time were informed by ideas derived from these ancient sources, a developing knowledge that helped to conceptualize and justify colonial exploration by providing models for dominating ‘savages’ and ‘barbarians’ (Hingley 2008a: 60–​66). This fashion for looking to the Roman past for the origins of contemporary civility ceased to be popular during the troubled decades of the early seventeenth century, but the idea was reinvented in the works of the eighteenth-​century antiquaries such as the Revd William Stukeley (for ‘Augustan’ England, see Ayres 1997: p. xiv). Early eighteenth-​century society was dominated by a landed aristocracy that drew deeply on classical Roman models (Ayres 1997: 2–​47). Stukeley’s ideas about the prehistoric henges of southern Britain, including Stonehenge, are more well known today than his contribution to Roman studies, which was of equal significance during his own lifetime. Fascinated by the Roman remains of Britain, Stukeley travelled along the Roman roads of the south, producing a volume entitled Itinerarium Curiosum (Stukeley 1724; cf. Sweet 2004: 166). This itinerary provided an account of the towns and remains that lay along the routes, glorifying the surviving remains of Roman civilization across Britain. Occasionally Stukeley’s tales about the early origins of Britain communicated the idea of a continuity of civilization that served to link the classical provincial past with early eighteenth-​century Augustan England. Very little excavated evidence was available to Stukeley but, in his account of a place that he calls ‘Mantantonis’ (Chichester), he describes an inscription that had been found the previous year during the digging of a cellar in the town (Figure 1.1) (Stukeley 1724: 194). This was the Cogidubnus inscription, which had already been recognized by another influential antiquary, Robert Gale, to be of considerable significance. Gale used the inscription to argue that Cogidubnus was a Roman citizen of British origin, who was ‘Romanized’ and took the name of his benefactor, the emperor Claudius (Gale 1723: 393–​394). Stukeley built on Gale’s account by arguing that the name ‘Pudens’ in the final line of the inscription referred to a man mentioned in

Early Studies in Roman Britain: 1610 to 1906    7

Figure 1.1  The Cogidubnus inscription from Chichester, Sussex. Source: from Gale (1723). Reproduced by permission of Durham University Library.

Martial’s Epigrams (Volume 1: 238-​9). Martial’s Pudens was married to a British woman called Claudia Peregrina (Martial Epigrams, Volume 2: 277), whom Stukeley assumes to have been identical to Claudia Rufina. Stukeley drew upon and transformed earlier traditions to suggest that Claudia was Cogidubnus’ heir and also a Christian (Stukeley 1724: 193; cf. Hingley 2008a: 187–​188). Earlier versions of this legend had claimed that Claudia Rufina was the daughter of Caratacus and had become a Christian while in Rome with her father. Stukeley later suggested that Claudia and Pudens invited St Paul to visit them in Chichester during the mid-​first century ad and that he preached to the local population (Stukeley 1740: 233). The reconstruction of the Roman history of Britain inspired Stukeley to reflect on an ancestral civility that mirrored the growing imperial ambitions of the British elite (Haycock 2002: 119). Stukeley sought to project the Roman remains of Britain into his neo-​classical present, with the apparent grandeur of these material traces, in turn, reflecting on the contemporary greatness of Augustan England (Ayres 1997: 96–​97; for the complexities of the term neo-​classical, see Sachs 2010: 30). Stukeley had access to all the Roman materials studied by Camden, since Britannia was republished in an expanded form in 1722, but he also recorded a variety of new discoveries. Stukeley was involved in the recording and illustration of a number of Roman towns, and he also encouraged friends and associates to uncover and document the remains of several Roman villas, such as Cotterstock and Weldon in Northamptonshire (Hingley 2008a: 171–17​2). During the 1740s, a local man called John Stair conducted the first excavations at the Roman city at Silchester and produced a plan that included the town walls, street system, and central forum (Hingley 2008a: 181–18​4; 2012b).

8

8   Richard Hingley

Roman Settlers and ‘Celtic Subalterns’ These early excavations began to provide evidence for the civil elements of Roman Britain, including towns and country houses. Important discoveries were reported at the meetings of the Society of Antiquaries and in regular short papers on Roman matters in the Philosophical Transactions and the Gentleman’s Magazine. The idea of the civil province was to be developed further as a result of more extensive excavations of villas and towns conducted at the end of the eighteenth century and during the early nineteenth. The increasing appreciation of a substantial civilian settlement across the Roman province was, however, often accompanied by the idea that these were the homes, not of Romanized Britons, but of Romans who had settled in Britain from overseas. How did this idea of two separate populations come about? Until the pre-​Roman date of ‘Celtic’ metalwork came to be clearly demonstrated during the early nineteenth century, the main source for thinking about the ancient Britons was the classical authors who referred (mostly dismissively) to the semi-​naked and animal skin-​covered barbarians of ancient Britain (Smiles 1994). In addition, the sixth-​century ad writings of the monk Gildas projected the idea of subservient semi-​barbarian and semi-​naked Britons living alongside the Romans, who lost their valour and fighting spirit before succumbing to the invasion of the fifth-​century Anglo-​Saxons (Hingley 2012a: 170–​171). Of course, some contrary views idealized the ancient Britons, including the antiquarian works that tied Stonehenge and the megalithic monuments in with the Druids (cf. Smiles 1994: 77–​79; Morse 2005: 41–​47). This image remained highly popular from the sixteenth century to the early twentieth. Despite Tacitus’ comments about the civilizing (and enslavement) of the Britons in Agricola (21), it was usually felt that these people did not become particularly Romanized under Roman tuition. From the late sixteenth century, antiquaries began to collect and study Latin inscriptions derived from sites within the Roman military frontier zone, including the monuments now known as Hadrian’s Wall and the Antonine Wall. These monuments were also surveyed and mapped at this time (Hingley 2008a: 110–​114; 2012a). Studies of the inscriptions derived from Roman sites across northern England and southern Scotland indicated that individual soldiers had travelled to Britain from different regions of the Roman Empire. Latin inscriptions from sites in the south of the province, such as the Cogidubnus example, were far rarer, and the vast majority of named soldiers derived from overseas (Sweet 2004: 181–​183; Hingley 2008a: 160). The urban sites that developed in southern Britain in the Roman period often had a military origin, and the Latin inscriptions referring to these soldiers were often taken to indicate stations occupied by a military population of incomers. The excavation of Roman towns and villas gradually led to the interpretation of these buildings as elements of the infrastructure of ‘stations’ occupied by the Roman officers who had settled in Britain as a result of their imperial and military duties.

Early Studies in Roman Britain: 1610 to 1906    9 Caesar, Tacitus, and other classical writers had presented accounts of significant events across the province that fitted with the military emphasis provided by the Roman inscriptions. As a result, from the seventeenth century to the nineteenth, most antiquaries viewed the Roman occupants of Britain as settlers from overseas who had brought their Roman identity and culture with them. During the eighteenth century, these ideas coincided with the increasing militarization of British society that accompanied the expansion of colonial territories overseas (Hingley 2008a: 161). At this time, a significant number of antiquaries, including Stukeley, began to develop a fascination with the network of Roman roads and stations across Britain, partly recorded by the Latin itineraries (Hingley 2008a: 161–​163). Roman mosaics began to be uncovered and recorded in some numbers during the early eighteenth century, but were occasionally interpreted as pavements used to floor the tents of Roman generals (Hunter 1995: 196). More observant antiquaries realized that mosaics were usually associated with substantial buildings, probably villas (Hingley 2008a: 166–​167, 169–​173). The dominant explanation continued to suggest that these elaborate buildings represented the homes of Roman generals or Roman gentlemen from overseas. In 1787, Major Hayman Rooke published an account of the remains of a substantial Roman building that he had uncovered at Mansfield Woodhouse (Nottinghamshire). He suggested that the building indicated that ‘the manners of Italy’ had been introduced to Britain by Roman settlers, but did not discuss the idea that these Romanized individuals could possibly be Britons (Rooke 1787: 375; cf. Hingley 2008a: 235–​236). Samuel Lysons conducted a remarkable campaign of excavations at the sites of several Roman villas in southern England between 1789 and 1819. Uncovering the remains of several substantial buildings, he produced information that led to a reassessment of the character of Roman culture in Britain. For the first time, remains of the foundations of buildings comparable to the better-​preserved classical remains of Rome, Pompeii, and Herculaneum were excavated on a large scale in the British countryside (Hingley 2008a: 247–​253). Lysons inferred that the builders and occupiers of these villas were Roman settlers from overseas, although, in the case of the impressive villa at Bignor, the proximity of the remains to Chichester caused him to speculate that it might have been the home of Cogidubnus (Lysons 1815: 219) (Figure 1.2). During the nineteenth century, however, the concept that the Roman buildings in the cities and villas of Roman Britain were the homes of Roman officers continued to dominate. Impressive buildings were found at the Roman towns of Bath, Cirencester, and London during the early and mid-​nineteenth century, indicating the widespread scale of the Roman investment in Britain, but these urban centres were usually considered to represent military ‘stations’ with a civilian element to their populations (Hingley 2008a: 279–2​83). Until the late nineteenth century, it was not possible to locate the homes and possessions of the pre-​Roman peoples of Britain. The few coins with inscriptions, including abbreviated names of several pre-​Roman leaders referred to by classical authors, were the only pre-​Roman items that antiquaries could identify with any confidence prior to the nineteenth century (Hingley 2008a: 29–​30). As a result, the image of skin-​clad

10

10   Richard Hingley

Figure 1.2  Plan of a Roman villa excavated by Samuel Lysons at Bignor in Sussex. Source: Lysons (1815: plate 19). Reproduced by permission of Durham University Library.

semi-​naked barbarian Britons survived well into the twentieth century, when it was challenged by new information about the settlements and material possessions of ‘Iron Age’ people (cf. Hingley 2011). In many accounts of the eighteenth and nineteenth centuries, the Britons of the Roman period were described as members of a subservient population living alongside the Roman settlers. This was an idea that survived into the twentieth century in some scholarly and popular accounts of pre-​Roman and Roman Britain (Smiles 1994: 146; Hingley 2012a: 223–22​6). Attested evidence for named ancient Britons on Roman inscriptions was very rare, and, consequently, most evidence indicated overseas origins for the Romans of Britain.

Contrasting Views of Roman Britain Although many antiquaries considered that the Roman villas and towns had represented the homes of Roman settlers from overseas, antiquaries occasionally drew upon the idea of the civilizing power of Roman rule in Britain. Thomas Warton suggested that Britons of rank might have built houses in the Roman style (Warton 1783: 59). Warton emphasized, however, that Roman ‘adventurers’ were mainly responsible for the Roman

Early Studies in Roman Britain: 1610 to 1906    11 buildings of Britain. Sir Richard Colt Hoare excavated several ancient ‘British villages’ on the chalk downs of Wiltshire, going against the trends of the time by deliberately exploring these extensive earthwork sites rather than focusing on the excavation of villas and towns (Hingley 2008a: 255). Colt Hoare planned and partly excavated a number of Roman-​period sites with less impressive remains (Figure 1.3), concluding that it was ‘the wise policy of the Romans to civilize, as well as conquer … after having taking possession of the British settlements, both conquerors and conquered resided together; the former introducing many arts, comforts and luxuries of life … to which the Britons had been strangers’ (Hoare 1821: 127). The potential relevance of the message of the introduction of Christianity to Britain also caused some Victorian antiquaries and clerics to draw different messages from the Roman past (Hingley 2008a: 271–​278). Stukeley’s claims for Claudia, Pudens, and St Paul at Chichester were reinvented and elaborated by John William, the Archdeacon of Cardigan, while other authors imagined that Claudia was the Christian daughter of Caratacus or Cogidubnus (Williams 1848; see Vance 1997: 205–​206; Hingley 2008a: 271–27​5). The solicitor and antiquary Henry Coote, who had developed an interest in Roman Britain, was determined to find a Roman root for British Christianity by drawing on an approach that argued for continuity in the urban centres of the province from the Roman period to medieval England (Coote 1878). These perspectives tied into a developing image that supported the concept of the mixed ancestral origins for the contemporary population of Britain, including the ancient Britons, Romans, Anglo-​Saxons, and Normans (Hingley 2000: 91–9​3). Despite the works of antiquaries such as Colt Hoare and Henry Coote, it remained difficult to mount a sustained challenge to the idea that the Roman population of Britain constituted incomers from overseas. Thomas Wright (181–​77) published his popular but problematic book, The Celt, the Roman and the Saxon, in 1852. Wright (1852: 266–2​71) portrayed the population of Roman Britain as a collection of distinct races living in their individual Roman military ‘stations’ among a population of enslaved and downtrodden Britons. His writing indicates that he considered British peasants and slaves to be genetically incapable of modifying their ways to accommodate themselves to the civilized lifestyles of the occupying power. Wright described the cities, villas, and roads, by contrast, as being occupied by a series of semi-​independent Roman republics, each derived from a different part of the empire and with contrasting racial identities (Wright 1852; cf. Hingley 2008a: 279–2​83). Henry Mengden Scarth published a short book on Roman Britain in 1883 in which he desperately sought evidence for Christianity in the country’s early history (Scarth 1883). He struggled to find much archaeological support for the idea that Roman civilization was transferred to the Britons in any meaningful way, although he had a greater comprehension of the possibility that Britons might have become Romanized than many other contemporary accounts (Hingley 2008a: 293). For example, Bertram Windle (1897: 11) published an account of Roman Britain, in which he observed: The comparison has justly been made between the Roman occupation of Britain and our own occupation of India, for in both cases the intention of the conquering race

12

Figure 1.3  Knook Castle and British villages, Wiltshire. Source: Hoare (1810). Reproduced by permission of Durham University Library.

Early Studies in Roman Britain: 1610 to 1906    13 has been, whilst firmly holding the dominions of which they have become possessed, to interfere as little as possible with the natives so long as they were content to submit quietly to the demands of their conqueror.

One of the most influential works on Roman Britain at the turn of the twentieth century was Rudyard Kipling’s collection of tales Puck of Pook’s Hill (1906). This projected a view of the Roman officers in Roman Britain as settlers from overseas or their descendants. The work had a deep impact on the teaching of Roman Britain in schools and was recommended as a teaching aid to generations of teachers (Hingley 2012a: 220). Kipling’s writings indicate that the image of Celtic subalterns and Roman officers was still current at the beginning of the twentieth century. Nevertheless, academic scholarship was beginning to turn to a different explanation for Roman culture in southern Britain. Indeed, Haverfield’s contribution to this debate was probably prompted by the publication of Puck of Pook’s Hill (Rivet 1976: 14).

Romanization: Solving a Contradiction By the beginning of the twentieth century, archaeologists were developing a far more detailed picture of the military works, towns, and villas of Roman Britain, Gaul, and Germany as a result of a substantial number of new excavations. Knowledge of pre-​ Roman culture was also improving. This accumulating information would gradually lead to a new view of Roman culture as a transformer of indigenous ways of life. Francis Haverfield is usually seen as largely responsible for the coherent new idea of Romanization that arose during the early twentieth century. Haverfield drew deeply upon the improving knowledge of Roman Britain that had resulted from a number of recent archaeological projects, including excavations on Hadrian’s Wall, at Aylesford, Cranborne Chase, and Silchester. He also drew upon the scholarship of the German ancient historian Theodor Mommsen, who had outlined an approach to Romanization developed by Haverfield in his own work (Hingley 2008a: 317–​318). In his article of 1906, Haverfield defined the way that Romanization was thought to have operated (cf. Hingley 2000: 114–​123). He argued that the Roman Empire became fully Romanized and that ‘the definite and coherent civilization of Italy took hold of uncivilized but intelligent men, while the tolerance of Rome, which coerced no one into conformity, made its culture the more attractive’ (Haverfield 1906: 188). Discussing the spread of Roman architecture and culture to the western parts of the empire, he argued that: ‘In material culture the Romanization advanced … quickly. One uniform fashion spread from Italy throughout central and western Europe, driving out native art and substituting a conventionalized copy of Italian art’ (Haverfield 1906: 188). This new knowledge enabled Haverfield to provide a well-​informed interpretation of the Romanization of the

14

14   Richard Hingley indigenous inhabitants of the southern and eastern parts of Roman Britain, which he called the ‘civil district’ (Haverfield 1906: 191–19​4). Although Haverfield was the first to apply Romanization in a sustained way to the archaeology of Roman Britain, he was not the first to think about the distinctions between the south and north of the province. The work of the antiquary John Collingwood Bruce had encouraged generations of northern English antiquaries to focus on the evidence for Hadrian’s Wall (Hingley 2012a: 200). Bruce drew a distinction between the Roman ‘camps’ and walls of the north of the province and the ‘cities’ of the south (Bruce 1860: 343). He also remarked on the ‘comparative security and luxury of those who were fortunate enough to live in the south’ and noted that no mosaic floors had been discovered in the three northernmost counties of England (Bruce 1860: 344). These observations were developed by Haverfield in the definition that he provided of the military and civil districts (Haverfield 1906: 192) (Figure 1.4). During the second half of the nineteenth century, excavations in London, Verulamium, Cirencester, and Silchester indicated that the Roman towns of southern Britain had complex sequences of deep and sustained occupation (Hingley 2008a: 279–​ 293; Hingley 2012b; cf. Hoselitz 2007: 173–17​4). It was realized that the military inscriptions, used by previous generations to provide evidence for the overseas origin of many ‘Romans’ in Britain, dated to the early stages of the military occupation of many sites, allowing a renewed emphasis upon a potentially indigenous contribution at these locations. Research on Hadrian’s Wall and the Antonine Wall indicated that the regular ‘stations’ along the lines of these two frontier works represented military forts rather than civilian or partially militarized cities (Hingley 2012a: 196–​199). Initially, the central building in the Roman fort at Chesters on Hadrian’s Wall was interpreted as the forum of a very small classical city (Figure 1.5). During the final years of the nineteenth century, however, it was argued that buildings of this type actually represented the headquarters buildings of the Roman forts that occurred at frequent intervals along Hadrian’s Wall. The excavation of forts in northern England, Scotland, and on the German Limes during the final years of the nineteenth and the early twentieth centuries led to an improved understanding of the military organization of the northern parts of Roman Britain (Hingley 2012a). In the south, the extensive excavations at the Roman town of Silchester were to prove equally vital to the changing new perspectives outlined by Haverfield. Extensive excavations were undertaken at the site from 1864 to 1878 by James Gerald Joyce and from 1890 to 1909 by the Society of Antiquaries. This work led to a gradually evolving knowledge of the archaeology of this Roman city (Hingley 2008a: 287–​289, 302–​6), providing very little evidence to support the idea of a Roman military population at any period of its history. Observing the irregular plan of the town and the variation of the house plans from those of Roman Italy, Haverfield (1894) proposed that Silchester represented ‘a native copy of a Roman town, such as occurs in countries ruled by a nation of higher civilization than the subject race’. Using this new approach to the Romanization of southern Britain, many urban centres were reinterpreted as civil centres of local self-​rule, the outcome of the transformation of pre-​Roman tribes into Roman civitates. Early military occupation at some of these sites was explained as

Isurium EBURACUM LEG VI and Col

Buxton DEVA LEG XX

Lindum Col. Leic

Viroconium

Glos.Col

ISCA LEC

Ciren. Bath

Camulodunum Verulam Col. Mun LONDINIUM Silch. Winch. Chich.

Exeter

Bltiens

HADRIANS WALL

EBURACUM LEG VI and Col.

Lindum Col

ON SHO

R

E

DEVA LEC XX Viroconium

X

Camulodunum

A

ISCA LEG

S Othona

Anderida SAX

H

OR

E

LONDINIUM

ON

S

Figure 1.4  The civil and military districts. Source: Haverfield (1912: figure 1). Reproduced by permission of Durham University Library.

16

16   Richard Hingley

Figure 1.5  The Roman site at Chesters (Cilurnum) showing the excavated areas and ‘forum’. Source: Budge (1903: opposite p. 98). Reproduced by permission of Durham University Library.

conquest period activity, the military units later moving further north and west to establish and occupy the military zone. The concept of Romanized Britons was also becoming popular outside the urban arena. In an important publication of the ‘Late Celtic’ pottery from the Aylesford cemetery in Kent, the archaeologist Arthur John Evans argued that the pre-​Roman ceramics showed ‘Romanizing influences’—​a phenomenon that was also evident in the contemporary coinage (Figure 1.6) (Evans 1890: 351, n.c. 356, 383). Evans and Haverfield

Early Studies in Roman Britain: 1610 to 1906    17

Figure 1.6  Sketch plan of the graves forming a ‘family circle’ at Aylesford, Kent. Source: Evans (1890: figure 4). Reproduced by permission of Durham University Library.

were both influenced by the excavations undertaken by Pitt-Rivers when he explored the homes of people he called ‘Romanised Britons’ on Cranborne Chase (Pitt-Rivers 1888: 65; cf. Hingley 2008a: 298). Pitt-Rivers’s innovative fieldwork drew upon Hoare’s earlier studies and was, in turn, influential in identifying the adoption of Roman pottery and personal ornaments by Britons who did not appear from the excavated settlement evidence to have had a particularly elevated social status (see also Swift, this volume, for further discussion on the development of Romano-​British artefact studies). The concept of Romanization was not used in Britain prior to the final decades of the nineteenth century, although authors since the early seventeenth century had sometimes adopted the term ‘Romanized’ (Hingley 2008b). I have argued that the concept of Romanization focused attention onto a much more directly evolutionary interpretation of Roman identity and cultural change that emphasized progress (Hingley 2000). Haverfield explored this idea in detail in 1906 to provide an account of how indigenous peoples in the lowland civil areas of Roman Britain could gradually adopt Roman ways. The south of the province was then thought to have become fairly fully Romanized, with even the peasants adopting Roman styles of pottery, artefacts, and building (Haverfield 1906: 198). The elite were seen as administrators of the towns who lived in the villas excavated across the lowlands. In the military zone, which covered much of Wales and central Britain (northern England and southern Scotland), evidence for Roman culture was generally found only

18

18   Richard Hingley on Roman military sites (Haverfield 1906: 191–192). In this region, villas were a very rare occurrence, and towns were scarce. Where urban centres did occur, evidence was found for continued military associations. The homes of the indigenous people in the military districts usually appeared to change relatively little as a result of Roman control. Excavation work uncovered round-houses, with Roman pottery and material culture rarely occurring. Consequently, the idea of Celtic subaltern and Roman officer continues to be popular for the military areas of Roman Britain (Hingley 2004).

Conclusion: Ancestral Tales The material remains that were available when Camden, Stukeley, and Haverfield wrote their accounts impacted deeply on their interpretations of Roman Britain, but the meaning of the Roman past had also been transformed by historical circumstances. Although the remains of Roman sites had been uncovered when Camden was writing, the techniques of excavation and site recording were generally unknown, and there was very little comprehension of the variety of Roman site types across the Roman Empire. The only material objects that enabled Camden to tell stories about pre-​Roman and Roman Britain were the Latin inscriptions and coins that had been found over the centuries. Haverfield had access to an additional three centuries of information from the surveying and excavation of archaeological sites and the study of inscriptions and classical texts. The careful excavation that had been undertaken at Roman sites in Britain and across Europe had identified a variety of different types of site, including towns, forts, villas, temples, industrial sites, and rural settlements. Also available by this time were the initial results of work on the chronology of ‘Late Celtic’ (late Iron Age) and Roman pottery. My previous work has explored the ways that ideas outlined by Camden, Stukeley, Haverfield, and others were used to define the national and imperial origin of Britain and the British (Hingley 2000, 2008a, 2011). This chapter has adopted a different approach by suggesting that Haverfield helped to establish an intellectually coherent and well-​informed account of the archaeology of Roman Britain, a body of work that challenged earlier understandings. This is why his work had so much impact upon later archaeologists (for further discussion of the development of Romano-​British archaeology after Haverfield, see Millett, this volume). Consequentially, the concept of Romanization remained popular at least until the 1990s; indeed, ideas about the progressive character of Roman cultural change remain influential in the second decade of the twenty-​first century. The establishment of the model for the civil and military districts also had a major impact on twentieth-​century archaeology (Hingley 2004; although see James 2001). Many of the archaeologists working on Roman Britain have now rejected the inherently progressive interpretations propounded by Romanization theory. It is important, however, to see Haverfield’s work in the context of its time. He was clearly reacting to

Early Studies in Roman Britain: 1610 to 1906    19 the influential origin myth that drew an analogy between British India and the Roman rule of Britain. Haverfield’s perspective built upon, contradicted, and transformed concepts of Roman Britain, developing and giving new attention to an interpretation that had originated with classical writings and had been worked on during the late sixteenth century by William Camden. Haverfield set an agenda that has only recently been challenged by authors who have produced accounts of Roman Britain that adopt a range of different perspectives (including James and Millett 2001, Webster 2001, and Mattingly 2006).

Acknowledgements I am very grateful to the Arts and Humanities Research Council and to Durham University for the research leave that initially enabled me to develop these arguments (see Hingley 2008a: p. vii). I am also grateful to the editors of this volume for the invitation to write this chapter and to Christina Unwin for help with editing and with the illustrations. Hingley (2008a) includes a fuller account of the topics covered in this chapter. The archaeology of Hadrian’s Wall is not covered in detail here, but is addressed in Hingley (2012a). For other writings about the antiquarian and archaeological study of Roman Britain during the period covered in this chapter, see Hingley (2000), Sweet (2004: 155–188), Freeman (2007), and Hoselitz (2007).

References Primary Sources Martial, Epigrams, Volume I, trans. D. R. Shakleton Bailey. Loeb Classical Library. Cambridge, MA: Harvard University Press, 1993. Martial, Epigrams, Volume III, trans. D. R. Shakleton Bailey. Loeb Classical Library. Cambridge, MA: Harvard University Press, 1993. Tacitus, De Vita Agricolae, ed. R. G. Ogilvie and I. Richmond. Oxford: Clarendon Press, 1976.

Secondary Sources Ayres, P. (1997). Classical Culture and the Idea of Rome in Eighteenth Century England. Cambridge: Cambridge University Press. Broklehurst, H., and Phillips, R. (2004). History, Nationhood and the Question of Britain. London: Palgrave Macmillan. Bruce, J. C. (1860). ‘ … observations … associated … with the proceedings of an archaeological week’, Archaeological Journal, 17: 342–​347. Bryson, A. (1998). From Courtesy to Civility:  Changing Codes of Conduct in Early Modern England. Oxford: Oxford University Press. Budge, E. A.  W. (1903). An Account of the Roman Antiquities Preserved in the Museum at Chesters, Northumberland. London: Gilbert and Rivington.

20

20   Richard Hingley Camden, W. (1610). Britannia, or a Chorographicall Description of the Most Flourishing Kingdomes, England, Scotland, and Ireland., trans. into English by Philémon Holland. London: Georgii Bishop and Ioannis Norton. Coote, H. C. (1878). The Romans of Britain. London: Frederic Norgate. Evans, A. J. (1890). ‘On a Late-​Celtic Urn-​Field at Aylesford, Kent, and on the Gaulish, Illyro-​ Italic and Classical Connexions of the Forms of Pottery and Bronze-​Work there Discovered’, Archaeologia, 52: 315–​388. Freeman, P. W. M. (2007). The Best Training-​Ground for Archaeologists: Francis Haverfield and the Invention of Romano-​British Archaeology. Oxford: Oxbow Books. Gale, R. (1723). ‘An Account of a Roman Inscription Found at Chicester’, Philosophical Transactions, 32: 391–​400. Haverfield, F. (1894). ‘The Foundation of Silchester’, Athenaeum, 3503 (15 December): 836. Haverfield, F. (1896). ‘Early British Christianity’, English Historical Review, 43: 417–​430. Haverfield, F. (1906). ‘The Romanization of Roman Britain’, Proceedings of the British Academy, 2: 185–​217. Haverfield, F. (1912). The Romanization of Roman Britain: New Edition, Revised and Enlarged. Oxford: Clarendon Press. Haycock, D. B. (2002). William Stukeley:  Science, Religion and Archaeology in Eighteenth Century England. Woodbridge: Boydell Press. Hingley, R. (2000). Roman Officers and English Gentlemen: The Imperial Origins of Roman Archaeology. London: Routledge. Hingley, R. (2004). ‘Rural Settlement in Northern Britain’, in M. Todd (ed.), A Companion to Roman Britain. Oxford: Blackwell, 327–​348. Hingley, R. (2005). Globalizing Roman Culture: Unity, Diversity and Empire. London: Routledge. Hingley, R. (2008a). The Recovery of Roman Britain 1586–​ 1906:  ‘A Colony so Fertile’. Oxford: Oxford University Press. Hingley, R. (2008b). ‘Not so Romanized? Tradition, Reinvention or Discovery in the Study of Roman Britain’, World Archaeology, 40: 428–​444. Hingley, R. (2010). ‘“The Most Ancient Boundary between England and Scotland”: Genealogies of the Roman Walls’, Classical Reception Journal, 2/​1: 25–​43. Hingley, R. (2011). ‘Iron Age Knowledge: Pre-​Roman Peoples and Myths of Origin’, in T. Moore and X.-​L. Armada (eds), Atlantic Europe in the First Millennium bc: Crossing the Divide. Oxford: Oxford University Press, 617–​637. Hingley, R. (2012a). Hadrian’s Wall: A Life. Oxford: Oxford University Press. Hingley, R. (2012b). ‘“A place that the lover of antiquity will visit with great delight”: From Caer Segonte to Calleva Atrebatum’, in M. Fulford (ed.), Silchester and the Study of Romano-​ British Urbanism. Journal of Roman Archaeology Supplementary Series 90. Portsmouth, RI: Journal of Roman Archaeology, 23–​40. Hoare, R. C. (1810). The History of Ancient Wiltshire: Part I. London: William Miller. Hoare, R. C. (1821). The Ancient History of Wiltshire: Volume II. London: Lackington, Hughes, Harding, Mavor and Lepard. Hoselitz, V. (2007). Imagining Roman Britain:  Victorian Responses to a Roman Past. Woodbridge: Boydell and Brewer. Hunter, M. (1995). Science and the Shape of Orthodoxy: Intellectual Change in Late Seventeeth-​ Century Britain. Woodbridge: Boydell Press. James, S. (2001). ‘Romanization and the Peoples of Britain’, in S. Keay and N. Terrenato (eds), Italy and the West: Comparative Issues in Romanization. Oxford: Oxbow Books, 187–​208.

Early Studies in Roman Britain: 1610 to 1906    21 James, S., and Millett, M. (2001) (eds). Britons and Romans:  Advancing an Archaeological Agenda. York: Council for British Archaeology. Kipling, R. (1906). Puck of Pook’s Hill. London: Macmillan and Co. Lysons, S. (1815). ‘Account of the Remains of a Roman Villa, Discovered at Bignor, in Sussex, in the Years 1811, 1812, 1813, 1814 and 1815’, Archaeologia, 18 (1817), 203–​221. Marchitello, H. (1997). Narrative and Meaning in Early Modern England: Browne’s Skull and Other Histories. Cambridge: Cambridge University Press. Mattingly, D. (2006). An Imperial Possession:  Britain in the Roman Empire. London: Allen Lane. Morse, M. (2005). How the Celts Came to Britain:  Druids, Ancient Skulls and the Birth of Archaeology. Stroud, Tempus. Pitt-​Rivers, A. L. F. L. F. (1888). Excavations in Cranborne Chase, near Rushmore, on the Borders of Dorset and Wilts, 1880–​1888: Volume 2. London: Harrison and Sons. Rivet, A. L. F. (1976). Rudyard Kipling’s Roman Britain: Fact and Fiction. Keele: Keele University Library. Rooke, H. (1787). ‘An Account of the Remains of Two Roman villae Discovered near Mansfield Woodhouse, in May and October, 1786’, Archaeologia, 8: 363–​376. Sachs, J. (2010). Romantic Antiquity:  Rome in the British Imagination. Oxford:  Oxford University Press. Samuel, R., and Thompson, P. (1990). The Myths We Live By. London: Routledge. Scarth, H. M. (1883). Early Britain: Roman Britain. London: Society for Promoting Christian Knowledge. Smiles, S. (1994). The Image of Antiquity:  Ancient Britain and the Romantic Imagination. London: Yale University Press. Spivak, G. C. (1994). ‘Can the Subaltern Speak?’, in P. Williams and L. Chrisman (eds), Colonial Discourse and Post-​Colonial Theory: A Reader. London: Harvester, 66–​111. Stukeley, W. (1724). Itinerarium Curiosum. Or, an Account of the Antiquities and Remarkable Curiositys in Nature or Art, Observed in Travels throu’ Great Britain. London: Printed for the author. Stukeley, W. (1740). ‘30 Aug., 1740 [diary entry]’, in W. C. Lukis (ed.), The Family Memoirs of the Rev. William Stukeley MD and the Antiquarian and Other Correspondences of William Stukeley and Roger and Samuel Gale. London: Surtees Society, 232–​233. Swann, M. (2001). Curiosities and Texts: The Culture of Collecting in Early Modern England. Philidelphia: University of Pennsylvania Press. Sweet, R. (2004). Antiquaries:  The Discovery of the Past in Eighteenth Century Britain. London: Hambledon and London. Vance, N. (1997). The Victorians and Ancient Rome. Oxford: Blackwell. Warton, T. (1783). Specimen of a History of Oxfordshire: The Second Edition. Oxford: J. Nichols, J. Robson, C. Dilly. Webster, J. (2001). ‘Creolizing the Roman Provinces’, American Journal of Archaeology, 105: 209–​225. Williams, J. (1848). Claudia and Pudens: An Attempt to Show that Claudia, Mentioned in St Paul’s Second Epistle to Timothy, was a British Princess. London: Privately printed. Windle, B. C. A. (1897). Life in Early Britain. London: David Nutt. Wright, T. (1852). The Celt, the Roman, and the Saxon: A History of the Early Inhabitants of Britain, down to the Conversion of the Anglo-​Saxons to Christianity. London: Arthur Hall Virtue and Sons.

22

Chapter 2

Roman Brita i n si nc e Haverf i e l d Martin Millett

Introduction There has been a considerable amount written in ten years concerning the history of the archaeology of Roman Britain, with a principal focus on the historical context of the ‘Romanization’ and its relationship with Britain’s own imperial past. Most interest has focused on the later nineteenth and earlier twentieth centuries, in particular the role of Francis Haverfield, who can be seen in many senses as the father of the modern discipline (Hingley 2000; Freeman 2007). Rather less attention has been paid to the development of the subject since the First World War, or to the trends in thought and the organization of research that led to the enormous growth of the discipline in the later twentieth century (see, however, Jones 1987; Freeman 2007: 581–​605; Fulford 2007). In this chapter I would like to explore two complementary issues in order to provide some historical context for the rest of the volume. First, who led research and how they were they networked? Although there is some danger this just provides a catalogue of ‘great men’ (and this is too often gendered), it would be a mistake to underestimate the importance of personal influence, teacher–​pupil links, and fieldwork-​based friendships in the development of the subject. Second, how has the subject changed over the last 100 years or so and how does this relate to a broader historical context?

Early Twentieth-​C entury Foundations Although there is a long history of antiquarian interest in Roman Britain, the origins of the present subject lie very much in the second half of the nineteenth century. Two key trends can be identified: first, an increase in the amount of evidence uncovered as

Roman Britain since Haverfield    23 a result of industrialization and the infrastructural expansion of the Victorian period. Second, the growth of a general interest in archaeology, combined with the accumulation of wealth by some that led to a dynamic period in archaeological exploration, with sites such as Silchester, Caerwent, Corbridge, and Wroxeter becoming the subject of substantial excavations, which were aimed at enhancing understanding of the Roman period. Although these excavations were methodologically crude by modern standards, the data accumulated have been of long-​term significance in feeding further interest and subsequent research. Equally, the Victorian and Edwardian impulse towards grand projects also led to the first systematic attempts at the collation of the archaeological evidence—​for instance, in the Victoria County Histories. This gathering-​together of evidence forms the foundation for modern understanding of Roman Britain. In the earlier twentieth century there were comparatively few people active in research on Roman Britain, and their connections, both personal and educational, are comparatively straightforward. Thus, as Camden Professor of Ancient History at Oxford, Francis Haverfield knew all those engaged in the subject—​some of whom had been students of his at Oxford—​although, of course, archaeology was not then a subject taught in its own right, so they were generally classicists or historians. Haverfield also utilized a strong network of local contacts, as reflected especially through his authorship of a periodic series of reviews of work on Roman Britain and his later annual reviews of fieldwork published by the British Academy in 1913 and 1914. The latter were predecessors to the annual survey published in the Journal of Roman Studies from 1921, which was funded by the Haverfield Bequest. The large-​ scale excavations undertaken in this period included those run at Corbridge with input from Haverfield (Bishop 1994). Elsewhere, the enormous project to excavate the whole of Silchester was organized by the Society of Antiquaries of London, which published the results in its journal Archaeologia. It also initiated publication of its Research Reports series in 1913, to publish the excavations at Wroxeter. These volumes provide the first monograph series devoted to archaeology in Britain and were largely devoted to the Roman period in their earlier years, publishing reports on a series of key projects. There was also important work done on sites in Scotland at this period, most notably the large-​scale excavations at Newstead, which were magnificently published (Curle 1911). George MacDonald, who went on to be a leading figure in the subject after the First World War, wrote The Roman Wall in Scotland (1911, revised 1934), providing the first systematic synthesis of the archaeology of the Roman frontier in Scotland. Haverfield’s own most enduring written contribution was his Romanization of Roman Britain, first delivered as a lecture to the (newly founded) British Academy and published in its Proceedings in 1906, subsequently appearing as a short book in 1912 (see also Hingley, this volume). This provided a comprehensive and accessible explanation for the incorporation of Britain into the Roman world, based on ideas about cultural change in the indigenous peoples rather than the implantation of Roman colonists. Although very heavily influenced by contemporary views of ‘civilization’, Haverfield’s ideas influenced the research agenda and academic debate for much of the following century.

24

24   Martin Millett The period leading up to the First World War was one of considerable investment in research on Roman sites and shows a dynamism that is difficult now to appreciate fully because of the way in which leaders of the upcoming generation, like G. L. Cheesman, were killed in the conflict of 1914–​18. The tragedy of the loss of the leaders of that generation was said to have broken Haverfield, who himself died in 1919 (Haverfield and MacDonald 1924: 31–​32; D. B. Webster 1991: 120; cf. Freeman 2007, 348–​424). We are left to speculate how the subject might have been different had some of his pupils survived to build on his intellectual foundations.

The Interwar Years The period after 1918 saw two key trends in the study of Roman Britain. First, despite the diminution of the enormous disposable wealth of the Edwardian era, there was a gradual acceleration in archaeological field research, characterized by a sequence of key excavations designed to elucidate the history of the province. These research-​based excavations were important in advancing field methods as well as providing an increasing resource for historical synthesis. Equally, we may note that the subject became more diversified, with dynamic research taking place in a number of different regions. The post-​First World War generation was small in number but also closely connected, with many also Oxford-​linked. The leading intellectual figure was undoubtedly R. G. Collingwood, an Oxford student of Haverfield’s, who, while professionally an academic philosopher (becoming Waynfleet Professor of Philosophy at Oxford), also made a very substantial contribution to the study of Roman Britain, both through his general writings—​which as elegant literary works have yet to be surpassed—​and through his work on the fundamental catalogue of inscriptions (The Roman Inscriptions of Britain). His intellectual influence on archaeology was very substantial too. As explained in his autobiography, he held that knowledge was based on the logic of question and answer, with information being only the result of the questions asked (Collingwood 1939). His work, together with younger scholars, on sorting out the history and chronology of Hadrian’s Wall (which was the subject of intense research through the 1920s and 1930s) exemplified this approach, with small-​scale excavations designed to answer particular questions and the results presented in a tight and disciplined intellectual framework. Such fieldwork built directly on the traditions of keyhole excavation on Hadrian’s Wall, which had been developed in Cumberland from 1894 to 1903. Collingwood’s general writings included a key and enduring practical manual (The Archaeology of Roman Britain, 1930) and the first serious synthesis on the province Roman Britain and the English Settlements (volume 1 of the Oxford History of England, published 1936, with J. N. L. Myers writing the English Settlements section). This set the stage for many subsequent works, providing a synthesis that drew together information from classical texts and archaeology to provide a narrative history of the province, complemented by chapters that dealt with themes such as towns and villas. Although later

Roman Britain since Haverfield    25 scholars have often been critical of his conclusions, it is remarkable how most, down to the present, have replicated this format. Interestingly, Collingwood—​himself brought up in a family active in the Arts and Crafts Movement—​was proudest of his chapter on art in this book (Collingwood 1939: 144). Although a piece of polemical writing that has attracted a lot of criticism, this has also provided the intellectual template for approaches to the subject down to today. The second key figure of the period was Mortimer Wheeler, a man of enormous energy and organizational ability, which he combined with a flair for publicity (Wheeler 1955; Hawkes 1982). Wheeler’s excavating career began in 1913 and included small-​scale excavations on the Balkerne Gate at Colchester, underneath a pub floor, while he was a soldier in 1917. After demobilization, he went on to increasingly bold projects, first in Wales, where he became director of the National Museum, and later in England, with notable work at Lydney, Verulamium, and Maiden Castle. Until her premature death in 1936, he worked jointly with his wife, Tessa Verney Wheeler, who made a fundamental contribution to the projects (Carr 2012). Mortimer Wheeler’s excavations combined clear strategic direction with rapid publication, ensuring that the sites had an immediate impact on thinking. Hence, it is immediately obvious that their Verulamium work was immensely influential on Collingwood’s account of town in Roman Britain and the English Settlements (Collingwood and Myres 1936). However, although Wheeler did write synthetic works of his own—​being responsible for Prehistoric and Roman Wales (1925) and key works on Roman London during his tenure of the Directorship of the London Museum—​they were not his forte. Wheeler was driven by grand narrative, and his excavations were designed to elucidate a story, with the result that they sometimes cut corners, especially in the presentation of the finds like pottery. Interestingly, his doctorate was awarded for work on Roman pottery in the Rhineland, among other things (Wheeler 1920). Nevertheless, he was a pioneer of large-​scale and systematic fieldwork in a manner that had not previously featured in the study of Roman Britain. Although Wheeler worked with a number of people who later made a significant contribution to the subject, he did not himself develop an active school of research students, even following his foundation of the Institute of Archaeology in London in 1937. Not only did this enterprise put archaeological education at the centre of things, but, unusually, it integrated Roman studies with the rest of archaeology (Potter 1987). Wheeler’s work certainly raised the standard of excavation, and those who had dug with him spread his methods, so that they dominated in the middle twentieth century. By contrast, Collingwood was the mentor of the new generation of researchers who followed his agenda—​the logic of question and answer—​in using small-​scale excavation and fieldwork to establish the history of Hadrian’s Wall. Most notable was Eric Birley, who was appointed to a post at Durham University in 1931. He was initially based at its outpost, Armstrong College in Newcastle (later King’s College then, in 1963, to become the University of Newcastle-​upon-​Tyne). Through the late 1920s and 1930s work on Hadrian’s Wall blossomed, and the foundations of what is often known as the Durham School developed, with a focus on both targeted excavation to establish the history of the frontier and the meticulous use of epigraphic material (James 2002). Eric Birley bought

26

26   Martin Millett land around Hadrian’s Wall, including the fort site of Chesterholm (Vindolanda); it is through this that his son Robin Birley developed the excavations from the 1970s, which produced the famous writing tablets. The principal focus of archaeological work in Roman Britain through this period remained military history and urban sites, with comparatively little interest being taken in the countryside, in the archaeology of ordinary people, or in art. There were exceptions to this, such as the excavations on villas such as Bignor (Winbolt and Herbert 1930) and Folkestone (Winbolt 1925), or the description of art objects in museum publications (e.g. British Museum 1922; Wheeler 1930); but, being outside the mainstream, they did not attract much attention. There was also a series of major excavations that continued the pre-​First World War traditions. J. P. Bushe-​Fox was the key excavator, a man of considerable skill who first dug with Sir William Finders Petrie in Egypt, then at Corbridge, taking a key role in the study of the pottery there. He went on to dig at Wroxeter and Hengistbury Head before the First World War and then undertook the key excavations at Richborough from 1922 to 1939. Elsewhere, for instance, at Wroxeter and at Caistor-​ by-​Norwich, some rather poor quality work was done in the same tradition. Another important departure that presaged later development was the growth of rescue archaeology in advance of construction work. There had long been archaeological work undertaken in response to building, and indeed, in London, the London Museum had employed a field officer to record sites since the 1920s (Fulford 2007: 360–​361). Such work developed with projects like the construction of the Colchester bypass in 1930–​1, where Christopher Hawkes—​who had first dug with Wheeler and had been taught by Collingwood at Oxford—​led a major campaign of excavations, exploring the late Iron Age settlement at Camulodunum. Although the publication of his major report was delayed until after the Second World War (and then in a much attenuated form because of the paper shortage), this proved a model for later work where excavations were undertaken in response to redevelopment work, as happened with the construction of airfields during the war.

The 1940s–​1960s In the period following the Second World War there was a gradual growth in the study of Roman Britain, which gathered pace through the 1950s and into the 1960s. The single biggest innovation through this period was the vast growth of archaeology beyond the purely academic sphere. This involved both excavation in advance of construction work and its growth as a leisure pursuit. Research-​based traditions were complemented both by university-​run training excavations and a rash of post-​war ‘rescue excavations’. All continued within a frame of reference still largely dominated by the desire to write narrative history. A key development was the excavation of occupied urban sites (unlike the greenfield sites such as Silchester, previously explored). Bomb damage to historic towns in England led to efforts to

Roman Britain since Haverfield    27 organize excavations to recover evidence of their early history. This difficult work was led by Aileen Fox at Exeter, W. F. Grimes in London, Kathleen M. Kenyon in Southwark, and S. S. Frere at Canterbury. It was revolutionary, not only in changing understanding of the history of the towns but in beginning to create a new route into archaeology. The excavations were comparatively poorly funded and organized through local committees, relying to a greater or lesser extent on volunteer labour. The techniques of deep urban excavation had to be developed and specialist teams brought together to study the finds. This pattern of organization was spread to other sites, with Frere subsequently moving on to major excavations in St Albans in the 1950s–​60s and individuals such as John Wacher and Alan McWhirr going on from there to undertake the excavations at Catterick and in Cirencester. The Verulamium excavations led by Frere were the most influential. As well as the high quality of results from his meticulous excavations, which provided a new model for understanding the origins and endings of towns in Roman Britain, many of the people who worked there went on to make major contributions to the study of Roman Britain. Such summer excavations were not only key training grounds but created a closely networked generation of younger archaeologists with a passion for the subject and a hunger for large-​scale excavation, as well as the knowledge and experience to undertake it. These excavations were rooted in rescue archaeology but developed research agendas, as did the field training schools of the period. Important also were big excavations at Fishbourne and Winchester, led by Barry Cunliffe and Martin Biddle respectively, and later, too, the pair of contrasting Birmingham University summer schools at Wroxeter run by Graham Webster (from 1955 to 1985) and Phil Barker (from 1966 to 1990). The Wroxeter excavations were heavily tied into the long-​standing tradition of liberal adult education, in which archaeology took a key role in the 1960s and 1970s, bringing in keen amateurs and training them through participation in major projects. This blossoming of the archaeology of Roman Britain is reflected in a different way with developments in the north. Eric Birley had a most distinguished war record, with service in 1939–​45 at Bletchley Park, where he used the methods developed for the study of Roman military epigraphy to interpret intelligence material relating to the battle order of the German army. Under his leadership the period after 1945 saw the study of Roman Britain flourish at Durham but with an emphasis closely linked to continental scholarship (James 2002). Significantly, Birley was instrumental in establishing the Limes Congress (first held in Durham in 1949), which linked him and his students with others working on the frontiers—​primarily in the German academic tradition—​ creating a broader tradition of Limesforschungen in Britain. The annual training excavation at Corbridge was one of the few places where students could come to learn the subject. Through these summer seasons from 1947 to 1973 generations of students formed enduring networks. Excavations generated significant volumes of finds that demanded study. Pottery, in particular, was identified as key in dating sites, a matter fundamental to the questions of historical reconstruction with which excavation was primarily concerned. Ceramic chronology, through the study of samian ware, had long been promoted by Eric Birley

28

28   Martin Millett (through whose initiative Stanfield and Simpson’s Central Gaulish Potters came to fruition in 1958). By contrast, coarse pottery was comparatively poorly understood. Important work was promoted on Bushe-​Fox’s excavations, but the first major typo-​ chronological studies were undertaken in the late 1930s and 1940s. The publication of Kathleen Kenyon’s study in the Jewry Wall volume (1948) and Christopher Hawkes and M. R. Hull’s work on the pottery from the pre-​war excavations at Camulodunum (1947) had set the agenda for such studies in southern England—​the latter in particular linking to continental literature. In Newcastle, John Gillam, a key figure in the Corbridge excavations and a student of Birley and Richmond, developed this field of study with the publication of his Types of Roman Coarse Pottery Vessels in Northern Britain, first published in Archaeologia Aeliana in 1957, and subsequently revised as a book in 1968. Gillam’s typological approach, like that reflected in Stanfield and Simpson’s work on samian ware, built on the historical chronology of the northern military sites, which had been established by previous excavations and created a monolithic framework that proved enduring, notwithstanding later changes in the dating of sites upon which it was based. The towering figure of the period was undoubtedly Ian Richmond, an Oxford graduate who had been sent by Collingwood to dig with Wheeler at Segontium in 1922. He was briefly Director of the British School at Rome (BSR) (1930–​2), and taught at the University of Durham’s King’s College, Newcastle, from 1935 to 1956—​b ecoming Professor in 1950—​w here he did key work on the frontier. He went on to be appointed as the first holder of the chair in the Archaeology of the Roman Empire at Oxford in 1956, a position that he held until his early death in 1965. His work was geographically and intellectually wide-​ranging, but he put great energy into his fieldwork in Britain, working on a range of military sites, including his notable excavations at Hod Hill (1951–​8 ) and Inchtuthill (1952–​65). At Oxford he inspired a new generation of students who went on to make key contributions to the subject. In contrast to some contemporaries, Richmond was also much wider ranging in his portfolio of publications. This included the first major post-​w ar synthesis—​Roman Britain, first published in 1955—​w hich remained in print in the Penguin History of Britain series for many years, very effectively popularizing the subject. Interestingly, the format and content of the book were very much influenced by Collingwood’s template, but, given its enduring character, Richmond’s little book had the effect of ‘fixing’ the orthodoxy for many years. Another legacy was indirect: as at Oxford, he had set about preparing a synthesis of the temple complex at Bath. With his death this project passed to the energetic young Barry Cunliffe, who ensured its rapid completion and publication in 1969, in the same way he had done with the fifth volume of Bushe-​Fox’s work at Richborough a year earlier, thereby establishing himself not only as the excavator of the spectacular site at Fishbourne, but as a key figure in Romano-​British studies. Uniquely for this period, Cunliffe’s training had been in Archaeology and Anthropology (rather than Classics) at Cambridge. His excavations at Fishbourne

Roman Britain since Haverfield    29 and Portchester were also an important training ground for the next generation of archaeologists. Another interesting figure closely associated with Richmond at the British School at Rome (BSR) in the 1920s and 1930s was Jocelyn Toynbee, who came from a distinguished intellectual family. She was influenced at the BSR by Eugénie Sellers Strong and prepared a magnificent book based on her Oxford doctoral thesis on Hadrianic sculpture. She had been an undergraduate in Cambridge, then taught at Oxford and Reading, before returning to Cambridge, where she was a Fellow of Newnham College and rose to become the Laurence Professor of Classical Archaeology in 1951–​62. She was not only a significant art historian but made a series of major contributions to Roman archaeology more broadly. In the context of Roman Britain, she wrote the first influential contribution to the study of Christianity in Britain (1953) as well as organizing a key exhibition in London on art in Roman Britain in 1962. The catalogue of this show (Toynbee 1962), and the major book that followed two years later (Toynbee 1964), provided the first full-​length treatment of the subject and had an enduring impact, not only by drawing together the evidence in one place where it could be more fully appreciated but also by establishing a genre of study. In many ways the Guildhall Museum exhibition organized by Jocelyn Toynbee is but one symbol of the dynamism and diversification of the subject in this period. Not only did other great projects come to fruition—​including the publication of the first volume of The Roman Inscriptions of Britain (1965), I. D.  Margary’s Roman Roads in Britain (1955–​7), the influential third edition of the Ordnance Survey Map of Roman Britain (1956), as well as the Ordnance Survey Map of Hadrian’s Wall (1964) —​but there was a series of other meetings and publications of key work that set an agenda for the future. An important conference held in Leicester in 1963 drew together the results of work on the towns of the province under the title The Civitas Capitals of Roman Britain (edited by J. S. Wacher, 1966). Religion, already noted among Toynbee’s interests, but that had hitherto been little explored, benefited from the synthesis of temples by M. J. T. Lewis (1966), a student of hers. Similarly, a contrasting approach, which laid the ground for later work and sought to emphasize indigenous traditions, appeared in the form of Ann Ross’s Pagan Celtic Britain (1967). The countryside, which had not figured very largely in previous work, was brought to the fore in contrasting ways. An epoch-​making volume Town and Country in Roman Britain was published in 1958 by A. L. F. Rivet, who was the Assistant Archaeology Officer of the Ordnance Survey and, as such, had been responsible for the production of the third edition of the Map of Roman Britain. He was a pupil of Collingwood and had dug with Wheeler. He used this work to present a coherent and powerful narrative on the Romano-​British countryside, providing an account that has never been surpassed for its originality. Later, the villas of the province were also the subject of his The Roman Villa in Britain (edited 1969), drawing together a variety of new evidence. Perhaps of most enduring value within this volume was David Smith’s synthesis of the evidence of mosaics, in which he sought to identify a series of regional artistic schools

30

30   Martin Millett in the province. Indigenous aspects of the countryside were also brought into focus for the first time through the Council for British Archaeology conference held in Oxford in 1965, which resulted in the publication, in 1966, of Rural Settlement in Roman Britain (edited by Charles Thomas). These initiatives were generally associated with the civilian areas of Roman Britain, and there was remarkably little synthesis of the extensive work done on the Roman frontiers in the north, although, in Wales, V. E. Nash Williams produced his key study The Roman Frontier in Wales (1954, revised M. G. Jarrett 1969). Work on Hadrian’s Wall and the Antonine Wall continued apace, but was mostly published in a stream of articles in the Proceedings of the Society of Antiquaries of Scotland, Archaeologia Aeliana, and the Transactions of the Cumberland and Westmoreland Archaeological Society. A  much-​read general review of the military was provided in Graham Webster’s book The Roman Imperial Army (1969), which was, in effect, largely based on Roman Britain. In contrast, new attitudes to the relationship between the Roman army and the indigenous peoples were first seen in Peter Salway’s volume The Frontier People of Roman Britain (1965) and, more generally, in the fieldwork by George Jobey in Northumbria. A whole range of different aspects of the archaeology of the province were also increasingly revealed through campaigns of aerial photography, most notably those initiated by J. K. St Joseph based in Cambridge. As well as revealing a whole swathe of new sites, especially forts and villas, he worked systematically to use aerial photography to explore the Roman military campaigns in the north, contributing much to the explicit research agenda of reconstructing the narrative history of the province. The culmination of this period of energetic research came in 1967 with the publication of Sheppard Frere’s Britannia: A History of Roman Britain. He had been a student of Classics at Cambridge, embarking on his major excavations in the post-​war period while teaching at Lancing (where incidentally Haverfield had also taught), before holding a Readership at the Institute of Archaeology from 1955 (in succession to Wheeler), being promoted to Professor in 1963, then moving to the Chair of the Archaeology of the Roman Empire in Oxford in 1966. He had a unique command of the subject and Britannia was a masterful volume, written in a style that summarized existing knowledge and understanding firmly and authoritatively. It not only told us what was known, but demonstrated how it should be understood, very much within the tradition of Britain’s imperial power. Following the format set by Collingwood, it remained as the key work for a generation.

The 1970s and 1980s Even as the undoubted successes of the post-​war generation were coming together, there was a series of changes taking place that was to have fundamental implications for the study of Roman Britain. These changes affected not only the evidence base of

Roman Britain since Haverfield    31 the subject but also how the province was to be studied. The trends of the 1960s continued for some time, with the publication of major syntheses like John Wacher’s Towns of Roman Britain (1974) and substantial works of reference including Rivet and Smith’s fundamental volume The Place-​Names of Roman Britain (1979) and A. R. Birley’s Fasti of Roman Britain (1981). The British volumes of the Corpus Signorum Imperii Romani also began to appear in 1977. The year 1970 also saw the publication of the first volume of the journal Britannia devoted to the study of Roman Britain, which was very much Frere’s creation. Finally, in a completely unpredicted way, excavations at Vindolanda in 1973 produced a mass of new textual evidence to enrich our knowledge of the Roman military (Bowman 1994). Nevertheless, the landscape had already begun to change, and orthodoxy was being questioned. With the economic growth of the 1960s, many new archaeological discoveries were made as a result of redevelopment work, whether building projects or motorway constructions. The scale of this activity put an enormous strain on the established response in these circumstances, which had seen central government, through the Ministry of Public Buildings and Works (later the Department of the Environment), fund individual archaeologists to dig threatened sites. It was becoming evident that this ad hoc system, which had involved both freelance diggers and academics being brought in to run the excavations, was now no longer sufficient, and increasing political pressure mounted for better provision. It was in these circumstances that the ‘rescue’ movement was born, with the objective of creating a professional state-​funded archaeological service to deal with the problem (Jones 1984). In a very British way the solution was not to create a state service, but a plethora of local solutions emerged with local authorities, museums, universities, and independent organizations establishing ‘excavation units’ piecemeal across the country to dig sites funded by the state. This boom did result in the wholesale recording of sites threatened by destruction and the establishment of a network of professional excavation teams, but there was massive variation in its extent and quality, and whether it was ever published. At its best, work of this period produced key new insights into urban development and hitherto poorly understood aspects of Roman Britain. For instance, work at Colchester gave a clear insight into the origins of the colonia (Crummy 1988). The move towards large-​scale rescue excavation went hand-​in-​hand with a philosophical shift among excavators. Whereas previous generations, whether explicitly or implicitly, had mostly excavated to answer research questions and contribute to understanding the past, there was a strong belief among many of the rescue era that excavating could be a ‘neutral’ activity that ‘saved’ evidence through ‘objective’ recording so that it could be thought about and interpreted later, accordingly promoting the concept of ‘preservation by record’. This form of excavation, promoted intellectually by Phil Barker (1977), especially in his excavations at Wroxeter, had a disastrous effect on publication, with many practitioners only belatedly discovering that, if you do not understand what you are digging while the evidence is in front of you, you will find it almost impossible to make sense of it later in publication. As a result, and with an overwhelming number of excavations, few sites made any real impact on the understanding of Roman Britain.

32

32   Martin Millett However, the rescue boom did demonstrate that Roman Britain had been very densely occupied by rural sites, so that few building projects did not find Roman material. This, in turn, finally disposed of the idea that the Romano-​British landscape had been dominated by vast tracts of uninhabited woodland with isolated clearings for towns and villas. The major samples of the landscape revealed by the construction of motorways also led to an increasing interest in field systems and the organization of the countryside—​an interest that moved the focus of rural archaeology away from elite buildings and towards a desire to understand whole landscapes (Fowler 1975). The intellectual impact of this was to encourage new forms of landscape research that attempted to see settlement more holistically and diachronically. This had a notable impact in some areas: for instance, through the pioneering work of Cunliffe at Chalton in Hampshire (Cunliffe 1973) and David Miles in the Upper Thames valley (Miles 1988). The downside of the excavation philosophy of this period was, to some extent, counterbalanced by the development of a view of the equality of the evidence. Previously, excavators had often been interested principally in finds that were of obvious value in establishing the chronology of their sites (namely, pottery and coins), or material that had some intrinsic ‘artistic’ interest. Now excavators who were less focused on reconstructing a narrative history sought to analyse a fuller range of finds, leading to a growth of studies not only of pottery and small finds but, increasingly, of the animal bones and botanical remains that held the key to understanding the agrarian economy and the environment (King 1978; Dimbleby and Jones 1981). This led to a greater appreciation of the potential of excavated finds in order to understand past society and stimulated approaches that increasingly drew on the natural sciences for the development of new modes of study. Notable in this respect were changes in the study of pottery pioneered by David Peacock at Southampton University. Building on his geological training, he developed methods that not only promoted the characterization and sourcing of ceramics, but also insisted that they should be used for understanding the past economy and not simply for dating (Peacock 1982). Similarly, traditional approaches to the coin evidence were complemented by a range of new studies pioneered by Richard Reece and John Casey, which sought to use the evidence to look at broader issues of archaeological interpretation (Casey and Reece 1974; Reece 1987). In all these areas the influence of alternative intellectual trends represented by the ‘New Archaeology’ began to be felt in the study of Roman Britain (Johnson 1999: 12–​33). Thoughts increasingly turned to thinking about Roman Britain as a social and economic system, while studies like those by Ian Hodder showed how the evidence could contribute to broader debates in archaeology (e.g. Hodder 1972; Hodder and Orton 1976). The sharp response to such work in Sheppard Frere’s valedictory lecture at Oxford in 1987 well illustrates how the study of Roman Britain had become a heavily contested subject. The changing landscape of higher education in the UK played an important role in the shift in emphasis away from historical narrative, as archaeology degrees in

Roman Britain since Haverfield    33 universities become more widespread. This was linked to the growth of a much broader-​based higher education system during the 1960s with the proliferation and growth of universities and a diversification of subjects taught. This widening of undergraduate education involved the founding of new universities and of archaeology departments. Existing departments grew and began to offer undergraduate degrees in archaeology (where most had previously taught it only as a subsidiary subject). In addition, newly created universities such as Southampton, Newcastle, and Leicester saw it as a key subject worthy of support. Thus, through the late 1960s and into the early 1970s, new degree courses came into being, and academic jobs were created and filled, many by those active in the study of Roman Britain. Some existing centres had also expanded or reinforced the discipline. Oxford had already created a chair in the Archaeology of the Roman Empire in 1956, bringing the study of Roman Britain to classics and history students. Durham moved to build a fully-​ fledged Department of Archaeology, while also retaining and expanding its work on Roman Britain. From the 1970s onwards, there was a series of growing centres of Roman teaching and research in British universities (most notably London’s Institute of Archaeology, Cardiff, Southampton, Reading, Leicester, Durham, Birmingham, Manchester, Oxford, and Newcastle-​upon-​Tyne, plus others that emerged later). Each of these came to produce groups of graduates who made distinctive contributions to particular areas of the subject, while appointments to lectureships and movements between universities created a complex web of interactions, so that today many of those in posts teaching Roman archaeology in Britain share linkages through who have taught them and which excavations they took part in (Table 2.1). Two particularly influential figures in the teaching of the subject were the contrasting figures of Richard Reece and Barri Jones. At the Institute of Archaeology, undergraduate archaeology was taught from 1968 with sub-​departments offering specialist training in their own areas. From 1974, the Roman Department comprised Professor John Wilkes (a Durham Ph.D.  graduate who had taught at Birmingham), Mark Hassall, a product of the Verulamium excavation who had studied in Oxford then London, and Richard Reece, who had studied biochemistry at University College London and taught in schools before doing his doctoral research on Roman coinage at Oxford. He had a unique, Socratic teaching style, stimulating critical thought and discussion among his students. In this period, the Institute attracted students who were already strongly committed to the subject, and, among the Romanists, Richard Reece developed a very strong following, with the result that significant numbers of them went on to take leading roles in the subject. Although Richard Reece was not himself an excavator on a large scale, his encouragement of generations of students to engage critically with primary material meant that many went on to do important work in the field and on finds, especially pottery and coins. Barri Jones, an Oxford pupil of Sir Ian Richmond’s, was also highly influential in the subject. He was a key player in the UK rescue movement in the early 1970s, having done important work in Italy and North Africa. He held a lecturing post at the University of

34

Table 2.1 The educational networks of the contributors to this volume 1960s 1970s

1980s

1990s

2000s

Oxford

•

* [S. S. Frere]

• * [S. S. Frere/​ J. Lloyd]

• * [A. Wilson]

Cambridge

•

-

•••

• # # * [J. Clackson] * [M. Millett] * [M. Millett] * [M. Millett]

Birmingham

•

•

Cardiff

•

* [W. Manning]

#

Sheffield

•#

* [G. Jones/​ D. Gilberstson]

•#

Institute of Archaeology

•••

Newcastle

•

Durham

•

* [C. F. C.Hawkes] * [S. S. Frere]

• • # # # * [R. Reece]

•#

• * [R. Jones] * [R. Jones]

Southampton

* [T. Champion]

Reading

• # * [J. Crow] * [P. van der Eijk] • • • # # # # # * [M. Millett/​C. Haselgrove/ ​C. Caple/​ J. Bayley] * [M. Millett/​ J. Price]

Bradford

* [R. Reece] * [R. Reece] * [C. Lockyear/​ S. Moorhead

• • # # # * [C. Haselgrove] * [R. Hingley/​ A. Leone] * [S. Lucy/​A. Millard] * [R. Hingley/​R. Witcher]

* [R. Jones]

# * [M. Millett] * [L. Revell] # * [M. Fulford/ ​A. Wallace-​Hadrill] * [M. Fulford] * [M. Fulford/​H. Härke]

2010s

Roman Britain since Haverfield    35 Table 2.1 Continued 1960s 1970s

1980s

1990s

2000s

2010s

•

Warwick Kent

• • # * [E. Swift/​A. Ward]

York

• # * [S. Roskams] * [D. Perring/​S. Roskams]

Wales–​ Newport

* [Aldhouse-​ Green]

Bournemouth

* [Darvill/​ O’Connor]

Leicester Other European

North America

•#

• # * [J. Bintliff/ ​B. van der Meer] •

••

* [Wailes]

Note: • = Undergraduate; # = Masters; * = Ph.D. [Supervisor(s)].

Manchester from 1964 and became Professor in 1971, after playing a key role in establishing the degree programme there. He was a dynamic fieldworker and an inspirational teacher, many of whose students from the 1970s went on to make key contributions to the study of Roman Britain, both as academics and in the newly growing excavation units. One of the key long-​term results of the growth of archaeology degrees and the boom in excavation in Britain was a change in attitude to the Roman past. Hitherto, most engaged in its study had been trained and immersed in Classics, thus giving priority to ancient texts in their studies and tending towards approaches that laid emphasis on reconstructing the narrative history from a Roman perspective. The new generation, which had often discovered the subject through digging, was brought up on ideas about using material evidence to approach the past in a variety of different ways, often putting emphasis on social and economic history, seeking the ordinary people rather than the elites and willing to expend efforts on new types of evidence. Equally, the structure of higher education meant that more people gained the opportunity to undertake doctoral research, so a range of major new studies could be undertaken, stretching the subject in new directions and often developing approaches that owed as much to the research traditions of prehistory as to the Classical world. Important in this regard was a series of Ph.D. projects that produced syntheses of aspects of the

36

36   Martin Millett archaeology of Roman Britain. For example, Mike Fulford produced a study of New Forest Roman pottery (1975), Tom Blagg wrote his thesis on architectural ornament (1982, published 2002), and Robert Philpott drew together the evidence for burial rites (1991). Thus narrative historical approaches no longer dominated the subject in the way that they had up to the 1970s. As these new approaches grew, there was a series of conferences and seminars that broke away from the consensus of the 1960s, questioning both conclusions and assumptions while offering examples of the value of new approaches. These included new studies of cemeteries (Reece 1977); different approaches to imperialism (Burnham and Johnson 1979); studies of the agrarian landscape (Miles 1982); and a questioning of the role of the military in the development of the province (Blagg and King 1984). Many of these ideas were also taken up in some of the professional excavation units, which saw a role not simply for gathering information but for forwarding understanding. Important studies of the military archaeology in the north of England and Scotland not only swept away the previous consensus about the chronology and function of the frontiers (Breeze and Dobson 1976; Hanson and Maxwell 1983), but also sought to explore the complex relationship between the army and local communities. The leading proponent of this approach was David Breeze, who had been a student of Eric Birley’s at Durham and went on to a career in heritage management in Scotland, but never lost sight of research priorities (Breeze 1982). The culmination of these processes was a group of alternative publications that came out as the members of the new generation established themselves in their careers (Henig 1984; Allason-​Jones 1989; Esmonde Cleary 1989; Hingley 1989; Mattingly and Jones 1990). Among these was my own Romanization of Britain (Millett 1990), designed as an attempt to provide a coherent alternative narrative to that with which I had been brought up (Millett 2003–​4).

The 1990s and Beyond In the academic sphere, the period since 1990 has been dominated by much spirited debate, which has found voice principally through the Theoretical Roman Archaeology Conference (TRAC), which began in Newcastle in 1991. Discussion has focused on critiques of the concept of Romanization, but, increasingly, TRAC has become a forum for a range of ideas that can loosely be characterized as post-​imperial approaches. The other strength of these meetings has been their increasing emphasis on more sophisticated interpretations of artefacts, moving them from the margins to the centre stage for discussion. Equally, there has been a strong trend towards evaluating the history of ideas in the subject and promoting alternative approaches. This has meant that research on Roman Britain now has a rare dynamism and excitement that is recognized by other branches of archaeology.

Roman Britain since Haverfield    37 In commercial archaeology a substantial change also took place in 1990 with the implementation of a system whereby, if required by the local planning authority, the developer of a site, rather than the state, was expected to pay for its archaeological investigation prior to destruction (see also Wilson, this volume). This completely changed the nature of commercial archaeology, as the pressure of competition on price came to bear. Excavation was no longer determined solely by academic strategies, and considerable autonomy was given to local authority advisers in determining what was required at particular sites. A positive benefit of this has been that, in some areas, extremely skilled and experienced archaeological teams have been able to carry through big projects on good budgets with excellent results that are changing our understanding. So, for instance, London is now probably among the most extensively explored and fully published of the cities in the Roman Empire. Equally, because of the need to justify the work undertaken, there has been a renewed emphasis, at a national and a regional level, on debating what we know about Roman Britain and encouraging commercial excavations to contribute to knowledge by addressing specific research agendas (James and Millett 2001). This has also required the beginnings of some systematic re-​evaluation of what is known about the province, as much information is not now published but archived. This all takes place against a background in which many full-​time field archaeologists are of the highest calibre and are contributing to understanding Roman Britain on a massive scale. Against this there are also areas, often in regions where property values are not high or where local authorities are weak, where the quality of archaeological work is poorer and little is being added to collective understanding. Equally, the lack of systematic publication makes it all but impossible to maintain a clear overview of the subject (Fulford 2011; Fulford and Holbrook 2011). In parallel with this there has also been a resurgence of excellent primary research, combining both the synthesis of existing evidence with first-​class new fieldwork and excavation. The publication of inscriptions has progressed with the completion of volumes 2 and 3 of The Roman Inscriptions of Britain, as well as completion of the definitive publication of the Vindolanda tablets (Bowman and Thomas 1983, 1994, 2003; Bowman, Thomas and Tomlin 2010, 2011). Anthony Birley has produced an updated version of his Fasti, albeit under a new title (2005). The four volumes of David Neal and Stephen Cosh’s Roman Mosaics in Britain are now in print, and the British volumes of the Corpus Signorum Imperii Romani are all now published. Further important and pioneering works have appeared, like Paul Tyers’s Roman Pottery in Britain (1996) and Jeremy Taylor’s Atlas of Roman Rural Settlement in Britain (2007), making this among the best-​ published provinces in the empire. Archaeology has also seen the development of new techniques of research—​some of which are now commonplace—​like large-​scale geophysics, which has been used extensively to map sites in Roman Britain, providing a wealth of information about a range of site types. Other new techniques like isotope studies are altering the questions we can answer, focusing attention on aspects of demography that had hitherto been a closed book.

38

38   Martin Millett Similarly there are now a series of alternative yet authoritative syntheses of the subject that provide contrasting and stimulating approaches to the province (Cool 2006; Creighton 2006; Mattingly 2006). These, and a rich range of other works, show how Roman Britain is now a contested subject and one in which much stimulating work is taking place. Given the scope and content of the present volume, which is designed to show the range and diversity of contemporary approaches to Roman Britain, it is unnecessary to provide any summary of current research here. It is worth noting how there is now a wide recognition that knowledge is the product of addressing questions, not simply data-​gathering, and that the questions we ask and the research we produce are themselves a product of contemporary perspectives. This leads to a postmodern diversity of approaches. Against this background we should also note that recognition of our information base is now much larger than ever before, not only as a result of commercial developer-​funded excavations but also because of the evidence from metal-​detecting now brought together by the ‘Portable Antiquities Scheme’ (see also Wilson, this volume). The challenge is how to make sense of this information in ways that help elucidate varied aspects of the history of Roman Britain.

References Allason-​Jones, L. (1989). Women in Roman Britain. London: British Museum Press. Barker, P. (1977). Techniques of Archaeological Excavation. London: B. T. Batsford. Birley, A. R. (1981). Fasti of Roman Britain. Oxford: Clarendon Press. Birley, A. R. (2005). The Roman Government of Britain. Oxford: Oxford University Press. Bishop, M. C. (1994). Corstopitum: An Edwardian Excavation. London: English Heritage. Blagg, T. F.  C. (2002). Roman Architectural Ornament in Britain. BAR British Series 329. Oxford: Archaeopress. Blagg, T. F. C., and King, A. C. (1984) (eds). Military and Civilian in Roman Britain. BAR British Series 136. Oxford: Archaeopress. Bowman, A. K. (1994). Life and Letters on the Roman Frontier:  Vindolanda and its People. London: British Museum Press. Bowman, A. K., and Thomas, J. D. (1983). Vindolanda: The Latin Writing-​Tablets. London: Britannia Monograph Series 4. Bowman, A. K., and Thomas, J. D. (1994). The Vindolanda Writing Tablets (Tabulae Vindolandenses II). London: British Museum Press. Bowman, A. K., and Thomas, J. D. (2003). The Vindolanda Writing Tablets (Tabulae Vindolandenses III). London: British Museum Press. Bowman, A. K., Thomas, J. D., and Tomlin, R. S. O. (2010). ‘The Vindolanda writing-​tablets (Tabulalae Vindolandenses IV, part 1)’. Britannia, 41: 187–​235. Bowman, A. K., Thomas, J. D., and Tomlin, R. S. O. (2011). ‘The Vindolanda writing-​tablets (Tabulalae Vindolandenses IV, part 2)’. Britannia, 42: 113–​144. Breeze, D. (1982). The Northern Frontiers of Roman Britain. London: B.T. Batsford. Breeze, D., and Dobson, B. (1976). Hadrian’s Wall. London: Allen Lane.

Roman Britain since Haverfield    39 British Academy (1977–​ to date). Corpus Signorum Imperii Romani. Oxford:  Oxford University Press. British Museum (1922). Guide to the Antiquities of Roman Britain. London: Trustees of the British Museum. Burnham, B. C., and Johnson, H. B. (1979) (eds). Invasion and Response: The Case of Roman Britain. BAR British Series 73. Oxford: Archaeopress. Carr, L. C. (2012). Tessa Verney Wheeler:  Women in Archaeology before World War II. Oxford: Oxford University Press. Casey, P. J., and Reece, R. M. (1974) (eds). Coins and the Archaeologist. BAR British Series 4. Oxford: Archaeopress. Collingwood, R. G. (1930). The Archaeology of Roman Britain. London:  Methuen and Company. Collingwood, R. G. (1939). An Autobiography. Oxford: Oxford University Press. Collingwood, R. G., and Myres, J. N. L. (1936). Roman Britain and the English Settlements. Oxford: Clarendon Press. Collingwood, R. G., Wright, R. P., Frere, S. S., Roxan, M. M., Tomlin, R., and Hassall, M. W. C. (1965–​to date). The Roman Inscriptions of Britain. Oxford: Clarendon Press. Cool, H. E.  M. (2006). Eating and Drinking in Roman Britain. Cambridge:  Cambridge University Press. Creighton, J. (2006). Britannia: The Creation of a Roman Province. London: Routledge. Crummy, P. (1988). ‘Colchester’, in G. Webster (ed.), Fortress into City. London: Batsford, 24–​47. Cunliffe, B. W. (1968). Fifth Report on the Excavations of the Roman Fort at Richborough, Kent. Oxford: Oxford University Press. Cunliffe, B. W. (1969). Roman Bath. London: Society of Antiquaries of London. Cunliffe, B. W. (1973). ‘Chalton, Hants.: The Evolution of a Landscape’, Antiquaries Journal, 53: 173–​190. Curle, J. (1911). A Roman Frontier Post and its People: The Fort of Newstead in the Parish of Melrose. Glasgow: James Maclehose and Sons. Dimbleby, G., and Jones, M. K. (1981) (eds). The Environment of Man: The Iron Age to the Anglo-​ Saxon Period. BAR British Series 87. Oxford: Archaeopress. Esmonde Cleary, A. S. (1989). The Ending of Roman Britain. London: B. T. Batsford. Fowler, P. J. (1975) (ed.). Recent Work in Rural Archaeology. London: Adams and Dart. Freeman, P. (2007). The Best Training-​Ground for Archaeologists: Francis Haverfield and the Invention of Romano-​British Archaeology. Oxford: Oxbow Books. Frere, S. S. (1967). Britannia: A History of Roman Britain. London: Routledge and Kegan Paul. Frere, S. S. (1987). Roman Britain since Haverfield and Richmond: A Lecture Delivered in All Souls College, 23 October 1987. Oxford:  Reprinted from History and Archaeology Review. Fulford, M. G. (1975). New Forest Roman Pottery. BAR British Series 17. Oxford: Archaeopress. Fulford, M. G. (2007). ‘The Grand Excavation Projects of the Twentieth Century’, in S. Pearce (ed.), Visions of Antiquity: The Society of Antiquaries of London 1707–​2007. London: Society of Antiquaries (=Archaeologia, 111), 33–​81. Fulford, M. G. (2011). ‘The Impact of Commercial Archaeology on the UK Heritage’, in B. W. Cunliffe (ed.), History for the Taking? Perspectives on Material Heritage. London:  British Academy, 33–​53.

40

40   Martin Millett Fulford, M. G., and Holbrook, N. (2011). ‘Assessing the Contribution of Commercial Archaeology to the Study of the Roman Period in England, 1990–​2004’, Antiquaries Journal, 91: 323–​345. Gillam, J. P. (1957). ‘Types of Roman Coarse Pottery Vessels in Northern Britain’, Archaeologia Aeliana, 4th ser., 35: 180–​251. Gillam, J. P. (1968). Types of Roman Coarse Pottery Vessels in Northern Britain. Newcastle-​ upon-​Tyne: Oriel Press. Hanson, W., and Maxwell, G. (1983). Rome’s North-​ West Frontier:  The Antonine Wall. Edinburgh: Edinburgh University Press. Haverfield, F. (1912). The Romanization of Roman Britain. Oxford: Clarendon Press. Haverfield, F., and MacDonald, G. (1924). The Roman Occupation of Britain. Oxford: Oxford University Press. Hawkes, C. F.  C., and Hull, M. R. (1947). Camulodunum:  First Report on the Excavations at Colchester, 1930–​1939. Oxford:  Report of the Research Committee of the Society of Antiquaries of London, 14. Hawkes, J. (1982). Mortimer Wheeler:  Adventurer in Archaeology. London:  Weidenfeld and Nicolson. Henig, M. (1984). Religion in Roman Britain. London: B. T. Batsford. Hingley, R. (1989). Rural Settlement in Roman Britain. London: Seaby. Hingley, R. (2000). Roman Officers and English Gentlemen: The Imperial Origins of Roman Archaeology. London: Routledge. Hodder, I. R. (1972). ‘Locational Models and the Study of Romano-​British Settlement’, in D. L. Clarke (ed.), Models in Archaeology. London: Methuen, 887–​909. Hodder, I. R., and Orton, C. R. (1976). Spatial Analysis in Archaeology. Cambridge: Cambridge University Press. James, S. T. (2002). ‘Writing the Legions: The Development and Future of Roman Military Studies in Britain’, Archaeological Journal, 159: 1–​58. James, S. T., and Millett, M. (2001) (eds). Britons and Romans: Advancing an Archaeological Agenda. CBA Research Report, 125. York: Council for British Archaeology. Johnson, M. H. (1999). Archaeological Theory: An Introduction. Oxford: Blackwell. Jones, G. D. B. (1984). Past Imperfect: The Story of Rescue Archaeology. London: Heinemann. Jones, R. F. J. (1987). ‘The Archaeologists of Roman Britain’, Bulletin of the University of London Institute of Archaeology, 24: 85–​97. Kenyon, K. M. (1948). Excavations at the Jewry Wall site, Leicester. Oxford:  Report of the Research Committee of the Society of Antiquaries of London. 15. King, A. C. (1978). ‘A Comparative Survey of Bone Assemblages from Roman Sites in Britain’, Bulletin of the University of London Institute of Archaeology, 15: 207–​232. Lewis, M. J. T. (1966). Temples in Roman Britain. Cambridge: Cambridge University Press. MacDonald, G. (1911). The Roman Wall in Scotland. Glasgow: J. Maclehose and Sons. Margary, I. D. (1955–​7). Roman Roads in Britain. 2 vols. London: Phoenix House. Mattingly, D. (2006). An Imperial Possession: Britain in the Roman Empire. London: Allen Lane. Mattingly, D., and Jones, G. D. B. (1990). An Atlas of Roman Britain. Oxford: Blackwell. Miles, D. (1982) (ed.). The Romano-​ British Countryside. BAR British Series 103. Oxford: Archaeopress. Miles, D. (1988). ‘Villas and Variety: Aspects of Economy and Society in the Upper Thames Landscape’, in K. Branigan and D. Miles (eds), The Economies of Romano-​British Villas. Sheffield: Department of Archaeology and Prehistory, 60–​72.

Roman Britain since Haverfield    41 Millett, M. (1990). The Romanization of Britain: An Essay in Archaeological Interpretation. Cambridge: Cambridge University Press. Millett, M. (2003–​4). ‘The Romanization of Britain:  Changing Perspectives’, Kodai, 13/​14: 169–​173. Nash Williams, V. E. (1954). The Roman Frontier in Wales. Cardiff: University of Wales Press. Neal, D. S., and Cosh, S. R. (2002–​2010). Roman Mosaics of Britain. Vols. 1–​4. London: Society of Antiquaries. Ordnance Survey (1956). Map of Roman Britain. Chessington: Ordnance Survey. Ordnance Survey (1964). Map of Hadrian’s Wall. Chessington: Ordnance Survey. Peacock, D. P. S. (1982). Pottery in the Roman World. London: Longman. Philpott, R. (1991). Burial Practices in Roman Britain. BAR British Series 219. Oxford: Archaeopress. Potter, T. W. (1987). ‘The Institute and Roman Archaeology: Past, Present and Future—​ an Outsider’s View’, Bulletin of the University of London Institute of Archaeology, 24: 71–​84. Reece, R. M. (1977) (ed.). Burial in the Roman World. CBA Research Report 22. London: Council for British Archaeology. Reece, R. M. (1987). Coinage in Roman Britain. London: Seaby. Richmond, I. (1955). Roman Britain. London: Penguin Books. Rivet, A. L. F. (1958). Town and Country in Roman Britain. London: Hutchinson University Library. Rivet, A. L. F. (1969) (ed.). The Roman Villa in Britain. London: Routledge and Kegan Paul. Rivet, A. L.  F., and Smith, C. (1979). The Place-​Names of Roman Britain. London:  B.T. Batsford. Ross, A. (1967). Pagan Celtic Britain: Studies in Iconography and Tradition. London: Routledge and Kegan Paul. Salway, P. (1965). The Frontier People of Roman Britain. Cambridge:  Cambridge University Press. Stanfield, J. A., and Simpson, G. (1958). Central Gaulish Potters. Oxford:  Oxford University Press. Taylor, J. (2007). Atlas of Roman Rural Settlement in Britain. York:  Council for British Archaeology. Thomas, C. (1966) (ed.). Rural Settlement in Roman Britain. London:  Council for British Archaeology. Toynbee, J. M. C. (1953). ‘Christianity in Roman Britain’, Journal of the British Archaeological Association. 3rd ser., 16: 1–​24. Toynbee, J. M. C. (1962). Art in Roman Britain. London: Phaidon. Toynbee, J. M. C. (1964). Art in Britain under the Romans. Oxford: Oxford University Press. Tyers, P. (1996). Roman Pottery in Britain. London: B. T. Batsford. Wacher, J. (1966) (ed.). The Civitas Capitals of Roman Britain. Leicester:  Leicester University Press. Wacher, J. (1974). Towns of Roman Britain. London: B.T. Batsford. Webster, D. B. (1991). Hawkeseye: The Early Life of Christopher Hawkes. Stroud: Alan Sutton. Webster, G. (1969). The Roman Imperial Army. London: Black. Wheeler, R. E. M. (1920). Comparative Notes on the Rhenish Pottery of the Roman Period (with an Appendix on the Balkerne Gate, Colchester). D.Litt. thesis, University College London. Wheeler, R. E. M. (1925). Prehistoric and Roman Wales. Oxford: Clarendon Press.

42

42   Martin Millett Wheeler, R. E. M. (1930). London in Roman Times. London: London Museum. Wheeler, R. E.  M. (1955). Still Digging:  Interleaves from an Antiquary’s Notebook. London: Michael Joseph. Winbolt, S. E. (1925). Roman Folkestone: A Record of Excavation of Roman Villas at East Wear Bay. London: Methuen and Co. Winbolt, S. E., and Herbert, G. (1930). The Roman Villa at Bignor, Sussex [with a Plan of the Villa]. Oxford: Clarendon Press.

Chapter 3

Romano-​B ri t i sh Archaeol o g y i n t h e E arly T wen t y-​Fi rst Century Pete Wilson

Introduction The early twenty-​first century is a period of rapid change even for such a venerable a subject as Romano-​British archaeology. Britannia has been studied since at least the time of Camden but, as a subject, can be argued to have been resistant to new thinking, dominated by a succession of ‘greats’ whose influence, in some cases unwittingly, often became an obstacle to the development of new interpretations and approaches. The tendency for the great and the good to dominate the subject extended well into the twentieth century: one has only to think of Eric Birley’s influence in relation to Hadrian’s Wall or the colossus-​like contributions of Sheppard Frere across a wide variety of subjects. Frere’s 1974 edition of Britannia is still this archaeologist’s first choice when checking ‘basic facts’ on a wide range of topics.

Archaeology in the United Kingdom Prior to 1990, responsibility for archaeology in the United Kingdom (UK) lay largely with the Department of the Environment, which funded most work on threatened sites. Some archaeological research, normally on unthreatened sites, was undertaken by universities and local archaeological societies. In November 1990, the government introduced Planning Policy Guidance 16: Archaeology and Planning (PPG 16) (HMSO

44

44   Pete Wilson 1990), which established archaeology as a ‘material consideration’ in the planning process and offered two options for important sites and remains: ‘preservation in situ’ (the preferred option) or excavation and ‘preservation by record’. It also established the principle of ‘polluter pays’, transferring the responsibility for funding excavation in advance of destruction from the state to the developer. In 2010, PPG 16 was superseded by Planning Policy Statement 5: Planning and the Historic Environment (PPS 5) (DCLG 2010), which was, in turn, replaced in 2012 by the National Planning Policy Framework (DCLG 2012). At a national level, following the creation of the Scottish Parliament and National Assembly for Wales in the late 1990s, responsibility for the historic environment (including archaeology) lies: in England, with Historic England (a non-​departmental public body sponsored by the Department for Culture, Media, and Sport); in Wales, with Cadw (the Welsh government’s historic environment service); and, in Scotland, with Historic Scotland (an executive agency of the Scottish government). These bodies are responsible for recommending sites for statutory protection through designation as scheduled monuments and for acting as their respective government’s national advisors. However, most archaeological sites—​recorded in Historic Environment Records (HERs), which are maintained by local authorities—​are undesignated and, where threatened by development, are provided with some level of protection through the local planning process, which can demand changes to development proposals or the funding of excavations in advance of partial or complete destruction (‘preservation by record’). Development-​led archaeology in the UK is largely undertaken by archaeological units, some of which may be established as charities or trusts, while others are fully commercial operations. Other bodies active in archaeological research include some university archaeology departments, long-​established local archaeological societies, and, increasingly, local community-​based groups, which are often funded—​at least in part—​by the Heritage Lottery Fund, and which may work with professional archaeologists who can act as mentors and advisors.

How We Got to Where We are Now While this chapter is focused on Romano-​British archaeology in 2016, it is perhaps worth spending a little time identifying and summarizing those factors that have been the key drivers of change. From the perspective of this writer, a major turning point is the increasing professionalization of field archaeology; first, through the ‘Ministry of Works Excavators’ such as John Wacher, Ernest Greenfield, and Philip Rahtz, and then through the rise of rescue archaeology that widened the pool of practitioners as regional units developed (see also Millett, this volume). Without decrying the contributions of universities, museum-​based colleagues, and amateur societies all of whom had to cede space to these new players; new thinking, more funding, and a greater intensity of work all contributed to a period of rapid change.

Romano-British Archaeology in the Early Twenty-First Century    45 The next step change came with the advent of developer funding. With the best will in the world, the resources that the Department of Environment was able to deploy, even when supplemented in the 1970s and 1980s by Manpower Service Commission money, could not cope with the scale of destruction resulting from development. The advent of PPG 16 in 1990 can be seen to have set the scene in development-​related archaeology up to the present day, and the practice and issues relating to developer-​funded work have been reviewed many times (see, e.g. Everill 2007). Parallel developments included a decline in amateur fieldwork and in particular excavations—​a decline often blamed on the rise of professionalism and developer funding—​and the apparent flight of many university-​based archaeologists to projects outside the UK, driven by a perceived greater value in research terms of foreign projects and, in some cases, the lower costs of fieldwork in some countries in comparison with the UK. Therefore, by the millennium, the UK had a substantial commercial sector undertaking work that was predicated on developers’ ambitions with respect to choice of location, albeit mitigated by the work of curatorial archaeologists and the planning and designation systems (Figure 3.1). University-​led research was, by that time, more restricted than it had been when Newcastle regularly excavated on Hadrian’s Wall and

Figure 3.1  Excavation of a major Roman building by Oxford Archaeology South in advance of housing development just outside the walled town of Cirencester in 2009. Source: © Oxford Archaeology South, with thanks to Berkeley Homes for agreement to reproduce.

46

46   Pete Wilson

Figure 3.2  Volunteers excavating Building 16 at Piddington villa, Northamptonshire, in 2012. Source: © Pete Wilson.

Birmingham worked at Wroxeter over a thirty-​year period; although Mike Fulford and Reading’s ongoing association with Silchester (Fulford 1989; Fulford and Clarke 2011) and Martin Millett and Peter Halkon’s work in the East Riding (Halkon and Millett 1999; Millett 2006) serve to demonstrate that important university-​led work continues. Bleaker still was the situation with respect to amateur fieldwork; this reflected in part the generally ageing profile of amateur societies, but also the difficulties or perceived difficulties of undertaking project work. However, again there were exceptions: for example, the Monmouth Archaeological Society at Caerleon (Burnham 2001: 215)—​where University of Wales College (Newport) was also active (Burnham 2001: 315–​316)—​or the long-​running Piddington excavations (Friendship-​Taylor and Friendship-​Taylor 2011) (Figure 3.2).

Research Frameworks In the decade from 2000 to ​2010 considerable effort and resource were expended on the development of regional research frameworks and strategies, which will be considered later in this chapter. However, the recognition of a need for lacunae to be identified and priorities defined goes back to at least the latter part of the twentieth century. For the purposes of this chapter, the invitation from the Directorate of Ancient Monuments and

Romano-British Archaeology in the Early Twenty-First Century    47 Historic Buildings to the Roman Society in 1975 to form a committee to ‘advise on academic priorities with respect to the rescue archaeology of the Roman period’—​which led to the publication of Priorities for the Preservation and Excavation of Romano-​British Sites (Roman Society 1985)—​can perhaps be seen as pivotal. The priorities defined in 1985 might, by today’s standards, be seen as generalized and limited, with a cynical reader perhaps feeling that he or she could identify the research interests of each of the committee members writ large in the document. Other researchers also promoted research issues that reflected their interests; for example, Vivien Swan in her The Pottery Kilns of Roman Britain (1984) devotes a chapter to ‘Kiln Studies: Current Techniques and Future Research’, providing cogent arguments that influence her subject to this day. In 1996, English Heritage launched Frameworks for Our Past (Olivier 1996), which envisaged the creation of regional and specialist research frameworks consisting of three elements: resource assessment, research agenda, and research strategy. Intellectual ownership was intended to lie with the research community that developed the framework, and, as a consequence, the documents reflect a level of variation, with several covering the (former) government regions, while colleagues in the south-​east have opted for a series of subregional documents. At the time of writing some areas lacked full coverage: for example, the research-​strategy element of the Yorkshire framework is still to be produced. Outside the basic regional structure, Stonehenge, Avebury, and Hadrian’s Wall have been the subject of targeted volumes (AAHRG 2001; Darvill 2005; Symonds and Mason 2009)—​the Hadrian’s Wall volume being cross-​referenced with the frameworks for the north-​west and north-​east (Brennand 2006, 2007; Petts and Gerrard 2006). Similarly some county-​level frameworks have been produced, such as those for Bedfordshire (Oake et al. 2007) and Surrey (Bird 2006), as well as others for smaller areas—​notably, for the Roman period, that for Fishbourne and its environs (Manley 2008). If we look beyond England, a research framework for Wales is in place (IfA Wales/​Cymru 2008) and one for Scotland is under development (accessed 25 May 2014). Pivotal in the development of thinking about research priorities for Roman Britain across the wider profession was a session at the Roman Archaeology Conference in 1999, which led to the publication of Britons and Romans: Advancing a Research Agenda (James and Millett 2001). More recently, Historic England has developed its own Research Strategy for the Roman Period Historic Environment http://​www.english-​heritage.org.uk/​professional/​research/​strategies/​research-​strategies/​roman (accessed 25 May 2014), drawing on the available regional research frameworks and in consultation with the sector. While it is not in doubt that the development of research frameworks has sharpened thinking with respect to academic objectives in projects of all types and at all levels, provided support for bids for funding, and assisted local and national curators in sustaining arguments in favour of appropriate and academically sound mitigation strategies, it is possible to argue those benefits are perhaps not their greatest value. The process of developing research strategies, or at least those that might be regarded potential models of good practice, have generally been inclusive, bringing together archaeologists from

48

48   Pete Wilson across the sector. This has led to colleagues from universities, curatorial organizations, museums, and commercial units—​as well as independent archaeologists and those from the voluntary sector (contra Moore 2006: 3)—​cooperating and debating in a way that few, if any, other processes or mechanisms have accomplished. In that sense, the research framework process can be seen to be crucial to the rekindling of some of the links and interactions between different elements in the sector that, as suggested earlier in this chapter, have been lost or weakened during the professionalization of archaeology.

The Specialist Beyond the regional and other location-​based research frameworks, specialist groups have developed research frameworks for their area of interest; notably, for the Roman period, the Study Group for Roman Pottery (SGRP) (Perrin 2011). The SGRP document acts as a powerful voice for archaeological specialists, not only setting out priorities for Roman pottery studies but also identifying the challenges that we face as a sector in ensuring that we retain and enhance the specialist knowledge that is crucial to our subject, as the issues identified for pottery studies are equally applicable to most other areas of work. Detailed consideration of the difficulties faced by specialists in general is provided by the Survey of Archaeological Specialists 2010–​11 (Aitchison 2011), but, if we look at the Roman period, many of the issues are all too clear: an ageing cadre of specialists who are often singleton researchers; limited opportunities for mentoring or developing a new generation of specialists; a decreasing number of posts (in common with the rest of the sector); and, while we remain in the grip of a recession, less work from developer-​ funded projects.

Developer-​funded Archaeology Prior to the publication of PPG 16 in 1990, it was generally recognized that resources for fieldwork in response to development were overstretched, and, depending on the interests and/​or location of commentators, it could be argued that significant archaeology was lost with limited or even no record. The rise of commercial archaeology increased the amount of resources going into fieldwork, particularly excavation (see, e.g. Darvill and Russell 2002), but also led to some asking questions of the process, given that sites were selected by developers. What is clear is that developer funding has led to even greater emphasis on the site (normally defined by the limits of the development) and, in extreme cases, to an atomization of the resource, with sites either side of a modern boundary feature being excavated by competing contracting units with little or no cooperation or exchange of ideas. Equally, the practice of different phases of the same project (evaluation, excavation, and post excavation) sometimes being undertaken by different

Romano-British Archaeology in the Early Twenty-First Century    49 organizations as a result of being separate tendered can further exacerbate the difficulties and, potentially, lead to an incomplete understanding of what has been excavated or destroyed. A further perception is that difficulties can arise from the willingness of many commercial units to range over a wide area in pursuit of work, fuelling an oft-​ repeated concern with respect to a potential lack of local knowledge. The argument is that operating in different parts of the country could lead to a lack of understanding of the local context, which could hamper the development of a project design appropriate to the circumstances of the site, impede understanding of the emerging results, and result in flawed analyses and interpretations in post-​excavation. These issues have been debated widely, and, in addition to Everill’s paper (2007), attention can be drawn to Tom Moore’s consideration of PPG 16 and cultural landscapes, wherein he suggests: ‘we are in danger of having archaeological landscapes made up of “dots” representing archaeological evidence, often of a fragmentary nature, which are not understood as part of a whole’ (Moore 2006: 5). The potential of developer-​funded work has been demonstrated on numerous occasions and in many circumstances, with site monographs from key urban centres making significant contributions to our understanding of Romano-​British sites. Examples include Causeway Lane, Leicester (Connor and Buckley 1999), or the more recent series of monographs on sites in London from Museum of London Archaeology Service/​ Museum of London Archaeology, Pre Construct Archaeology, Wessex Archaeology, and others. Similarly, in the context of the rural areas of Roman Britain, while individual site reports may enlighten us about the specifics of individual sites (for example, the recent volume on Faverdale, Darlington (Proctor 2012)), perhaps our greatest advances have come from work on linear schemes such as roads and pipelines. Taking Yorkshire as an example, the combined data for the Iron Age and Romano-​British periods from the M1–​A1 Link Road (Roberts et al. 2001), Holmfield/​Ferrybridge Interchange (Roberts 2005), and the A1(M) Darrington to Dishforth scheme (Brown et al. 2007), represent an unprecedented and highly informative sample of the rural landscape in an area where other major projects are working towards publication. As samples of landscape examined in detail and analysed and published at length (with substantial discussion sections occupying between 10 and 20 per cent of the volumes), they complement and inform extensive survey projects, such as those undertaken on the crop marks of the Magnesian Limestone (Roberts et al. 2010) or the Swale–​Ure Washlands (Bridgland et al. 2011) and funded through the Aggregates Levy Sustainability Fund. The wide range of projects undertaken and the number of contractors undertaking work with developer funding make it rather invidious to pick out ‘highlights’, but it is perhaps necessary to cite some examples. Looking to the south of England, Oxford Archaeology’s work at Lankhills (Booth et al. 2010) has challenged our understanding of a site that has figured large in the literature of late Roman Britain. Similarly, the work of Cotswold Archaeology (formerly Cotswold Archaeological Trust) over many years and increasingly through developer-​funded work has, with others, built on the work of the Cirencester Excavation Committee (see, e.g. Holbrook 1998, 2008). It would be possible to go on, but in the context of this chapter it is clear that commercially funded projects

50

50   Pete Wilson of all sizes, where the results are appropriately analysed and disseminated, are adding substantially to our understanding of most aspects of Roman Britain. On a less positive note, Moore (2006) addresses the issue of grey literature and the lack of integration of the results of commercially funded work into academic and other syntheses. Since Moore’s paper was published, we have had Richard Bradley’s ground-​breaking study (2007), which sought to integrate the evidence for prehistoric periods from the grey literature with pre-​existing data. For the Roman period, we have already seen the results of the first two phases of the Roman Grey Literature Project (Holbrook 2010a, b; Fulford and Holbrook 2011; Hodgson 2012). Reading University and Cotswold Archaeology, with funding from Leverhulme Trust and English Heritage, are now engaged in the third phase of the project: an England-​wide study of data from grey literature and fully published reports intended to produce a new account of the rural settlement of Roman Britain.

Non-​publication Recognition of the potential of development archaeology is nothing new. For many years the Ministry of Works and its successors (the Department of the Environment (Ancient Monuments Branch) and, most recently, Historic England) have—​through targeted support for ‘unexpected discoveries’—​ put considerable resources into responding to threats posed by development, including the work of the ‘Ministry of Works Excavators’. The 1970s are viewed by many as the ‘heroic age’ of rescue archaeology that led to the development and perceived stability of regional archaeology units and, subsequently, to the present-​day commercially led situation, as already discussed. However, all too often the promise of the fieldwork was not fully realized, projects were not adequately analysed or published, and the phenomenon known as ‘backlog’ was on us. The problem also extended to much of the work done during the Manpower Service Commission-​funded fieldwork boom in the late 1970s. Many of the publications from that time (of both Roman and other sites) were a product of the individual determination of those responsible; but all too often important work remains largely unknown and substantially inaccessible. The issue is particularly difficult with respect to some of our most important towns and cities, although work on key rural sites also remains unpublished. Set against that situation, the issues around grey literature already discussed seem almost esoteric. The information in grey literature, for all the difficulties of using it, could be accessed through HERs or the Archaeology Data Service (ADS) if the researchers whose work might benefit from it were to take the trouble to seek it out. That said, after the replacement of PPG 16 with Planning Policy Statement 5 (PPS 5), it would now be difficult to justify non-​publication of projects while the associated Practice Guide remains in force, as it states (in the specification for the ‘content of a written scheme of investigation’) that it should contain: ‘proposals for the preparation and publication of a suitable report on the investigation, its results and the advancement in understanding that those results bring’ (DCLG 2010: 38). It is hoped that assurances

Romano-British Archaeology in the Early Twenty-First Century    51 that the force of PPS 5 will survive the implementation of the National Planning Policy Framework (DCLG 2012) are borne out.

Research for its Own Sake While it has to be accepted that all archaeology, at least if well done and appropriately disseminated, contributes to our collective knowledge and/​or academic discourse, in times of resource pressure those bodies responsible for managing the archaeological resource (be they national agencies, curatorial archaeologists, or organizations that own the resource such as the National Trust) are increasingly forced to focus on what is needed for management purposes or what can be regarded as ‘reasonable’ in terms of a brief for a development-​related site. Although academic knowledge will be gained, it may also be constrained in terms of the resources available for analysis (if a commercial project) or limited by the project design (if driven by management questions). One example with a positive benefit, the recent Miner–​Farmer Landscapes Project in the North Pennines (an Area of Outstanding Natural Beauty), sought to examine environmental threats from erosion and climate change and, as part of that study, has produced a re-​examination of Whitley Castle Roman fort and the areas surrounding it (Ainsworth and Went 2009). While a level of ongoing engagement of university-​based colleagues has already been acknowledged, it is possible to argue that, in recent years, there has been re-​engagement on the part of universities with the archaeology of Roman Britain. This suggestion is prompted, in part, by the development of a number of high-​profile projects, including those undertaken by the University of Newcastle within the Hadrian’s Wall World Heritage Site (initially with Historic England), and work by Durham University at Binchester, the University of Cambridge at Aldborough, and the University of Nottingham at Caistor St Edmund. It is notable that all those projects have included opportunities for non-​student volunteers, delivering that better engagement referred to earlier in the chapter. Given the success of the South Shields and Vindolanda volunteer programmes—​which demonstrate an international enthusiasm for Roman Britain—​it is perhaps surprising that there are not more joint projects in the UK along the lines of the Binchester partnership involving Stanford University (see (accessed 13 July 2012)).

Community Archaeology, the Media, and Key Stage 2 Public engagement with archaeology in all its forms has never been higher, and, in part, this reflects the profile of archaeology on television. In the UK, the success of

52

52   Pete Wilson programmes such as Meet the Ancestors and Time Team made archaeology mainstream entertainment, and Roman-​period sites seem to have a particular fascination. In part this may simply reflect the fact that Roman-​period structures can be quite ‘tele-​visual’, often being constructed of masonry and frequently associated with substantial numbers of objects, many of which are readily recognizable and visually attractive. The fact that the Roman period figures in the national curriculum in Key Stage 2 may well also influence the thinking of programme-​makers. Roman-​period stories play well with news media, often for the same reasons of good visual appeal and recognizable material culture, although the latter is too often reduced (at least in the headlines) to a focus on ‘value’ expressed in purely financial terms, and the media may not always treat our subject in ways with which we are comfortable! The Daily Mail (22 May 2002) reported our publication of Catterick (Wilson 2002 a, b) with the headline ‘Meet Crossus Dressus—​ our first transvestite’, in reference to Hilary Cool’s suggestion that a burial might represent a gallus (a priest of Cybele). Since the 1950s, when Sir Mortimer Wheeler and Glyn Daniel were unlikely media stars, archaeologists have grappled with the conflict between the necessity of working with the media to popularize and ‘make the case for’ our subject and the charges of ‘dumbing down’ from colleagues suspicious of possibly simplistic ‘popular’ treatments of often complex issues. These arguments continue to the present day (see, e.g. Clack and Brittain 2007; Whittington 2010). The attributes of Roman-​period archaeology that appeal to the media also make it attractive to amateur, community, and public archaeology projects. The value of the contribution of the voluntary sector is difficult to gauge because much of the work is almost invisible at a national level (and also locally in some cases), even if the groups themselves are not unknown: for example, the Council for British Archaeology (CBA) has identified the existence of in excess of 2,000 community archaeology groups (Thomas 2010). Good relationships exist between many professionals and non-​professional groups; in many cases Heritage Lottery funding or other sources facilitate direct cooperation and, where necessary, mentoring. In terms of the present-​day archaeological landscape, professional archaeologists and academics need to engage with the enthusiasm that the voluntary sector represents. This is not simply as a potential ‘project resource’ but, more pressingly, to ensure that the results that derive from voluntary-​sector projects are fed into the literature and other resources available to researchers and also to offer, where needed, support to ensure that those results are of a standard that represents a return of lasting value for the work and enthusiasm expended. PPS 5 has pushed the profession as a whole towards better and higher-​quality engagement with the public; our challenge is to deliver that without compromising the standards and academic integrity that we value so highly. The nature of many Roman sites and associated collections provides us with an almost unrivalled ‘marketing tool’ for our subject. The significance of Hadrian’s Wall as an icon of the past cannot be overemphasized; along with Stonehenge, it may be seen to represent ‘archaeology’ in the UK. The ability of Hadrian’s Wall to engage the wider public is demonstrated by its popularity as a tourist draw, but, more significantly in the context of this chapter, the wish of people to engage further is difficult to overstate. For example,

Romano-British Archaeology in the Early Twenty-First Century    53

Figure 3.3  Volunteers excavating the sequence of barracks in the north-​west quadrant of the fort at Vindolanda in 2011. Source: © The Vindolanda Trust.

the long-​term association of South Shields with Earthwatch brings paying international volunteers into industrial Tyneside (accessed 13 July 2012), while all 650 places for paying voluntary diggers during a recent season at Vindolanda (Figure 3.3) were filled in less than three hours (accessed 13 July 2012).

Portable Antiquities Scheme (PAS) One area where amateurs dominate is amongst metal detectorists, and it is not necessary here to re-​examine the history of the often troubled relationships between archaeologists and detectorists. The looting of Wanborough Roman temple (Surrey) and the continued depredations at Icklingham Roman town (Suffolk) are well known (Thomas 2009; Wilson 2015) and serve to illustrate the vulnerability of Roman-​period sites to illicit detecting. On a more positive note, responsibly detected and appropriately recorded metal-​detector finds are adding much to our knowledge, particularly with respect to numismatic studies (see, e.g. Walton 2012). Detectorists are also revealing the existence of previously unknown sites and, where these are responsibly detected,

54

54   Pete Wilson the finds recorded, and the data passed to the appropriate Historic Environment Record, they are contributing significantly to research (see, e.g. Brindle 2009). For example, the Portable Antiquities Scheme (PAS) records 90 Masters dissertations and 63 Ph.D. theses either in progress or completed using Scheme data (accessed 7 July 2012). However, perhaps only 40 per cent of detectorists record their finds with the PAS. Monitoring of eBay by PAS has revealed significant levels of non-​reporting of potential treasure finds, which there is a clear duty in law to report; so one has to ask: ‘what is being lost unrecorded?’ In addition, detectorists often argue that they are ‘saving the heritage’ by removing metalwork from the soil where it is under attack from fertilizers and other chemicals, citing the poor state of recent finds compared to those made twenty or so years earlier in the same location. While it is probable that agricultural and other chemicals are having an adverse effect on buried material culture, it may also be the case that sites are getting ‘detected out’, with only the poorly preserved material remaining after decades of detecting. Metal detectors are not going to go away, but they are contributing quite legally to the erosion of the archaeological resource, particularly when finds are not recorded. This presents an ongoing challenge to archaeologists. Should we encourage and support ‘responsible metal detecting’ while also demanding greater levels of recording, or risk seeing our common past—​of all metal-​using periods—​disappear into the antiquities trade or private collections? Even given the success of the PAS, legal metal detecting provides other challenges to archaeologists and museum professionals, not least the pressure to acquire key finds for public display. While not necessarily taking resources directly away from curatorial work or archaeological projects, the staff time and effort required to acquire finds such as the Frome hoard of Roman coins (Moorhead et al. 2010), which under the Treasure Act process was valued at £320,250 with a further £100,000 needed for conservation, represents a major challenge. The state of museum funding (particularly acquisition budgets) means that many museums may struggle to find the resources to acquire even single locally found objects that would enhance their collections and displays.

Approaches and Techniques The advent of the metal detector is only one of many innovations of which archaeologists of all types have had to take account in recent decades. The introduction of new technologies and developments has made immense and far less controversial contributions. The potential of aerial photography has long been known and, through the work of Professor St Joseph, David Wilson, and many others, Romano-​British sites have provided some of its most productive targets for many years, with many new sites identified and much new detail about known sites added. Since 1988, large-​scale intensive mapping of air photographs by the English Heritage National Mapping Programme and others has

Romano-British Archaeology in the Early Twenty-First Century    55 added immensely to our stock of knowledge. For example, taking the National Mapping Programme alone, in excess of 40 per cent of England had been mapped by 2012 (accessed 25 May 2014), which allows present-​day researchers widened opportunities to think at a landscape level. Furthermore, the potential to combine the aerial photographic evidence with satellite and multi-​spectral imagery, LIDAR data (accessed 25 May 2014) and large-​ scale geophysical surveys (see Powlesland 2009; (accessed 25 May 2014), provides the potential for complementary datasets, which can give previously unimaginable detail (Figure 3.4). The contributions of individual remote-​sensing techniques are well known in the literature: for example, the extensive geophysical surveys on the forts and vici of the Hadrian’s Wall World Heritage Site, at Birdoswald (Biggins and Taylor 1999; 2004a), Maryport (Biggins and Taylor 2004b), and elsewhere, or those undertaken as part of the Cadw-​sponsored Roman Forts Environs Project (see Hopewell 2005). As already suggested, the use of aerial photography in combination with geophysical survey has revolutionized our knowledge of many sites and their context (see, e.g. Wilson 2002a:  34–​45) for work designed to contextualize excavation data). Furthermore, projects designed from the start to utilize a range of remote-​sensing techniques in combination with surface collection, excavation, and geographical and topographic modelling, as at Wroxeter (Gaffney and White 2007) or in the Vale of Pickering (Powlesland et al. 2006), allow a further step change in understanding and provide models for future work.

Figure 3.4  Combined geophysical surveys undertaken by the Landscape Research Centre with the support of Historic England and the Aggregates Levy Sustainability Fund on the southern side of the Vale of Pickering, North Yorkshire; depicts more than 1.5 kilometres of late Iron Age and Roman trackside ribbon development or ‘ladder settlement’ in addition to a wealth of other prehistoric and post Roman features. Source: © Landscape Research Centre.

56

56   Pete Wilson

Developing Understanding Having moved on from being a subject dominated and, some would say, constrained by the ‘great and the good’, Romano-​British archaeology in the second decade of the twenty-​first century is in an interesting position. We undoubtedly have a new generation of ‘greats’, but one perhaps less dominating and more open to debate. The appearance of books and papers that challenge or question old orthodoxies, in whole or in part, is now common, and those new works and their authors are likely to be met with questions and searching reviews from colleagues (see, e.g. Freeman (1993) on Martin Millett’s The Romanization of Britain (1990)), rather than facing offhand dismissal as ‘upstarts’ by those whose opinions they challenge. That, viewed with the fact that there is space in the literature for a volume running to over 600 pages that seeks to present ‘an experimental, speculative and heretical’ view of Roman Britain (Mattingly 2006: p. xv), suggests strongly a subject that has matured nicely. This openness to new ideas may, in part, reflect the success of the Theoretical Roman Archaeology Conference (TRAC) from its inception in 1991 in providing exposure for new ideas, different approaches, and (mostly) younger researchers. Writers such as Mattingly (2006: 491–​528) and Taylor (2007) beat the drum for the recognition of diversity within Roman Britain. On one level this is not new—​as Mike Fulford (2007) states in his review of Mattingly (2006), where he points to the Peoples of Roman Britain series published in the 1970s and 1980s by Duckworth—​and, in addition, regional variation in aspects of material culture has long been recognized. However, what we are seeing now is the presentation of a more nuanced and complex picture that is beginning to show variables operating at a local level. Such understanding can be enhanced only by the full incorporation by researchers of all types of data, including that held by the PAS and data derived from voluntary-​sector work and commercially led interventions. Roman Britain continues to spring surprises, such as the discovery or recognition in the decade to 2010 of a further three Roman military sites at Calstock, Lostwithiel, and near St Austell, in Cornwall (Ordnance Survey 2011); overturning long-​held academic belief but seemingly having little impact on popularly accepted ‘facts’—​hence ‘The Romans never settled here’ (accessed 25 May 2014). Furthermore, Roman Britain is, or at least those studying Roman Britain are, gradually reconnecting with the wider empire. The Limes Congress has, since 1949, brought together archaeologists working on Roman military sites, and related topics and scholars of Roman Britain have figured large in the various Congress programmes. This may be a result of the obvious research potential of similarities and differences between military sites and broad issues of military strategy, deployment, and supply across the empire. In contrast, for many years British archaeologists working on the civilian archaeology of Roman Britain were apparently less willing to make those linkages and engage with colleagues working on

Romano-British Archaeology in the Early Twenty-First Century    57 similar subjects elsewhere in the Roman Empire. That said, the fact that there were exceptions among excavators and others is accepted, as is the strength of international cooperation on particular topics such as the Corpus Inscriptionum Latinarum or terra sigillata. The reasons behind this more limited engagement are debatable, but for decades it appeared that most archaeologists of Roman Britain were talking to themselves, or rather to like-​minded colleagues whose thinking was insular by inclination. By 1966, Sir Mortimer Wheeler saw ‘Romano-​British archaeology [as having] reached an uncommonly high level of technical excellence, but [having] diverted into a relatively insignificant channel talent which might have been more amply employed’ (Freeman 2007: 625). While we might debate vigorously the insignificance of our study, I for one would accept that in seeking to better understand a minor province on the outer edge of the empire, archaeologists of Roman Britain need to engage more closely with those working on other provinces: twenty-​first-​century Roman Britain cannot be an insular study. However, engagement is happening increasingly through contacts with foreign colleagues and also through greater direct cooperation and the internationalization of key conferences such as the Roman Archaeology Conference (RAC) and Theoretical Roman Archaeology Conference (TRAC). While there have often been non-​UK speakers at both conferences, the holding of the conferences outside the UK (for example, TRAC 2008 in Amsterdam, RAC/​TRAC in Ann Arbor in 2009 and Frankfurt in 2012) promotes greater exposure of UK archaeologists to a much wider archaeological community and to a broad range of work and approaches, both academic and in terms of field procedures.

Challenges for the Future We live in uncertain times, and across all parts of the historic environment sector we are seeing unprecedented change. We cannot yet know what the long-​term impact of changes to university fees will mean for individual archaeology departments and for the expectations of students still drawn to the subject. Similarly, the implications of the Government’s 2010 comprehensive spending review on national bodies are still being worked through and at the time of writing the full impacts are as yet unclear, and, in a climate where further cuts are possible, we face continuing uncertainty. There is also uncertainty as to the impact of the proposed Welsh Heritage Bill or the possible merger of the Royal Commission on the Ancient and Historical Monuments of Scotland with Historic Scotland (accessed 13 July 2012). In England, as elsewhere, local authorities are under immense pressure from austerity measures, and it is Historic Environment Services, museums, and other non-​statutory services that are bearing much of the pressure. Romano-​British archaeology is not affected by this any more than the archaeology of any other period, but we are possibly looking at a future

58

58   Pete Wilson with a very different curatorial regime that will potentially have severe consequences for all periods. In conclusion, it is difficult to try and predict where Romano-​British archaeology is going—​at least in the short term—​given the issues that have been outlined. However, there is a sense that our subject (while suffering along with the rest of the heritage sector) has a resilience that will ensure it retains its popularity with the public—​as providing potentially good television, evocative places to visit, and a strong focus for community and public archaeology. Those in academic posts, unit-​based archaeologists, and professionals in curatorial roles can perhaps only ‘wait and see’ as to how university departments fare, what impact the National Planning Policy Framework and the Localism Act in England have on development-​led archaeology, and what the future prognosis for funding is. The lights may be flickering, but, hopefully, whatever reinventions the sector goes through, new and exciting work will emerge in the commercial sector as development returns and resources will be found to continue and develop both universityand community-​based field projects focused on our Roman past. Academically, we are stronger than ever, but it is possible that, across the sector, we need to demonstrate that Britannia is far from being a ‘grand old lady’ that has ‘ceased to matter as a major historical problem’ (Wheeler 1961: 159).

References AAHRG (Avebury Archaeological and Historical Research Group) (2001). Archaeological Research Agenda for the Avebury World Heritage Site. Salisbury: Trust for Wessex Archaeology (accessed 25 May 2014). Ainsworth, S., and Went, D. (2009). ‘Whitley Castle, Tynedale, Northumberland:  An Archaeological Investigation of the Roman Fort’, English Heritage: Research Department Report 89/​2009. Aitchison, K. (2011). Survey of Archaeological Specialists 2010–​11. Sheffield: Landward Research (accessed 25 May 2014). Biggins, J. A., and Taylor, D. J.  A. (1999). ‘A Survey of the Roman Fort and Settlement at Birdoswald, Cumbria’, Britannia, 30: 91–​110. Biggins, J. A., and Taylor, D. J.  A. (2004a). ‘Geophysical Survey of the vicus at Birdoswald Roman Fort, Cumbria’, Britannia, 35: 159–​178. Biggins, J. A., and Taylor, D. J. A. (2004b). ‘The Roman Fort and vicus at Maryport: Geophysical Survey, 2000–​2004’, in R. J. A. Wilson and I. D. Caruna (eds), Romans on the Solway. Essays in Honour of Richard Bellhouse. Cumberland and Westmorland Antiquarian and Archaeological Society Extra Series 31. Kendal: Trustees of the Senhouse Roman Museum, Maryport, 102–​133. Bird, D. (2006). Surrey Archaeological Research Framework 2006. Kingston on Thames and Guildford:  Surrey County Council and Surrey Archaeological Society (accessed 25 May 2014). Booth, P., Simmonds, A., Boyle, A., Clough, S., Cool, H. E. M., and Poore, D. (2010). The Late Roman Cemetery at Lankhills, Winchester:  Excavations 2000–​2005. Oxford Archaeology Monograph 10. Oxford: Oxford Archaeology.

Romano-British Archaeology in the Early Twenty-First Century    59 Bradley, R. (2007). The Prehistory of Britain and Ireland. Cambridge:  Cambridge University Press. Brennand, M. (2006) (ed.). ‘The Archaeology of North-​West England:  An Archaeological Research Framework of North-​West England. Volume 1: Resource Assessment’, Archaeology North-​West, 8/​18  (accessed 25 May 2014). Brennand, M. (2007) (ed.). ‘Research and Archaeology in North-​ West England:  An Archaeological Research Framework of North-​West England. Volume 2: Research Agenda and Strategy. Archaeology North-​West, 9/​19  (accessed 25 May 2014)  and (accessed 25 May 2014). Bridgland, D., Innes, J., Long, A., and Mitchell, W. (2011). Late Quaternary Landscape Evolution of the Swale-​Ure Washlands, North Yorkshire. Oxford: Oxbow Books. Brindle, T. (2009). ‘Amateur Metal Detector Finds and Romano-​ British Settlement: A Methodological Case Study from Wiltshire’, in M. Driessen, S. Heeren, J. Hendriks, F. Kemmers, and R. Visser (eds), TRAC 2008: Proceedings of the Eighteenth Annual Theoretical Roman Archaeology Conference, Amsterdam 2008. Oxford: Oxbow Books, 53–​72. Brown, F., Howard-​Davis, C., Brennand, M., Boyle, A., Evans, T., O’Connor, S., Spence, A., Heawood, R., and Lupton, A. (2007). The Archaeology of the A1 (M)  Darrington to Dishforth DBFO Road Scheme. Lancaster Imprints 12. Lancaster:  Oxford Archaeology North. Burnham, B. C. (2001). ‘Roman Britain in 2000. 1.  Sites Explored. 1.  Wales’, Britannia, 32: 312–​318. Clack, T., and Brittain, M. (2007). Archaeology and the Media. Walnut Creek, CA: University College London Institute of Archaeology Publications. Connor, A., and Buckley, R. (1999). Roman and Medieval Occupation in Causeway Lane, Leicester. Leicester Archaeology Monograph Series 5. Leicester: Leicester Archaeology. Darvill, T. (2005). Stonehenge World Heritage Site:  An Archaeological Research Framework

​ (accessed 25 May 2014). Darvill, T., and Russell, B. (2002). Archaeology after PPG 16: Archaeological Investigations in England 1990–​1999. Bournemouth University School of Conservation Sciences Research Report 10. Bournemouth and London: Bournemouth University/​English Heritage. DCLG (Department of Communities and Local Government) (2010). PPS5: Planning for the Historic Environment Practice Guide. London: HMSO. DCLG (Department of Communities and Local Government) (2012). National Planning Policy Framework. London: HMSO. Everill, P. (2007). ‘British Professional Field Archaeology:  Antiquarians and Labourers; Developers and Diggers’, in Y. Hamilakis and P. Duke (eds), Archaeology and Capitalism: From Ethics to Politics. Walnut Creek, CA: Left Coast Press, 158–​187. Freeman, P. W. M. (1993). ‘“Romanisation” and Roman Material Culture’, Journal of Roman Archaeology, 6: 438–​445. Freeman, P. W. M. (2007). The Best Training-​Ground for Archaeologists. Oxford: Oxbow Books. Frere, S. S. (1974). Britannia: A History of Roman Britain. London: Cardinal. Friendship-​ Taylor, R. M., and Friendship-​ Taylor, D. E. (2011). Iron Age and Roman Piddington: Ninth Interim Report. Piddington: Upper Nene Archaeological Society.

60

60   Pete Wilson Fulford, M. (1989). The Silchester Amphitheatre. Excavations of 1979–​85. Britannia Monograph Series 10. London: Roman Society. Fulford, M. (2007). ‘An Insular Obsession’, Britannia, 38: 367–​369. Fulford, M., and Clarke, A. (2011). Silchester: City in Transition. The Mid-​Roman Occupation of Insula IX c.AD 125–​250/​300: A Report on Excavations undertaken since 1997. Britannia Monograph Series 25. London: Roman Society. Fulford, M. G., and Holbrook, N. (2011). ‘Assessing the Contribution of Commercial Archaeology to the Study of the Roman Period in England, 1990–​2004’, Antiquaries Journal, 91: 1–​23. Gaffney, V. L., and White, R. H. (with Goodchild, H.) (2007). Wroxeter, the Cornovii, and the Urban Process:  Final Reports on the Wroxeter Hinterland Project 1994–​1997. Volume 1:  Researching the Hinterland. JRA Supplementary Series 68. Portsmouth, RI:  Journal of Roman Archaeology. Halkon, P., and Millett, M. (1999). Rural Settlement and Industry:  Studies in the Iron Age and Roman Archaeology of Lowland East Yorkshire. Yorkshire Archaeological Report 4. Leeds: Yorkshire Archaeological Society. HMSO (Her Majesty’s Stationery Office) (1990). Planning Policy Guidance 16: Archaeology and Planning. London: HMSO. Hodgson, N. (2012). ‘Assessing the Contribution of Commercial Archaeology to the Study of Roman South and West Yorkshire, 1990–​2004’, Yorkshire Archaeological Journal, 84: 38–​58. Holbrook, N. (1998) (ed.). Cirencester: The Roman Town Defences, Public Buildings and Shops. Cirencester Excavations V. Cirencester: Cotswold Archaeological Trust. Holbrook, N. (2008) (ed.). Excavations and Observations in Roman Cirencester 1998–​2007. Cirencester Excavations VI. Cirencester: Cotswold Archaeological Trust. Holbrook, N. (2010a). ‘The Contribution of Commercial Archaeology to the Study of Roman Somerset, 1990–​2004’, Proceedings of the Somerset Archaeological and. Natural History Society, 154: 33–​50. Holbrook, N. (2010b). ‘The Contribution of Commercial Archaeology to the Study of Roman Essex, 1990–​2004’, Essex Archaeology and History (Fourth series), 1: 1–​15. Hopewell, D. (2005). ‘Roman Fort Environs in North-​West Wales’, Britannia, 34: 225–​269. IfA Wales/​Cymru (2008). A Research Framework for the Archaeology of Wales (accessed 25 May 2014). James, S., and Millett, M. (2001) (eds). Britons and Romans:  Advancing an Archaeological Agenda. CBA Research Report 125. York: Council for British Archaeology. Manley, J. (2008) (ed.). The Archaeology of Fishbourne and Chichester:  A  Framework for its Future. Lewes:  Sussex Archaeological Society (accessed 25 May 2014). Mattingly, D. (2006). An Imperial Possession:  Britain in the Roman Empire, 54 BC–​AD 409. London: Allen Lane. Millett, M. (1990). The Romanization of Britain:  An Essay in Archaeological Interpretation. Cambridge: Cambridge University Press. Millett, M. (2006) (ed.). Shiptonthorpe, East Yorkshire: Archaeological Studies of a Romano-​ British Roadside Settlement. Yorkshire Archaeological Report 5.  Leeds:  Yorkshire Archaeological Society. Moore, T. (2006). Following the Digger:  The Impact of Developer-​Funded Archaeology on Academic and Public Perceptions of Cultural Landscapes. Newcastle: Cultural Landscapes in

Romano-British Archaeology in the Early Twenty-First Century    61 the 21st Century. Forum, UNESCO, University and Heritage 10th International Seminar, International Centre for Cultural Heritage Studies. Newcastle:  International Centre for Cultural Heritage Studies. Moorhead, S., Booth, A., and Bland, R. (2010). The Frome Hoard. London:  British Museum Press. Oake, M., Luke, M., Dawson, M., Edgeworth, M., and Murphy, P. (2007). Bedfordshire Archaeology. Research and Archaeology: Resource Assessment, Research Agenda and Strategy. Bedfordshire Archaeology Monograph 9. Bedford: Bedfordshire Archaeology (accessed 25 May 2014). Olivier, A. C. H. (1996). Frameworks for our Past: A Review of Research Frameworks, Strategies and Perceptions. London: English Heritage. Ordnance Survey (2011). Roman Britain Historical Map. 6th edn. Southampton:  Ordnance Survey. Perrin, R. (2011). A Research Strategy and Updated Agenda for the Study of Roman Pottery in Britain. Study Group for Roman Pottery Occasional Paper 1  (accessed 25 May 2014). Petts, D., and Gerrard, D. (2006). Shared Visions: The North-​East Regional Research Framework for the Historic Environment. Durham: Durham County Council (accessed 25 May 2014). Powlesland D. (2009). ‘Why Bother? Large Scale Geomagnetic Survey and the Quest for “Real Archaeology”’, in S. Campana and S. Piro (eds), Seeing the Unseen: Geophysics and Landscape Archaeology. Abingdon: Taylor and Francis, 167–​182. Powlesland, D., Lyall, J., Hopkinson, G., Donoghue, D., Beck, M., Harte, A., and Stott, D. (2006). ‘Beneath the Sand: Remote Sensing, Archaeology, Aggregates and Sustainability: A Case Study from Heslerton, the Vale of Pickering, North Yorkshire, England’, Archaeological Prospection, 13/​4: 291–​299. Proctor, J. (2012). Faverdale, Darlington: Excavations at a Major Settlement in the Northern Frontier Zone of Roman Britain. London: Pre-​Construct Archaeology. Roberts, I. (2005) (ed.). Ferrybridge Henge: The Ritual Landscape. Archaeological Investigations on the Site of the Holmfield Interchange of the A1 Motorway. Yorkshire Archaeology 10. Leeds: Archaeological Services WYAS. Roberts, I., Burgess, A. and Berg, D. (2001) (eds). A New Link to the Past: The Archaeological Landscape of the M1–​A1 Link Road. Yorkshire Archaeology 7.  Leeds:  West Yorkshire Archaeology Service. Roberts, I., Deegan, A., and Berg, D. (2010). Understanding the Cropmark Landscapes of the Magnesian Limestone. Leeds: Archaeological Services WYAS. Roman Society (1985). Priorities for the Preservation and Excavation of Romano-​British Sites. London: Roman Society. Selkirk, A. (2010). ‘Who Champions the Amateur?’, Current Archaeology, 247: 48–​49. Swan, V. G. (1984). The Pottery Kilns of Roman Britain. RCHME Supplementary Series 5. London: HMSO. Symonds, M. F.  A., and Mason, D. J.  P. (2009) (eds). Frontiers of Knowledge:  A  Research Framework for Hadrian’s Wall. Durham: Durham County Council and Durham University

(accessed 25 May 2014). Taylor, J. (2007). An Atlas of Roman Rural Settlement in England. CBA Research Report 151. York: Council for British Archaeology.

62

62   Pete Wilson Thomas, S. (2009). ‘Wanborough Revisited:  The Rights and Wrongs of Treasure Trove Law in England and Wales’, in S. Thomas and P. G. Stone (eds), Metal Detecting and Archaeology: ‘Heritage Matters’ [Volume 2]. Woodbridge: Boydell, 153–​165. Thomas, S. (2010). Community Archaeology in the UK: Recent Findings. Council for British Archaeology (accessed 25 May 2014). Walton, P. J. (2012). Rethinking Roman Britain:  Coinage and Archaeology. Wetteren, Belgium: Moneta 137. Wheeler, R. E. M. (1961). Review of: The Journal of Roman Studies, 50 (1960) Jubilee Volume, Antiquity, 35 (138): 157–​159. Whittington, K. (2010). ‘Through a Glass Lens Darkly’, Archaeologist, 77 (Autumn): 41. Wilson, P. (2002a). Cataractonium:  Roman Catterick and its Hinterland. Excavations and Research 1958–​1997: Part 1. CBA Research Report 128. York: Council for British Archaeology. Wilson, P. (2002b). Cataractonium:  Roman Catterick and its Hinterland. Excavations and Research 1958–​1997: Part 2. CBA Research Report 129. York: Council for British Archaeology. Wilson, P. (2015). ‘Metal Detectors: Friends or Foes?’, in P. Everill and P. Irving (eds), Rescue Archaeology 40 Years On. Hertford: Rescue, 159–172.

Chapter 4

The Devel opme nt of Artefact St u di e s Ellen Swift

Introduction This chapter surveys the evolution of artefact studies from the antiquarian period to the early twenty-first century. Following a brief overview of antiquarian interest in objects and collecting, I investigate the place of artefact studies within the initial development of archaeological methodology during the later nineteenth century. The classification of objects and the development of find reports within published site excavation reports are then considered, followed by an analysis of the main trends in scholarly interpretation. Each section is informed by a consideration of the brief biographical details of scholars engaged in artefact research in order to place trends in artefact studies within a wider social perspective. Particular themes relating to data presentation and interpretation and the status of artefact studies are then discussed at more length. As will become apparent, the position of artefact research in archaeology results from its complex historical development alongside other aspects of archaeological method and interpretation, and is influenced not only by trends in academic research but by a range of other factors, including the lingering influence of antiquarian and early twentieth-​century studies, the status and profile of artefact researchers, and the wider conditions of academia and museums in the later twentieth century.

The Antiquarian Interest in Artefacts Antiquarian interest in artefacts was very much the interest of the collector. Following the founding of the Society of Antiquaries, early members sometimes met to discuss items in their personal collections (Herendeen 2007: 90), although initially there was little in the nature of systematic description and recording. In general, collections

64

64   Ellen Swift were private ones, and antiquarians were hampered by their lack of access to artefacts beyond their own objects. Although information could be exchanged with other scholars, established connections with other collectors, access to their finds, and an ability to travel were essential in developing artefact research. In an early attempt to disseminate knowledge (also followed by others (C. Evans 2007: 274)), the late eighteenth–​early nineteenth-​century French numismatist Mionnet made and sold cast copies of Roman coins (Babelon 1901: 200). The illustration of artefacts, however, was already increasing during the seventeenth century (Hingley 2008: 40). Watercolours and engravings were the principal methods used until the introduction of photographs in the mid-​ nineteenth century (Smiles 2007: 124). Eighteenth-​century collectors and antiquarians, such as Richard Warner, who published a collection of antiquities from Roman Bath (Warner 1797), focused on inscriptions and coins, as well as the more large-​scale and spectacular finds such as mosaic floors and pieces of sculpture. Above all, they were interested in items that could be linked to Roman history (Kendrick 1950:  166), yet the tendency to see artefacts through the prism of ancient history led to some spectacular misconceptions. Faussett (1856 [1773]), for instance, thought Anglo-​Saxon artefacts were Roman. Thanks to the Western admiration for Classical culture, however, some Roman artefacts, such as coins, were well documented from comparatively early on: for example, the substantive publication of coin collections during the eighteenth century across Europe (Babelon 1901). The discovery of sites such as Pompeii and Herculaneaum was an added stimulus to eighteenth-​century interest in Roman remains (Hingley 2008: 233–​4). The initial focus on historical artefacts and their existence as tangible links to the dramas of past times was similar to the value attached more widely by antiquarians to ‘curiosities’ as a prop for storytelling (Crane 1999: 187). As a consequence of the history of collecting, artefact studies played a comparatively large role in research: for instance, forming about half of all the papers published in the Society of Antiquaries journal Archaeologia between 1820 and 1850 (Hingley 2007: figure 59). Of nearly 600 entries on small finds collected for this chapter from the Society of Antiquaries subject index (maintained until 1988), around 40 per cent were nineteenth century or earlier. As Tyers (1996: 2) notes for early studies of Roman pottery, in general there were two principal approaches in antiquarian publications, each of which was to persist in later artefact research: first, the interpretation of objects through reference to ancient texts in which they were mentioned and, secondly, a focus on the documentation and description of objects as a material record of the historical past.

The Beginning of Archaeological Approaches to Artefact Study During the nineteenth century, and particularly its second half, more systematic approaches to artefact studies developed, thanks especially to two major figures: Charles

The Development of Artefact Studies    65 Roach-Smith and Augustus Lane Fox Pitt-Rivers. Both were private collectors who became interested in the site context of objects and their activities are recognizable as archaeology, rather than antiquarianism. A dependence on ancient texts is still evident in this period, but this was influenced by nineteenth-​century scholarly endeavours across a range of parallel subjects, from natural history to ethnology. For instance, in his catalogue of artefacts from Richborough, Roach-​Smith draws upon Pliny’s description of spoon handles being used to pierce eggshells, but goes on to compare this to folk customs in East Anglia, in which eggshells were apparently pierced to prevent witches sailing in them (Roach-​Smith 1850: 206). This illustrates the contemporary interest in ethnographic analogy and the implicit identification of ancient peoples with those who were perceived to be their modern equivalents. Pitt-​Rivers collected both antiquities and ethnographic material and participated in the learned societies of both disciplines. The comparison of ancient and modern artefacts was fundamental to his methodology (Chapman 1985). Pitt-​Rivers was also an early exponent of experimental reconstruction, but his most important contribution to artefact studies, influenced by figures such as Darwin, was the development of typology as a means of understanding the evolution and dating of objects (Bowden 1991: 54–​55). Both Pitt-​Rivers and Roach-​Smith also had a good understanding of issues related to dating: Pitt-​Rivers pointed out the significance of artefacts in dating archaeological sequences and Roach-​Smith knew that only the latest coins in a deposit were useful for dating (Bowden 1991: 120; Roach-​Smith 1868a: 18–​19). Interest in geographical distribution is also seen in this period; for instance, Akerman’s listing of coin finds in south-​east Britain (Akerman 1844; J. Evans 1956: 275). Although classification of natural history was, by this date, well established, it was only with the recovery of greater quantities of artefacts as a result of the building of nineteenth-​century railways and canals that it became possible for more systematic archaeological classification to be made (Piggott 1989:  153). Reports on categories of artefacts as part of a site report start appearing towards the end of the nineteenth century (C. Evans 2007: 287). Lucas (2001: 68) suggests that, deriving from the previous tradition of artefact collection, finds took ‘pride of place’ in early site reports; yet when we examine reports for the Roman period they do not always include the finds (e.g. Ashby et al. 1904). Where finds are listed, it is evident from references in passing to ‘other finds’ that only a small selection is actually being published (Pevensey Excavation Committee 1908: 14; Bushe-​Fox 1913: 31). The prevailing attitude is summed up nicely in Fox’s 1893 report on excavations at Silchester, in which he asserts that ‘the objects in bone and glass call for no special remark’ and omits them (Fox 1893: 471). Bone artefacts start to be systematically included in site reports only from the 1920s (see Table 4.1). In addition to deciding what to include, those publishing catalogues had to choose how to arrange their material. This was not just a presentational matter. There was an evidently dialectical relationship between classification and interpretation (discussed in more detail later in this chapter). Roach-​Smith (1850, 1859) arranged artefacts mostly in categories according to their perceived function, which suited his use of ancient texts to discuss how the artefacts had been used in everyday Roman living. Pitt-​Rivers, the other pioneer of artefact studies in the nineteenth century, used illustrated captions in

66

Table 4.1 Finds categories in Roman site reports up to the 1960s Contents and order of occurrence of finds categories (excluding animal bone, human bone, botanical remains, etc., unless these Main organizing include worked principle of small objects) finds

Sections with authorship credited to others apart from excavator/​s

Reference (in chronological order)

Site

Fox (1893)

Silchester

Gold finger-​ring/​ None intaglios/​coins/​ other metal small finds/​pottery/​coin hoard/​silver hoard

No

Appendices on coin hoards and silver hoard

Pevensey Excavation Committee (1908)

Pevensey

Pottery/​stamped None tiles/​ coins/​bronze small finds/​other small finds, mostly metal

Partly (location of selected finds noted in description of excavations)

No

Curle (1911)

Newstead

Stone objects Functional including categories inscriptions/​ dress and armour/​ weapons/​ Samian/​other pottery/​tools and implements/​ transport and harness/​ miscellaneous/​ ornaments/​ appendix on vegetable remains including baskets/​ appendix on coins

Yes

Appendices on coins and vegetable remains

Haverfield (1911)

Corstopitum

Inscribed objects/​other stone objects/​ bronze small finds (starting with brooches)/​ iron small finds/​ lead small finds/​ miscellaneous objects of wood, ceramic or enamel/​samian/​ decorated grey ware pottery sherd/​coins

Only for the pottery

All, including pottery and coins

Material

Stratigraphic context of finds

Table 4.1 Continued

Reference (in chronological order) Bushe-​Fox (1913)

Site Wroxeter

Contents and order of occurrence of finds categories (excluding animal bone, human bone, botanical remains, etc., unless these Main organizing include worked principle of small objects) finds

Stratigraphic context of finds

Sections with authorship credited to others apart from excavator/​s

Small finds (starting with metal finds, brooches listed first)/​glass/​ samian/​other pottery/​coins

Material

Yes

Pottery and coins

Baddeley (1922) Cirencester

Bronze small finds/​coins/​ pottery

Material

No

No

Winbolt (1925)

Folkestone villa

Coins/​samian/​ other pottery/​ bronze small finds/​bone (including bone artefacts among animal bone)/​ silver/​iron/​glass/​ tile stamps/​ architectural

Material

Yes

No

Bushe-​Fox (1926)

Richborough

Sculptured & Material architectural (though not very stonework/​lead pig systematic) with inscription/​ bronze brooches/​ other metal small finds/​other small finds/​lamps/​glass vessels/​samian/​ other pottery/​coins

Yes

Pottery and coins

Wheeler and Wheeler (1928)

Caerleon Inscriptions on Material amphitheatre stone and lead/​ stamped tiles/​ statuettes/​ brooches/​other metal and bone small finds/​ intaglios/​glass/​ lamps/samian/​ other pottery/​coins

Yes

Inscriptions and stamped tile

(continued)

68

Table 4.1 Continued

Reference (in chronological order)

Site

Contents and order of occurrence of finds categories (excluding animal bone, human bone, botanical remains, etc., unless these Main organizing include worked principle of small objects) finds

Stratigraphic context of finds

Sections with authorship credited to others apart from excavator/​s

Fieldhouse, May, and Wellstood (1931)

Tiddington

Bronze small finds Material (starting with brooches)/​bronze and iron small finds/​iron small finds/​other metal small finds/​glass, stone and bone objects/​pottery/​tile

No

No

Whiting, Hawley, and May (1931)

Ospringe cemetery

Samian/​other pottery/​bronze small finds (starting with brooches)/​lead and pewter objects (coins were published separately in an earlier article)

Material

Yes

Pottery

Bushe-​Fox (1932)

Richborough

Bronze brooches/​ miscellaneous other small finds/​ glass/​lamps/​ pottery

None

Yes

Small finds, samian, potter’s stamps, coin summary, coin hoard

Corder and Kirk (1932)

Langton villa

Bronze small Material finds (starting with brooches)/​ bone small finds/​ iron small finds/​ stone small finds/​ pottery/​coin

No

Coins

Table 4.1 Continued

Reference (in chronological order)

Site

Contents and order of occurrence of finds categories (excluding animal bone, human bone, botanical remains, etc., unless these Main organizing include worked principle of small objects) finds

Stratigraphic context of finds

Sections with authorship credited to others apart from excavator/​s

Wheeler and Wheeler (1932)

Lydney

Architectural and Material worked stone/​ (though not very bronze small systematic) finds (starting with brooches)/​ miscellaneous small finds/​iron objects/​ inscribed objects/​ pottery/​coins

Yes

Coins found before 1928-​ 9, coin hoard 1

Kenyon (1948)

Leicester Jewry wall

Samian/​other Material pottery/​bronze small finds (starting with brooches)/​bone small finds/​jet and shale small finds/​ gold finger-​ring/​ bronze figurine/​ spindle-​whorls/​ miscellaneous small finds including those of iron and stone/​ tiles/​coins

Yes

Samian, potter’s stamps and coins

Bushe-​Fox (1949)

Richborough

Bronze terret/​ Material bronze small finds (starting with brooches, intaglios also in this section)/​ miscellaneous small finds/​glass vessels/​lamps/​ iron/​pottery/​coins

Yes

Small finds, pottery and coins

(continued)

70

Table 4.1 Continued Contents and order of occurrence of finds categories (excluding animal bone, human bone, botanical remains, etc., unless these Main organizing include worked principle of small objects) finds

Reference (in chronological order)

Site

Oswald (1952)

Margidunum

Via Quintana/​ Context area extending to south rampart/​ four outer ditches/​ defensive area of outer ditches

Wedlake (1958)

Camerton

Samian/​coarse Mainly by type of pottery/​figurines/​ object brooches/​coins/​ architectural fragments/​querns/​ whetstones/​ spindlewhorls/​ counters/​beads/​ bracelets/​rings/​ spatula/​styli/​ miscellaneous bronze objects/​ studs and nails/​ toilet accessories/​ glass/​locks and keys/​pins/​ needles/​spoons/​ shale objects/​ miscellaneous stone objects/​ iron knives/​ miscellaneous iron objects/​Belgic brick or daub/​ Roman brick

Gould (1966–​7) Wall

Coins/​glass/​ Mainly material brooches/​samian/​ coarse pottery/​ medieval pottery/​ military bronzes/​ other metal objects/​stone and shale objects

Stratigraphic context of finds

Sections with authorship credited to others apart from excavator/​s

Yes

No

Samian, coins

Yes

Coins, glass, brooches, samian, military bronze

The Development of Artefact Studies    71 Table 4.1 Continued

Reference (in chronological order)

Site

Cunliffe (1968) Richborough (director of excavation B. W. Pearse)

Contents and order of occurrence of finds categories (excluding animal bone, human bone, botanical remains, etc., unless these Main organizing include worked principle of small objects) finds Masonry and Material sculptural fragments/​ brooches/​other bronze small finds/​bone small finds/​iron objects/​ stone mortars/​ Saxon sword/​ iron age pottery/​ Roman coarse pottery/​samian/​ other pottery/​ coins

Stratigraphic context of finds Yes

Sections with authorship credited to others apart from excavator/​s All except Roman coarse ware

© Ellen Swift.

his series Excavations in Cranborne Chase (C. Evans 2007: 286), arranging the objects shown in the plates in material categories (Pitt-​Rivers 1887). Hierarchies of material that were already beginning to appear in the first antiquarian studies were consolidated in the publication of reports that listed inscriptions and figurative art objects first; followed by pottery and then the other material or functional categories (e.g. Roach-​Smith 1859). These hierarchies were still evident over half a century later:  for instance, inscribed objects came first in the 1911 Newstead and Corstopitum reports; inscriptions and stamped tile were listed first in the Caerleon amphitheatre report of 1928; while sculptural/​architectural stonework and inscribed objects were first in the Richborough report of 1926 (see Table 4.1). Lucas (2001: 79) observes that objects that were susceptible to typological analysis also tended to receive special attention, creating a self-​reinforcing cycle in which they were then perceived as worthy of further study. This is certainly true of pottery and brooches in the Roman period, which quickly become prominent in the categories of object recorded, a convention still evident in Roman site reports in the late twentieth century (see Table 4.1). Publication of site reports and catalogues had the inevitable effect of distancing people from the actual artefacts (C. Evans 2007: 288–​289), although it was also very

72

72   Ellen Swift enabling to research (especially in cases where the illustrations were of high quality), reducing the necessity of travelling to visit collections and disseminating knowledge of artefacts to a much wider audience. Artefact expertise remained, however, in the hands of those actually in contact with the material—​invariably private collectors of independent means. Those most active in publishing notes on Roman artefacts during the nineteenth century (drawn from a survey of entries on categories of small finds in the Society of Antiquaries of London subject catalogue) include Albert Way, Revd Charles King, Revd John Cox, Augustus Wollaston Franks, and Robert Soden Smith. All were university-​educated, from relatively privileged backgrounds, and probably developed their interest in Roman antiquities through education in the Classics. Only King held a university position, as a research fellow at Cambridge (Wilson 1997: 3; Burton 2004; Nurse 2004a, b; Wroth 2004). The presence of King and Cox is a good indication of the preponderance of clergymen among nineteenth-​century antiquarians interested in the Roman period. Only two definite instances of women reporting on Roman antiquities in this period can be found: the Dowager Duchess of Cleveland, who seems to be reporting items on behalf of ‘His Grace’ (Cleveland 1850) and Hannah Jackson Gwilt, who was more closely involved with the Royal Astronomical Society (Beech 2006). As artefacts were gradually transferred from private to public collections, museum curators—​who were likely to have previously been private collectors (as with Soden Smith, Franks, and A.  J. Evans)—​became the first ‘professional’ artefact specialists. Franks, for instance, in spite of a wide range of other chronological interests, published a large number of notes on Roman artefacts in antiquarian journals, which are indexed in the Society of Antiquaries subject catalogue. These are typical of the scholarship of the day: most merely note the find-​spot and record the presentation of the artefact(s) at a society meeting. Where Franks contributes a more substantial article, the artefacts are evaluated through reference to Classical texts and depiction on Roman monuments, and parallels from other sites are noted (see, e.g. Franks 1858, 1867–​70). Typological studies are also evident: for example, a consideration of brooch development by Arthur John Evans, Keeper of the Ashmolean Museum (A. J. Evans 1896). It would be a mistake, however, to characterize later-​nineteenth-century ​studies as obsessed with typology. Pitt-​Rivers, for example, proposed new fields of enquiry, such as the study of pottery fabric and distribution and site assemblages as a means to evaluate the status of the occupants (Tyers 1996: 7–​8). These were not taken up as serious endeavours, however, until well into the twentieth century. Roach-​Smith meanwhile (who, unusually, was not a man of private means but a chemist) made a ground-​breaking study of samian pottery (Tyers 1996: 4–​6). This study—​using evidence from kiln sites, samian stamps, and the form and decoration of the vessels—​established that samian was an import into Britain from Roman Gaul (Roach-​Smith 1868b). It is interesting to contrast the background of later-​nineteenth-​century archaeology in Britain—​dependent on knowledgeable amateurs—​to that in Germany, where archaeology enjoyed more institutional support (Freeman 2007: 131–​132). Dragendorff, for instance, who published a very important paper on samian classification towards the

The Development of Artefact Studies    73 end of the nineteenth century (Tyers 1996: 8–​9), was an academic who had completed a Ph.D. on the subject and been taught by professors of archaeology in Germany.

Approaches to the Classification and Interpretation of Artefacts in  the Earlier Twentieth Century In the early twentieth century Romano-​British archaeology was dominated by Francis Haverfield, the first university-​based specialist on Roman Britain (Freeman (2007) is a detailed treatment of his life and work). Although he is more well known for his synthetic work and work on inscriptions, he had a good knowledge of categories of artefacts such as brooches and samian pottery. He tried to encourage the publication of museum catalogues and corpuses of material (Freeman 2007: 233–​235). With regard to artefacts, he set an example through his publication detailing the Romano-​British holdings of provincial museums in Britain (Haverfield 1891), and towards the end of his career he published a catalogue of the brooches in the Tullie House Museum (Haverfield 1919). Museum catalogues incorporating Romano-​British finds within wider artefact categories also started to be produced by the British Museum (e.g. Walters 1908; Marshall 1911; Dalton 1912); although Dalton himself noted that the authors were yet to establish expertise in their subjects (Miller 1973: 316; Freeman 2007: 90), and some catalogues were later criticized for their inaccuracies (see, e.g. Jenkins 1985). Independent, often wealthy, figures such as the lawyers James and Alexander Curle (Graham Ritche 2002: 19–​20) and Bushe-​Fox also contributed significantly to the publication of excavated finds. As Lucas (2001: 65–​66) notes, in the nineteenth and early twentieth centuries finds assemblages were usually written up by the excavators themselves, while as the twentieth century progressed an increasing proportion was allocated to specialists. We do commonly find separately authored sections on coins and/​or pottery from the very early twentieth century, however: e.g. Newstead and Richborough (Curle 1911; Bushe-​Fox 1926) (see also Table 4.1, a larger sample of site reports than considered by Lucas, whose focus is not specifically on the Roman period). Site finds were sometimes published separately to the excavation report (see, e.g. J. Evans 1894; Haverfield 1911). In addition to contributions discussing particular types of artefacts, Haverfield incorporated discussion of brooches and pottery into his influential synthetic account of the ‘Romanization’ of Britain (Haverfield 1912). He focused particularly on the ‘Celtic’ influence in these categories of artefacts, evident in their artistic style (Haverfield 1912: 39–​ 40); this had also been remarked on previously by others (e.g. A. J. Evans 1896; Romilly Allen 1901; Curle 1911: 317) and is something of a recurring theme as discussed later in this chapter. Artefacts were also included in a popular and influential synthesis The Roman Era in Britain (1911) by John Ward, Keeper of the National Museum of Wales,

74

74   Ellen Swift who included a much wider range of objects than Haverfield. Ward’s arrangement of the chapters on artefacts is an inconsistent mixture of material and functional categories. Other than the pottery section, which discusses kilns and methods of manufacture as well as descriptions of form, the information is limited to a description of the artefacts and their varied forms, rather than discussing how the finds inform our wider picture of Roman Britain. This generation of scholars was from a very similar background to those of the nineteenth century: mostly educated at public schools and Oxford or Cambridge but without the opportunity to study at university, let alone doctoral level in archaeology, and gaining knowledge through a mixture of practical experience and informal training from others in the field. Bushe-​Fox, for instance, gained his experience through excavation. On the basis of his knowledge, he was later appointed Inspector of Ancient Monuments (Wheeler 1955: 154–​155). Although such accounts fade out as the twentieth century progresses, we encounter some interesting imaginative reconstructions in early twentieth-​century site reports, based, in part, on the information from the finds assemblage. These show the continued existence of a nineteenth-​century strand of interpretation from the perspective of ancient history. Curle (1911), for instance, tends to begin each finds section of his report from Newstead with a preamble that uses the artefacts to inform a suggested picture of daily life at the fort. On the evidence of the finds, and against the received wisdom that was to prevail for many succeeding decades, he suggests the presence of women at the fort—​an idea that was not revisited until the late twentieth century (Curle 1911: 317; Allason-​Jones 1995). An integrated approach to finds is also seen in Winbolt’s discussion (1925: 167–​180) of everyday living at the Folkestone Roman villa—​an account very reliant on sources like Pliny but also drawing on the small finds from the excavations. The incorporation of finds is likely to stem, partly, from the functional arrangement of material facilitating the reconstruction of site-​level activities and correlation with ancient texts and, partly, from the fact that the authors also wrote the finds sections. Despite these early examples of functional categorization, it became much more common in the twentieth century to arrange artefact catalogues instead by material (see Table 4.1). The origins of this, according to Lucas (2001: 69), lie in nineteenth-​century and even earlier antiquarian approaches to classification of all kinds of objects. Division into material categories fostered the development of separate branches of artefact specialism, which became increasingly inward-​looking (focused, for example, on typological studies), with little reference across material categories or consideration of more holistic questions; for instance, artefacts as site assemblages (Webster 1977: 318; Lucas 2001: 79). Material categories did not map easily onto possible function and thus activity at a site. Classification by material inevitably, then, constrained interpretation. It is also important to note, though, that the traditional site-​report format, while discouraging contextual approaches, did not actively prevent them. Many relatively early finds reports list the material by context, in addition to the finds sections and/​or provide basic contextual details after each find entry. Oswald’s report (1952) on Margidunum presents the finds only by context, though this is unusual (see Table 4.1). The information was

The Development of Artefact Studies    75 there for contextual studies, but, despite this, they were not a focus of research: context was valued only for dating purposes. It may be useful briefly to consider how the term ‘small finds’ became established. As noted by Evans (C. Evans 2007: 287–​288), the ‘site-​report formula’ came into existence only from the early twentieth century. Examining these early reports, particularly the series of research reports of the Society of Antiquaries, we can trace the development of the term. Initially, the word finds refers not only to the artefacts within a site report but also to structural evidence and other material: that is, to anything that has been found. We see this, for instance, in the title of some early publications (e.g. Donovan 1935). At around the same time, it is evident that the term is apparently becoming a colloquial way of referring to the artefacts from an excavation. Some site reports from the 1920s and 1930s use it to refer to the collection of artefacts found, but place it in quotation marks, as if it is not quite formal written English or established as a term (Winbolt 1925, 1935; Wheeler and Wheeler 1932). Others use it as a section or volume title for specialist reports (Corder and Kirk 1932; Nash-​Williams 1932). However, there is a good deal of variety in practice, and many excavation reports use other headings: for instance, ‘Objects of Metal and Bone’ (Wheeler and Wheeler 1928: 161); ‘Objects Found during the Excavations’ (Baddeley 1922); ‘Various Metal and Bone Objects’ (Whiting et al. 1931: 99). ‘Small’ seems to be used first as an adjective. Haverfield (1911), for instance, subtitles his report on artefacts from Corstopitum ‘Smaller Objects’ and Bushe-​Fox’s series (1913, 1926, 1932, 1949) of Society of Antiquaries research reports use the heading ‘Small Objects in Metal, Bone, etc.’ or ‘Small Objects in Metal, Bone, Glass, etc.’. We first see ‘small finds’ as part of a longer phrase ‘Brooches and Other Small Finds’ in the excavation report on Lydney by Mortimer and Tessa Wheeler (Wheeler and Wheeler 1932: 68). From the 1940s, ‘small finds’ is used fairly consistently by Kathleen Kenyon in her site reports (Kenyon 1948, 1950, 1959) and now seems to have become an accepted term for the artefacts section of a report other than pottery. Kenyon (1952: 132, 149) defines the category in her methodological guide to excavation as objects that are not bulk finds ‘such as coins, brooches, ornaments’ or as referring to ‘special objects’. Its use in a guide such as this presumably helped to cement the term in common usage, although we can note that its usage to include coins as well does not match with the current established terminology, which tends to use it for small artefacts other than coins. There is still plenty of variation in practice for most of the twentieth century: for instance, the Society of Antiquaries research reports using a main heading ‘The Finds’ for the section of the report dealing with artefacts and subheadings ‘Objects of . . . ’ using material categories (e.g. Hawkes and Hull 1947; Wacher 1969). Following the use of the title ‘Small Finds’ for the Archaeology of York series volumes on this material (e.g. Macgregor 1976, 1978) and for Crummy’s seminal 1983 report (Crummy 1983), it then becomes established in use and is especially seen in many excavation report finds volumes from the 1990s (e.g. Manning et al. 1995; Bishop 1996; Cool and Philo 1998; Rees et al. 2008). In coin studies, a shift to a different method of classification relatively early on was particularly enabling to research. The most widely used coin reference book in the initial years of excavation reports was Cohen’s complete alphabetically arranged list of the western

76

76   Ellen Swift Roman coinage, achieved in the mid-​nineteenth century (Babelon 1901: 219). In the 1920s, the authors of the Roman Imperial Coinage (soon to become the standard reference work) introduced a new system: listing coins by emperor and date or mint (Mattingly et al. 1923–​ 84). This facilitated research, since this method of classification corresponded to a much greater extent with the factors involved in the original production of the coinage.

Scholars and Their Studies from  the 1930s to 1970s The continuation of previous approaches driven by ancient history, in which daily life is the subject of study, can be seen in synthesis books produced through most of the twentieth century that do incorporate finds in their discussion of everyday activities (Quennell and Quennell 1937; Birley 1964; Liversidge 1968). Liversidge, in particular, draws heavily on Romano-​British artefacts in addition to ancient texts, material from Pompeii, and so on, including discussion of artefacts found in context such as shop and workshop contents, grave assemblages, and votive deposits. However, in contrast to the work of her contemporary Jocelyn Toynbee (e.g. Toynbee 1962), this work was later undervalued and not incorporated to any significant degree in later synthesis volumes (Collingwood and Richmond 1969; Todd 1989). Liversidge was also omitted in a later survey of the major figures of Romano-​British archaeology (Jones 1987; Wallace 2002). The wider failure to value her work may be related to the lesser status of Romano-​ British archaeologists among Cambridge academics of the period (Richard Reece and Catherine Hills, pers. comm.) and/​or the focus of her research on social history. If we turn to scholarly articles on artefacts from the 1930s up to the end of the 1960s, the main focus is still on the compilation and descriptive publication of material, sometimes also discussing distribution and dating (e.g. Hawkes 1940; Hildyard 1950; Charlesworth 1959; Hawkes and Dunning 1961; Boon 1966). The article by Hawkes and Dunning, though it also includes a typology, is rare in its consideration of the archaeological context of the material. Continuing from the nineteenth-​century practice of publishing notes on small finds displayed at learned societies, most published papers are short descriptive ‘notes’ on individual or small groups of finds (data compiled from Society of Antiquaries subject catalogue). In pottery research, there is a focus on pottery kilns and distribution from the 1930s onwards (Tyers 1996: 16), applying the earlier work done on samian to other pottery types. Articles on small finds also sometimes show an interest in trade routes and/​or the origin of artefacts or of characteristic forms or motifs (e.g. Dalton 1922; Tonnochy and Hawkes 1931; Curle 1933; Hildyard 1945). An interest in culture-​historical style is also seen, evident through continued discussion of the ‘Celtic’ nature of Romano-​British brooches (e.g. Collingwood 1930a; Savory 1956). O’Neil’s paper (1935) on coins is an early attempt to evaluate and compare the coin site-​find profile of four different sites (not taken further until Reece’s work of the 1970s onwards),

The Development of Artefact Studies    77 although he himself points out the difficulties: namely, the old-​fashioned publication convention of publishing only selective data and the scarcity of people available to carry out the research needed (O’Neil 1935: 76–​79). We will return to these topics later, but this comment also raises the question of who is engaged in finds research during this period. If we examine the profile of those contributing articles on artefacts to scholarly journals, it is evident that it is dominated by museum personnel, who were also instrumental in passing on knowledge and providing opportunities for new generations of artefacts scholars. Dalton and Tonnochy held positions at the British Museum, as did Hawkes until he transferred to a chair at Oxford in 1946 (Champion 2004); Stevenson was at the National Museum of Antiquities Scotland; Liversidge was an Honorary Keeper of the University Museum of Archaeology and Anthropology at Cambridge before she was appointed as a Research Fellow of Newnham College and later a Cambridge faculty lecturer (Archives Hub 2012; Catherine Hills, pers. comm.); Hull was Curator at Colchester Museum (Hawkes 1982); Charlesworth worked in a number of museums and at the Ancient Monuments Inspectorate (Serpell 1982); Boon followed Nash-​Williams at the National Museum of Wales (Manning 1995); Harden employed Kirk at the Ashmolean (Young 2007) and later moved to the London Museum (Painter and Thompson 1994). Collingwood is rare as someone with expertise on artefacts who had always been based in a university. We can also note the appearance of women; to comply with changes in the law, women were allowed to become Fellows of the Society of Antiquaries from 1920 (J. Evans 1956: 388–​9). Although there is a noticeable dip in the 1930s and 1940s, probably linked in some way to the disruptions of the Second World War, women increase noticeably as a proportion of those working on small finds from the 1950s onwards (see Figure 4.1(a)). As before, most of these scholars came from privileged backgrounds and were privately educated, mostly at public schools, before studying at Oxford or Cambridge; although few had the opportunity to study for a doctorate. Boon is typical in that he had a degree in Latin but no postgraduate qualifications, building up his expertise through museum posts (Manning 1995:  p.  xiii). Webster continued the tradition of the self-​ taught expert. Initially an engineer, he later gained a post at the Grosvenor Museum, Chester, based on his experience gained through archaeological excavation (Henig and Soffe 2002). It is instructive to compare the two editions of the core textbook The Archaeology of Roman Britain (Collingwood 1930b; Collingwood and Richmond 1969) to gauge archaeological approaches to artefact studies across the same period. Both editions include chapters on coins, samian, coarse pottery, brooches and a chapter on weapons, tools, and utensils. In neither is there any attempt to integrate the artefacts into a wider picture, and their value is mainly seen in terms of dating. However, the potential of pottery and coins as economic evidence is acknowledged. Collingwood’s chapter (1930b: 185–​199) on coins begins with a discussion of the archaeological study of coinage, as opposed to numismatic history or art-​historical approaches. It is evident in the revised chapter and its bibliography that, while hoarding has been studied further, more work has been done in the intervening forty-​odd years on the structure of the coinage and the monetary system:

78

(a)

Percentage

100 90 80 70 60 50 40 30 20 10 0 –8 8 80

–7 9

19

70

–6 9

19

60

–5 9

19

50 19

–4 9 40

19

–3 9 30

19

–2 9 20

–1 9

19

10 19

19

00

–0 9

Women Men

(b) 100 90 80 Percentage

70 60

Women Men

50 40 30 20 10 0

(c)

1970s

1980s

Coins, pottery or small finds Animal bone, botanical remains or scientific analysis Mix of finds and other Other

Figure 4.1  Gender profiles in artefact studies. (a) Percentage of men and women with published articles on small finds listed in the Society of Antiquaries subject catalogue from 1900 to 1988 (end date of catalogue). (b) Percentage of men and women publishing articles in Britannia 1970–​89 (excluding reviews). (c) Subject of articles by women in Britannia 1970–​89 (excluding reviews). Source: © Ellen Swift.

The Development of Artefact Studies    79 that is, not those areas Collingwood prioritized as ‘archaeological’. There is still no mention of studies of coins as site finds, despite O’Neil’s work. We can see here the influence of long traditions of numismatic study overriding a nascent awareness of the potential of other approaches. In the chapter on samian in the later edition, continuity of traditions of study is also evident. Samian was first valued for dating; it is noted that decorated sherds can now be dated more precisely and stratified decorated samian is seen as the publication priority (Collingwood and Richmond 1969: 236, 249–​250). The revised chapter on coarse pottery includes a new section on production centres and a greater range of fabric-​type descriptions, reflecting the work done on these areas by scholars such as Gillam and Hartley in the intervening period. The other finds sections show very few changes, which is surprising, since Richmond himself encouraged his students to take up artefact research (Richard Reece, pers. comm.). In 1969 (as in 1930), the importance of brooches lies in typology, dating, and distribution, but, whereas Collingwood, in 1930, is also interested in ‘artistic style’ and its importance for understanding cultural history, Richmond dismisses it as not the proper focus of archaeologists (Collingwood 1930b: 243; Collingwood and Richmond 1969: 286). The final chapter, on weapons, tools, and utensils, is descriptive in both cases—​a listing of the various types of items found in the Roman period—​but the later edition incorporates more information on manufacture and use of the artefacts (Collingwood 1930b: 261–​274; Collingwood and Richmond 1969: 304–​325). Frere’s synthesis volume of the same period incorporates finds research to a much greater extent, integrated within the narrative rather than occurring in separate chapters. Artefact research contributes, for instance, to questions of production and trade: for example, the market competition of the various pottery industries and consideration of imported and exported objects (Frere 1967: 29–​294). Social history of the kind done by Liversidge was not easily incorporated within the framework of political and economic history that structured these books.

The Development of Artefact Studies from the 1970s Onwards Alongside the continuing production of essential corpora of small finds (e.g. Henig 1974; Manning 1976; Green 1978), the 1970s saw a number of developments in studies of both pottery and coins broadly in line with wider developments in archaeological interpretation that focused on evidence as scientific data (notably D. Clarke 1968). The Study Group for Roman Pottery was founded in 1971, and new techniques in pottery, such as thin-​section analysis, combined with massive increases in available data (Tyers 1996: 2–​22), contributed to established research areas (e.g. origin and distribution). This combination of valued research questions, new data, and developments in methodology—​including an emphasis on scientific methods—​ensured that pottery studies flourished, focusing on the economy and trade. Webster (1977) saw the priorities

80

80   Ellen Swift as the use of more objective descriptive terms (replacing the undesirable subjectivity of extant fabric descriptions, for instance), and more work on manufacture and distribution. The increase in data from the 1970s onwards also necessitated a consideration of quantification methods (Tyers 1996: 22–2​3). In coin studies, new methodologies for the study of site finds were developed by Reece and Casey, including the allocation of coins to specific period divisions, which enabled statistical comparison (e.g. Reece 1973; Casey 1974), and culminating in the edited volume Coins and the Archaeologist (Casey and Reece 1974), setting out specifically archaeological—​as opposed to numismatic—​ approaches to coin finds. The edited book Roman Crafts (Strong and Brown 1976) illuminates the contribution of scientific and technical studies to artefacts research in this period. Reference to contemporary craft and manufacture enabled a detailed consideration of production techniques. The title reflects the wider popularity of handcrafts, folk art, and so on in the 1970s. Of the contributors on artefacts in Roman Crafts and elsewhere, Manning, Price, and Wild have a university affiliation, showing a gradual spread of artefact researchers into academia—​a trend also demonstrated by coins and pottery specialists such as Reece, Casey and Peacock. As Jones (1987: 90) notes, most of the 1970s generation of academics in Roman archaeology were unable to study archaeology as undergraduates, but, in career trajectory, they benefited from the expansion of universities in the 1960s and 1970s. Yet they are still outnumbered by contributors based in museums or otherwise outside the universities. Contributors to the edited volume Coins and the Archaeologist (Casey and Reece 1974) also include only three university researchers, among many museum-​based researchers. In the 1980s, approaches to pottery include comparisons to ethnographic data in addition to a strong interest in the details of production methods and technologies (Peacock 1982). More systematization was achieved through new guidelines for recording and the increasing use of computer programs (Tyers 1996: 23). There was also renewed interest in functional pottery studies in this decade (Tyers 1996: 43). We can see pottery studies diverging further from Roman archaeology as a particular subdiscipline in the foundation of the Journal of Roman Pottery Studies in 1986. In small-​finds studies, Crummy’s Colchester report (1983) was the first major finds catalogue since Curle’s Newstead report (1911) to arrange artefacts in functional rather than material categories—​potentially encouraging the study of artefacts in relation to each other as functional assemblages rather than in the abstract as typological sequences. An interest in artefact function is evident in some journal articles of the 1980s (e.g. Jackson 1985); however, change was very slow, with a continued emphasis on traditional approaches in small-​finds studies (typology, distribution) and the publication of previously unpublished artefacts (Swift 2007: 22; data from the Society of Antiquaries subject catalogue). Notwithstanding new cataloguing trends, the dominance of material categories in already published site reports (see Table 4.1) also continued to affect the way that material was interpreted more widely. For example, in Philpott’s analysis of burial rites, burials with vessels made from different materials are considered in separate chapters, even though Philpott himself points out that the material from

The Development of Artefact Studies    81 which grave goods are made might not be ritually significant (Philpott 1991: 127, 117). The separation inevitably fosters a discussion of status as expressed through more or less prestigious materials, whereas a consideration of graves with similar vessel forms, irrespective of material, might have resulted in a greater emphasis on ritual practice. In journal articles, books, and theses there was a continuing tendency to study only one type of artefact or material category (e.g. Lloyd-​Morgan 1981; Allason-​Jones 1989a), rather than integrated studies of assemblages of diverse finds, although, as we have seen in earlier periods, a separate strand of research concerned with everyday life does sustain a consideration of artefacts as functional groups (e.g. Allason-​Jones 1989b). Museum expertise continues in the 1980s in both the documentation and interpretation of finds (e.g. Johns et al. 1980; Boon 1983), and independent scholars also continue to make significant contributions (e.g. Hattatt 1982, 1985). We also see publications based on theses on Roman finds completed in the later 1970s and 1980s (e.g. Lloyd-​Morgan 1977; Green 1978; White 1988; Allason-​Jones 1989a). Indeed, artefact study comprises around half of all doctoral research in Roman archaeology in the 1980s (Swift 2007: 21–2​ 2); this increase in research perhaps contributed to the founding of the Roman Finds Group in 1988. Yet there are still comparatively few artefact specialists in university posts, and these are mostly coin and pottery researchers (e.g. Casey, Reece, Fulford, Greene, Hartley, Peacock, Gillam) who were appointed in the 1970s and earlier. It is important to note the wider dearth of new posts in archaeology at this time (Mattingly 2008). Although there was a good deal of finds research being carried out by the 1980s, the wider context was different; unlike pottery researchers in the period of university expansion in the 1970s, researchers on finds in the 1980s were unable to gain university posts, thanks to contraction and retrenchment during this time. To consider gender profiles:  by the 1980s, women are apparently overrepresented in finds studies, making up around a quarter of all finds researchers (Figure 4.1(a)). Although this is only slightly greater than the proportion of women publishing in the journal Britannia on any subject in the 1970s and 1980s (Figure 4.1(b)), it is evident that these published Britannia contributions by women are themselves heavily biased towards artefact and other specialist reports (Figure 4.1(c)). By contrast, in the same period only about a quarter of the articles by men in Britannia are on finds topics. The connection between women and artefact studies was noted at the time (Jones 1987: 95) and is further discussed later in this chapter. A synthesis volume summarizing the major research in Romano-​British archaeology between 1960 and 1989 (Todd 1989) provides an opportunity to examine the integration of finds research into the wider picture of Roman Britain in this period. Although there are no chapters specifically concerned with artefacts, or with daily life, a number of contributors incorporate research on pottery, coins, and/​or precious metal objects into overviews of various topics, including art, the economy, and religion in both the early and late Roman periods (Blagg 1989: 209–​210; Esmonde Cleary 1989; Fulford 1989; Henig 1989). We can see here the continued existence of hierarchies in material. Research on other objects is mentioned by only a couple of authors (Breeze 1989: 38; Keppie 1989: 68–​69). In the bibliography, publications on artefacts are concentrated in

82

82   Ellen Swift a section on industries and crafts (Todd 1989: 261–26​2), probably because this had been the focus of new research in the 1960s and 1970s (most notably Strong and Brown 1976). The 1990s and early 2000s saw a number of developments both in methodological issues and in interpretation. Roman finds work continued to be characterized by a strongly empirical approach (Hingley and Willis 2007: 7–​8). Systematization of data recording continued, with the establishment, for instance, of the National Roman Pottery Fabric Reference Collection (Tomber and Dore 1998) and a research framework for Roman pottery (Willis 2004: 1). The application of comparative statistical methods to artefacts datasets is seen, and in this we also see computer analysis moving beyond coins and pottery to other classes of artefact (e.g. Cool and Baxter 2002). From the late 1990s onwards, site reports more widely adopt the functional artefact categories used by Crummy (1983). More widely in Roman archaeology, the 1990s is the decade that saw the increasing incorporation of theoretical approaches that had already become established in other branches of archaeology (James 2003: 180). Explicit theoretical approaches are, unsurprisingly, most evident in finds-​based articles in the proceedings volumes of the Theoretical Roman Archaeology Conference (TRAC) from 1991 onwards. Earlier artefact-​based contributions to the TRAC series focus on questioning established assumptions (e.g. Allason-​Jones 1995; S. Clarke 1997), while in later volumes there is more of a tendency to bring in theoretical frameworks more explicitly, particularly in relation to social, cultural, and ethnic identity. This is clearly the dominant focus of interest, incorporating, for instance, postcolonial or consumption theory (see Fincham et al. 2000; Fenwick et al. 2008). A particularly influential book is Hodder’s Symbols in Action (1982). Most of the articles consider one artefact type or category such as pottery vessels, despite the call for a more integrated approach to finds assemblages in the first TRAC proceedings (J. Evans 1995). Some old foci of interest are revisited, notably the ‘Celtic’ nature of some Romano-​British brooches, reframed as a possible example of ‘resistance’ to Roman culture (Jundi and Hill 1998). In addition, many articles focus on contextual information to reinterpret what might be termed ‘pragmatic’ or functionalist explanations of evidence in more post-​processual terms, particularly focusing on ritual behaviour (e.g. S. Clarke 1997; Hutcheson 1997; Dungworth 1998; van Driel Murray 1999) and clearly heavily influenced by Hill’s work (1995) on structured deposition and/​ or Merrifield’s Archaeology of Ritual and Magic (1987). Ironically, given the developing focus here and elsewhere on contextual site studies of Roman material culture (e.g. Allison 2004; Hingley and Willis 2007), a major resource for new small finds and coins data from the later 1990s has been the Portable Antiquities Scheme (PAS), founded in 1997, which records stray finds—​mostly metal-​detected material—​and publishes a web database that has seen rapid expansion (Clark 2008; PAS 2012). Distribution maps, which used to be the product of laborious data collection, can now be mapped instantly and thus are likely to continue to dominate PAS research, as they have wider artefact research for much of the twentieth century, though within new interpretative frameworks such as landscape studies. PAS data will be particularly important in transforming our picture of rural sites (Hingley and Willis 2007: 11–​12;

The Development of Artefact Studies    83 Clark 2008:  19; PAS 2012) and listings on the website of finds research undertaken using PAS data confirm that it is flourishing as a research area. However, PAS data are only slowly being incorporated into wider artefact research (e.g. Eckardt and Crummy 2008: app. 6; Swift 2010: app. 4), rather than being the subject of stand-​alone projects. In an article in 2007, I surveyed, through a questionnaire, prevailing views of Roman-​ finds research and attempted to evaluate the profile of artefact researchers. About a quarter of researchers had a university affiliation at either staff or student level; slightly more were engaged as freelance or unit-​based specialists, and 30 per cent were based in museums. Strong gender bias was evident, with women over-​represented in finds research compared to their numbers in archaeology overall and with a sharp divide between men specializing in coins or pottery and women focusing on small finds or small finds and pottery (interestingly, the gender bias in coins study is not evident on the continent: Richard Reece, pers. comm.). A widespread negative perception of finds research was documented, and it was suggested that this might be related to: first, the persistence of old-​fashioned views of what finds research was or could do; secondly, the position of finds research on the margins of academic study (including its position as ‘women’s work’); and, thirdly, wider trends in theoretical approaches that attempted to undermine empiricism. More broadly, it was also suggested that the negative perception of finds research was connected with hierarchies of value within archaeological research, in which fieldwork and archaeological theory have been dominant, and in historical research, which has privileged political and economic history over social history. The prevalence of women in small-​finds research in particular, documented in this chapter to exist in the 1970s and 1980s as well as more recently, is clearly connected to this. Women, historically, have tended to focus on social history and may also have taken a different attitude to established research hierarchies or been constrained by those hierarchies and wider prejudices into working in what have been perceived to be ‘less important’ areas (Swift 2007). My consideration of a longer time span in the current chapter also provides additional evidence that social history has been a focus of research by women scholars that has generally been sidelined in Romano-​British archaeology. In addition, a negative view of small-​finds research becomes unsurprising when one considers that archaeologists may have been basing their opinions, for instance, on the extremely limited picture, barely updated from the 1930s, given in one of the major textbooks of the later twentieth century (Collingwood and Richmond 1969). Both in my article and elsewhere multiple causes are suggested for well-​documented declining numbers of artefact specialists and artefact research (see, e.g. Allason-​Jones 2001; Willis 2004); the general undervaluing of finds studies, which is a contributory factor (among other things such as retrenchment in the 1980s) to reduced numbers of researchers in universities; researchers in museums not being accorded the same status as university academics; structural problems, such as the lack of a defined career path for artefact researchers outside the universities; work commitments in unit and museum archaeology necessitating a focus on essential catalogues of material rather than synthetic and/​or interpretative articles; and an increasing drift in museum expertise and priorities away from artefact knowledge and research towards education and

84

84   Ellen Swift outreach (Cooper 2007; Johns 2007; Swift 2007). Since there have been fewer researchers altogether, often only one person has studied a particular artefact type in depth (Allason-​Jones 2001: 24). Consequently, methodologies and interpretations have not been debated, and the development of research—​or lack of it in a particular area—​has been very dependent on a single individual. There have, however, been some attempts to address essential skills shortages in artefact identification; for instance, through an Arts and Humanities Research Council grant scheme for doctoral funding specifically limited to finds research (Fulford 2007) and through training as part of the PAS. The latter has contributed enormously in changing the picture in very recent years, not only providing training but reconnecting museums and artefact research through museum-​based posts, encouraging the reporting of new material, and providing new datasets for M.A. and Ph.D. research projects (Clark 2008).

Discussion A number of recurring, interlinked issues that have affected artefact studies can be drawn out for further discussion, evident from the earliest days of antiquarian studies to the early twenty-first century: the availability and presentation of data; the status and profile of researchers; traditions of scholarship, and the influence of wider interpretative trends in academia. Artefact research is unlike text-​based scholarship in that it depends on material objects. Given that few people have been able to travel extensively to view collections, publication and description of the objects have always been priorities: from the coins dictionaries of the eighteenth and nineteenth centuries, through the publication of catalogues of major museum collections of artefacts in the twentieth, to today’s web databases. Researchers have adopted new technologies accordingly: from cast copies of coins to 3-​D scanning of artefacts. Yet there have always been problems with selectivity of publication and in the categorization and quantification of data, which have particularly affected inter-​site comparison of material. Hierarchies in the way that data are published and arranged have been very persistent, affecting what has been studied and how, with some finds categories suffering persistent neglect. Effectively, one generation has unwittingly decided what could be of interest to the future scholars of another. In the earliest site reports, functional grouping of finds assemblages related initially to a wider interest in the interpretation of objects through reference to ancient texts, yet enabled the integration of the artefacts into the interpretation of a site. Later, the almost universal decision to shift to material categories—​together with an increasing quantity of data—​encouraged fragmentation into narrow research specialisms and publications on one category of artefact or type of material, discouraged the integration of finds data into site interpretations, and fostered particular interpretations based on material differences. Conventions established in the 1970s have also not had solely positive effects. Specialist new methodologies have meant that artefact researchers have effectively become further separated from generalist interpreters

The Development of Artefact Studies    85 and excavators (Johns 2007: 30). As Tyers (1996: 22) notes, the use of data from quantified pottery reports has been hampered by the frequent decision not to quantify samian and mortaria and by a plethora of different methods of quantification. In coins studies, the use of different period divisions by Casey and Reece (followed by those they taught) potentially affects site comparisons; though, since full coins list have usually been published, in practice this is not so much of a problem. Lockyear (2007: 212–2​14) discusses other issues with regard to the cataloguing of coins: for instance, condensation of data and the provision of summary information only. To reiterate: artefact research is unlike text-​based scholarship in that it depends on material objects. This is, undoubtedly, one reason why artefact specialists have more often been private collectors, museum curators, and independent finds experts working in unit archaeology rather than university-​based academics. Museum scholarship has had different priorities from university research, with, for instance, an emphasis on connoisseurship (Johns 2007: 29–​31) and the publishing of catalogues of essential data. Arguably, this is another factor that contributes to the differing trajectory of artefact research (especially small-finds research) compared to other scholarly work on Roman Britain, in which there has been more of a drive towards synthesis and broad interpretation. Notwithstanding, there are a number of core topics to which finds evidence can make a strong contribution, and interpretation has consisted of different iterations of these topics: studies of style; the study of production and trade; the reconstruction of daily life; studies of ancient technology; and consideration of artefact function. We can see a slightly differing emphasis on one topic or another through time, related to wider trends in archaeological interpretation such as culture history, processualism, and post-​processualism, which have also affected the way the topic has been approached; but each topic in itself has been present from the nineteenth to the twenty-​first centuries. The documentation and description of material, with further interpretation limited to typology and distribution, have also been persistent traditions. This is not to denigrate the value of this work; such studies are a fundamental stage in any finds research. Documentation through images and description is obviously essential, and, as regards typology and distribution, we cannot hope to understand the objects in their social, cultural, and economic context unless this is defined as closely as possible in time and space. However, we can see that the focus on description and typology as an end in itself retarded the take-​up of new approaches. In all areas of finds research, there was innovation in both methodology and interpretation that was overlooked for decades because it could not be accommodated within prevailing habits of scholarship. It is also clear that, while work done on aspects such as production, technology, and trade was acknowledged more widely in archaeological scholarship, artefact research on everyday life, in which the archaeological context of material was considered, has been less valued. This has particularly affected small-​finds research. If we take the long view of artefact research across the twentieth century and notwithstanding the many problems that remain, it actually seems quite encouraging that so much has been achieved in the face of so many obstacles. The first decade of the twenty-first century has seen a developing, more positive approach to material culture studies—​at

86

86   Ellen Swift least in the domain of academic archaeology (see, e.g. Burnham et al. 2001: 74–​75; Hingley and Willis 2007). New edited compilations focus on artefact studies in particular and not just on pottery or coins (Aldhouse-​Green and Webster 2002; Hingley and Willis 2007; Allason-​Jones 2011). Multiple categories of finds data are increasingly featuring in published artefact research on Roman Britain (including the study of finds assemblages), and artefacts are more often integrated with other types of evidence (e.g. Cool 2006; Gardner 2007; Pitts 2010). There is more effort to provide in-​depth interpretation of a finds assemblage in relation to the site overall, particularly with regard to contextual analysis of notable deposits (e.g. Cool 1998, 2004; Eckardt 2006; Rees et al. 2008), although the failure to do this more widely or to point out this shortcoming in reviews of excavation reports is still seen as a significant problem (Cool 2006: 244–​245). The reconciliation of empirical methodologies with post-​processual interpretative approaches has been essential to the integration of theory within finds studies that is needed for innovation in artefact research (Hingley and Willis 2007: 8; Swift 2007). The first decades of the twenty-​first century may be the beginning of a new phase in small-​finds research—​similar to the step change in coins and pottery studies in the 1970s—​where a new value for material culture studies and social history coincides with a massive expansion in available and high-​quality data, new analytical methods for its interrogation, and theoretically informed approaches to interpretation—​that is, provided the problems that we inherit can be overcome.

Acknowledgements Many thanks to Peter Guest for providing me with an overview of coins literature in relation to Roman Britain from the 1970s onwards, Anne Alwis and Richard Reece for useful comments on the text, and Catherine Hills for providing additional detail on Joan Liversidge.

References Akerman, J. (1844). Coins of the Romans Relating to Britain, Described and Illustrated. 2nd edn. London: J. R. Smith. Aldhouse-​Green, M., and Webster, P. (2002). Artefacts and Archaeology: Aspects of the Celtic and Roman Worlds. Cardiff: University of Wales Press. Allason-​ Jones, L. (1989a). Earrings in Roman Britain. BAR British Series 201. Oxford: Archaeopress. Allason-​Jones, L. (1989b). Women in Roman Britain. London: British Museum Press. Allason-​Jones, L. (1995). ‘“Sexing” Small Finds’, in P. Rush (ed.), Theoretical Roman Archaeology: Second Conference Proceedings. Aldershot: Avebury, 22–​32. Allason-​Jones, L. (2001). ‘Material Culture and Identity’, in S. James and M. Millett (eds), Britons and Romans:  Advancing an Archaeological Agenda. CBA Research Report 125. York: Council for British Archaeology, 19–​25.

The Development of Artefact Studies    87 Allason-​ Jones, L. (2011). Artefacts in Roman Britain:  Their Purpose and Use. Cambridge: Cambridge University Press. Allison, P. M. (2004). Pompeian Households:  An Analysis of Material Culture. Los Angeles: Cotsen Institute of Archaeology, University of California, Los Angeles. Archives Hub (2012). Joan Liversidge: Administrative/​Biographical History (accessed 27 January 2012). Ashby, T., Hudd, A. E., and Martin, A. T. (1904). ‘VII:  Excavations at Caerwent, Monmouthshire, on the Site of the Romano-​British city of Venta Silurum, in the years 1901-​ 1903’, Archaeologia, 59: 87–​124. Babelon, E. (1901). Traité des Monnaies Grecques et Romaines [volume  1]. Bologna:  Forni Editore. Baddeley, W. S. (1922). ‘The Excavation at Cirencester, February–​March 1922’, Transactions of the Bristol and Gloucestershire Archaeological Society, 44: 101–​115. Beech, M. (2006). ‘The cometarium by John Taylor’, Bulletin of the Scientific Instrument Society, 88: 1–​6. Birley, A. (1964). Life in Roman Britain. London: Batsford. Bishop, M. C. (1996). Finds from Roman Aldborough:  A  Catalogue of Small Finds from the Romano-​British Town of Isurium Brigantium. Oxford: Oxbow Books. Blagg, T. (1989). ‘Art and Architecture’, in M. Todd (ed.), Research on Roman Britain, 1960–​ 1989. Britannia Monograph Series 11. London: Society for the Promotion of Roman Studies, 203–​218. Boon, G. C. (1966). ‘Gilt-​Glass Beads from Caerleon and Elsewhere’, Bulletin of the Board of Celtic Studies, 22: 104–​109. Boon, G. C. (1983). ‘Potters, Oculists and Eye-​Troubles’, Britannia, 14: 1–​12. Bowden, M. (1991). Pitt Rivers:  The Life and Archaeological Work of Lieutenant-​General Augustus Henry Lane Fox Pitt Rivers, DCL, FRS, FSA. Cambridge:  Cambridge University Press. Breeze, D. J. (1989). ‘The Northern Frontiers’, in M. Todd (ed.), Research on Roman Britain, 1960–​89. Britannia Monograph Series 11. London:  Society for the Promotion of Roman Studies, 37–​60. Burnham, B., Collis, J., Dobinson, C., Haselgrove, C., and Jones, M. (2001). ‘Themes for Urban Research’, in S. James and M. Millett (eds), Britons and Romans: Advancing an Archaeological Agenda. CBA Research Report 125. York: Council for British Archaeology, 67–​76. Burton, A. (2004). ‘Smith, Robert Henry Soden (1822–​1890)’, Oxford Dictionary of National Biography. Oxford: Oxford University Press (online edn). Bushe-​Fox, J. P. (1913). Excavations on the Site of the Roman Town at Wroxeter, Shropshire in 1912. Research Report of the Society of Antiquaries of London, volume 1. Oxford: Oxford University Press for the Society of Antiquaries of London. Bushe-​Fox, J. P. (1926). First Report on the Excavation of the Roman Fort at Richborough, Kent. Research Report of the Society of Antiquaries of London, volume 6. Oxford: Oxford University Press for the Society of Antiquaries of London. Bushe-​Fox, J. P. (1932). Third Report on the Excavation of the Roman Fort at Richborough, Kent. Research Report of the Society of Antiquaries of London, volume 10. Oxford: Oxford University Press for the Society of Antiquaries of London. Bushe-​Fox, J. P. (1949). Fourth Report on the Excavation of the Roman Fort at Richborough, Kent. Research Report of the Society of Antiquaries of London, volume 16. Oxford: Oxford University Press for the Society of Antiquaries of London.

88

88   Ellen Swift Casey, P. J. (1974). ‘The Interpretation of Romano-​British Site Finds’, in P. J. Casey and R. Reece (eds), Coins and the Archaeologist. BAR British Series 4. Oxford: Archaeopress, 37–​51. Casey, P. J., and Reece, R. (1974). Coins and the Archaeologist. BAR British Series 4. Oxford: Archaeopress. Champion, T. (2004). ‘Hawkes, Charles Francis Christopher (1905–​1992)’, Oxford Dictionary of National Biography. Oxford: Oxford University Press (online edn). Chapman, W. (1985). ‘Arranging Ethnology: A. H. L. F. Pitt Rivers and the Typological Tradition’, in G. W. Stocking (ed.), Objects and Others: Essays on Museums and Material Culture. History of Anthropology Series, volume 3. Wisconsin: University of Wisconsin Press, 15–​48. Charlesworth, D. (1959). ‘Roman Glass in Northern Britain’, Archaeologia Aeliana, 4th edn. 37: 33–​58. Clark, K. (2008). ‘A Review of the Portable Antiquities Scheme’, Portable Antiquities Scheme Website. Museums, Libraries and Archives Council (accessed 23 May ​2012). Clarke, D. (1968). Analytical Archaeology. London: Methuen. Clarke, S. (1997). ‘Abandonment, Rubbish Disposal and “Special” Deposits at Newstead’, in K. Meadows, C. Lemke, and J. Heron (eds), TRAC 96:  Proceedings of the Sixth Annual Theoretical Roman Archaeology Conference, Sheffield 1996. Oxford:  Oxbow Books, 73–​81. Cleveland, Dowager Duchess of (1850). ‘Note on Four Gold Rings (One Roman) Found at Pierse Bridge, Durham and at Wanborough, Prudhoe Castle, and Colchester near Cambridge, Northumberland’, Archaeological Journal, 7: 191–​192. Collingwood, R. G. (1930a). ‘Romano-​Celtic Art in Northumbria’, Archaeologia, 80: 37–​58. Collingwood, R. G. (1930b). The Archaeology of Roman Britain. London: Methuen. Collingwood, R. G., and Richmond, I. A. (1969). The Archaeology of Roman Britain. 2nd edn. London: Methuen. Cool, H. E. M. (1998). ‘Life in Roman Castleford’, in H. E. M. Cool and C. Philo (eds), Roman Castleford. Volume 1: The Small Finds. Yorkshire Archaeology Series 4. Wakefield: Yorkshire Archaeology Service, 355–​373. Cool, H. E. M. (2004). The Roman Cemetery at Brougham. Britannia Monograph Series 21. London: Society for the Promotion of Roman Studies. Cool, H. E.  M. (2006). Eating and Drinking in Roman Britain. Cambridge:  Cambridge University Press. Cool, H. E.  M., and Baxter, M. J. (2002). ‘Exploring Romano-​British Finds Assemblages’, Oxford Journal of Archaeology, 21/​4: 365–​380. Cool, H. E.  M., and Philo, C. (1998) (eds). Roman Castleford. Volume 1:  The Small Finds. Yorkshire Archaeology Series 4. Wakefield: Yorkshire Archaeology Service. Cooper, N. J. (2007). ‘Promoting the Study of Finds in Roman Britain: Democracy, Integration and Dissemination: Practice and Methodologies for the Future’, in R. Hingley and S. Willis (eds), Roman Finds: Context and Theory. Oxford: Oxbow Books, 35–​52. Corder, P., and Kirk, J. (1932). A Roman Villa at Langton, near Malton, East Yorkshire. Leeds: Roman Antiquities Committee of the Yorkshire Archaeological Society. Crane, S. (1999). ‘Story, History and the Passionate Collector’, in M. Myrone and L. Pelz (eds), Producing the Past:  Aspects of Antiquarian Culture and Practice 1700–​ 1850. Woodbridge: Ashgate, 187–​203. Crummy, N. (1983). The Roman Small Finds from Excavations in Colchester. Colchester Archaeological Reports 2. Colchester: Colchester Archaeological Trust.

The Development of Artefact Studies    89 Cunliffe, B. (1968). Fifth Report on the Excavations of the Roman Fort at Richborough, Kent. Research Report of the Society of Antiquaries of London, volume 23. Oxford:  Oxford University Press for the Society of Antiquaries. Curle, J. (1911). A Roman Frontier Post and its People: The Fort of Newstead in the Parish of Melrose. Glasgow: James Maclehose and Sons for the Society of Antiquaries of Scotland. Curle, J. (1933). ‘An Inventory of Objects of Roman and Provincial Roman Objects Found on Sites in Scotland not Definitely Associated with Roman Constructions’, Proceedings of the Society of Antiquaries of Scotland, 66: 277–​400. Dalton, O. M. (1912). Catalogue of the Finger-​Rings, Early Christian, Byzantine, Teutonic, Medieval and Later, in the British Museum. London: Longmans. Dalton, O. M. (1922). ‘Roman Spoons from Dorchester’, Antiquaries Journal, 2: 89–​92. Donovan, H. (1935). ‘Roman Finds in Bourton-​on-​the-​Water’, Transactions of the Bristol and Gloucestershire Archaeological Society, 57: 234–​259. Dungworth, D. (1998). ‘Mystifying Roman Nails: Clavus annalis, defixiones and Minkisi’, in C. Forcey, J. Hawthorne, and R. Witcher (eds), TRAC 97: Proceedings of the Seventh Annual Theoretical Roman Archaeology Conference, Nottingham 1997. Oxford:  Oxbow Books, 149–​159. Eckardt, H. (2006). ‘The Character, Chronology and Use of the Late Roman Pits: The Silchester Finds Assemblage’, in M. Fulford, A. Clarke, and H. Eckardt (eds), Life and Labour in Late Roman Silchester:  Excavations in Insula IX since 1997. Britannia Monograph Series 22. London: Society for the Promotion of Roman Studies, 221–​244. Eckardt, H., and Crummy, N. (2008). Styling the Body in Late Iron Age and Roman Britain: A Contextual Approach to Toilet Instruments. Instrumentum Monograph Series 36. Montagnac: Instrumentum. Esmonde Cleary, S. (1989). ‘Constantine I  to Constantine III’, in M. Todd (ed.), Research on Roman Britain, 1960–​1989. Britannia Monograph Series 11. London:  Society for the Promotion of Roman Studies, 235–​244. Evans, A. J. (1896). ‘On Two Fibulae of Celtic Fabric from Aescia’, Archaeologia, 55: 179–​198. Evans, C. (2007). ‘“Delineating Objects”: Nineteenth Century Antiquarian Culture and the Project of Archaeology’, in S. Pearce (ed.), Visions of Antiquity: The Society of Antiquaries of London 1707–​2007. London: Society of Antiquaries, 267–​305. Evans, J. (1894). ‘IX: On Some Iron Tools and Other Articles Formed of Iron Found at Silchester in the Year 1890’, Archaeologia, 54: 139–​156. Evans, J. (1956). A History of the Society of Antiquaries. Oxford: Oxford University Press. Evans, J. (1995). Roman Finds Assemblages: Towards an Integrated Approach?’, in P. Rush (ed.), Theoretical Roman Archaeology: Second Conference Proceedings. Aldershot: Avebury, 33–​58. Faussett, B. (1856 [1773]). Inventorium Sepulchrale: An Account of some Antiquities Dug up at Gilton, Kingston, Sibertswold, Barfriston, Beakesbourne, Chartham, and Crundale, in the County of Kent, ed. C. Roach-​Smith. London: T. Richards. Fenwick, C. Wiggins, M., and Wythe, D. (2008) (eds). TRAC 2007: Proceedings of the Seventeenth Annual Theoretical Roman Archaeology Conference, London 2007. Oxford: Oxbow Books. Fieldhouse, W. J., May, T., and Wellstood, F. C. (1931). A Romano-​British Industrial Settlement near Tiddington, Stratford-​upon Avon. Birmingham: Corporation of Stratford-​upon-​Avon and the Trustees and Guardians of Shakespeare’s Birthplace. Fincham, G., Harrison, G., Holland, R., and Revell, L. (2000) (eds). TRAC 99:  Proceedings of the Ninth Annual Theoretical Roman Archaeology Conference, Durham 1999. Oxford: Oxbow Books.

90

90   Ellen Swift Fox, G. E. (1893). ‘Excavations on the Site of the Roman City at Silchester, Hants, in 1894’, Archaeologia, 54: 439–​494. Franks, A. (1858). ‘On Bosses of Roman Shields Found in Northumberland and Lancashire’, Archaeological Journal, 15: 55–​58. Franks, A. (1867–​70). ‘Note on a Roman Enamel Bowl Found at Broughing, Herts’, Proceedings of the Society of Antiquaries of London. 2nd ser. 4: 514–​516. Freeman, P. W. M. (2007). The Best Training Ground for Archaeologists: Francis Haverfield and the Invention of Romano-​British Archaeology. Oxford: Oxbow Books. Frere, S. S. (1967). Britannia: A History of Roman Britain. London: Routledge and Kegan Paul. Fulford, M. (1989). ‘The Economy of Roman Britain’, in M. Todd (ed.), Research on Roman Britain, 1960–​1989. Britannia Monograph Series 11. London: Society for the Promotion of Roman Studies, 175–​202. Fulford, M. (2007). ‘Foreword’, in R. Hingley and S. Willis (eds), Roman Finds: Context and Theory. Oxford: Oxbow Books, pp. ix–​x. Gardner, A. (2007). An Archaeology of Identity: Soldiers and Society in Late Roman Britain. Walnut Creek, CA: Left Coast Press. Gould, J. (1966–​7). ‘Excavations at Wall (Staffordshire) 1961–​63 on the Site of the Early Roman Forts and of the late Roman Defences’, Transactions of the Lichfield and South Staffordshire Archaeological and Historical Society, 8: 1–​38. Graham Ritche, J. N. (2002). ‘James Curle (1862–​1944) and Alexander Ormiston Curle (1866–​ 1955):  Pillars of the Establishment’, Proceedings of the Society of Antiquaries of Scotland, 132: 19–​41. Green, M. J. (1978). A Corpus of Small Cult Objects from the Military Areas of Roman Britain. BAR British Series 52. Oxford: Archaeopress. Hattatt, R. (1982). Ancient and Romano-​British Brooches. Oxford: Dorset Publishing Company. Hattatt, R. (1985). Iron Age and Roman Brooches:  A  Second Selection of Brooches from the Author’s Collection. Oxford: Oxbow Books. Haverfield, F. (1891). ‘Roman Remains in Local Museums’, Antiquary, 24: 168–​172. Haverfield, F. (1911). ‘Report of the 1910 Excavations at Corstopitum:  Smaller Objects’, Archaeologia Aeliana, 3rd ser. 7: 176–​202. Haverfield, F. (1912). The Romanization of Roman Britain. 2nd edn. Oxford: Clarendon Press. Haverfield, F. (1919). ‘The Tullie House fibulae’, Transactions of the Cumberland and Westmorland Antiquarian and Archaeological Society, ns 19: 1–​16. Hawkes, C. F. C. (1940). ‘An “Aylesford” La Tene III Brooch from Arundel Park and the Dating of the Type’, Antiquaries Journal, 20: 492–49​5. Hawkes, C. F. C. (1982). ‘Mark Reginald Hull, 1897–​1976’, Essex Archaeology and History, 3/​ 14: 1–​2. Hawkes, C. F. C., and Hull, M. R. (1947). Camulodunum: First Report on the Excavations at Colchester 1930–​39. Research Report of the Society of Antiquaries of London, volume 14. Oxford: Oxford University Press for the Society of Antiquaries of London. Hawkes, S., and Dunning, G. (1961). ‘Soldiers and Settlers in Britain’, Medieval Archaeology, 5: 1–​70. Henig, M. (1974). A Corpus of Roman Engraved Gemstones from British Sites. BAR British Series 8. Oxford: Archaeopress. Henig, M. (1989). ‘Religion in Roman Britain’, in M. Todd (ed.), Research on Roman Britain, 1960–​1989. Britannia Monograph Series 11. London: Society for the Promotion of Roman Studies, 219–​234.

The Development of Artefact Studies    91 Henig, M., and Soffe, G. (2002). ‘Graham Webster—​Archaeologist’, Bulletin of the Association for Roman Archaeology, 12: 3–​5, 8. Herendeen, W. (2007). William Camden: A Life in Context. Woodbridge: Boydell. Hildyard, E. J. W. (1945). ‘The Ancestry of the Trumpet-​Fibula’, Antiquaries Journal, 25: 154–​158. Hildyard, E. J. W. (1950). ‘The Date of the Malton Brooch’, Antiquaries Journal, 30: 74. Hill, J. D. (1995). Ritual and Rubbish in the Iron Age of Wessex. BAR British Series 242. Oxford: Archaeopress. Hingley, R. (2007). ‘The Society, its Council, the Membership and Publications, 1820–​50’, in S. Pearce (ed.), Visions of Antiquity:  The Society of Antiquaries of London 1707–​2007. London: Society of Antiquaries, 173–​198. Hingley, R. (2008). The Recovery of Roman Britain. Oxford: Oxford University Press. Hingley, R., and Willis, S. (2007). Roman Finds: Context and Theory. Oxford: Oxbow Books. Hodder, I. (1982). Symbols in Action:  Ethnoarchaeological Studies of Material Culture. Cambridge: Cambridge University Press. Hutcheson, A. J.  R. (1997). ‘Native or Roman? Ironwork Hoards in Northern Britain’, in K. Meadows, C. Lemke, and J. Heron (eds), TRAC 96:  Proceedings of the Sixth Annual Theoretical Roman Archaeology Conference, Sheffield 1996. Oxford:  Oxbow Books, 65–​72. Jackson, R. (1985). ‘Cosmetic Sets from Late Iron Age and Roman Britain’, Britannia, 16: 165–​192. James, S. (2003). ‘Roman Archaeology: Crisis and Revolution’, Antiquity, 77: 178–​184. Jenkins, F. (1985). ‘A Group of Silvered Bronze Horse-​Trappings from Xanten’, Britannia, 16: 141–​164. Johns, C. (2007). ‘The Last Chance’, in R. Hingley and S. Willis (eds), Roman Finds in Context and Theory. Oxford: Oxbow Books, 29–​34. Johns, C., Thompson, H., and Wagstaff, P. (1980). ‘The Wincle, Cheshire, Hoard of Roman Gold Jewellery’, Antiquaries Journal, 60: 48–​58. Jones, R. F.  J. (1987). ‘The Archaeologists of Roman Britain’, Bulletin of the Institute of Archaeologists of London, 24: 85–​97. Jundi, S., and Hill, J. D. (1998). ‘Brooches and Identities in First Century AD Britain: More than Meets the Eye?’, in C. Forcey, J. Hawthorne, and R.Witcher (eds), TRAC 97: Proceedings of the Seventh Annual Theoretical Roman Archaeology Conference, Nottingham 1997. Oxford: Oxbow Books, 125–​137. Kendrick, T. D. (1950). British Antiquity. London: Methuen. Kenyon, K. (1948). Excavations at the Jewry Wall Site, Leicester. Research Report for the Society of Antiquaries of London, volume 15. Oxford: Oxford University Press for the Society of Antiquaries of London. Kenyon, K. (1950). ‘Excavations at Breedon Hill, Leicestershire 1946’, Transactions of the Leciestershire Archaeological Society, 26: 17–​82. Kenyon, K. (1952). Beginning in Archaeology. London: Phoenix House. Kenyon, K. (1959). Excavations in Southwark 1945–​ 47. Research Papers of the Surrey Archaeological Society 5. Guildford: Surrey Archaeological Society. Keppie, L. (1989). ‘Beyond the Northern Frontier: Roman and Native in Scotland’, in M. Todd (ed.), Research on Roman Britain 1960–​1989. Britannia Monograph Series 11. London: Society for the Promotion of Roman Studies, 61–​74. Liversidge, J. (1968). Britain in the Roman Empire. London: Routledge and Kegan Paul. Lloyd-​ Morgan, G. (1977). Mirrors in Roman Britain. BAR British Series 41. Oxford: Archaeopress.

92

92   Ellen Swift Lloyd-​Morgan, G. (1981). ‘Jet and Shale in the Archaeological Collections of the Grosvenor Museum, Chester’, Journal of the Chester Archaeological and Historical Society, 64: 41–​46. Lockyear, K. (2007). ‘Where do we Go from here? Recording and Analysing Roman Coins from Archaeological Excavations’, Britannia, 38: 211–​224. Lucas, G. (2001). Critical Approaches to Fieldwork: Contemporary and Historical Archaeological Practice. London and New York: Routledge. Macgregor, A. (1976). Finds from a Roman Sewer System and an Adjacent Building in Church Street. The Archaeology of York, volume 17, fasc. 1. York: Council for British Archaeology for the York Archaeological Trust. Macgregor, A. (1978). Roman Finds from Skeldergate and Bishopshill. The Archaeology of York, volume 17, fasc. 2. York: Council for British Archaeology for the York Archaeological Trust. Manning, W. H. (1976). Catalogue of Romano-​British Ironwork in the Museum of Antiquities, Newcastle upon Tyne. Newcastle upon Tyne:  Department of Archaeology, University of Newcastle upon Tyne. Manning, W. H. (1995). ‘George C. Boon (1927–​1994)’, Britannia, 26: pp. xiii–​xiv. Manning, W., Price, J., and Webster, J. (1995). The Roman Small Finds: Report on the Excavations at Usk 1965–​1976, Volume 7. Cardiff: University of Wales Press. Marshall, F. H. (1911). Catalogue of the Jewellery, Greek, Etruscan and Roman, in the Departments of Antiquities, British Museum. London: British Museum Department of Greek and Roman Antiquities. Mattingly, D. (2008). ‘John Dore:  Archaeologist of Northern Britain and Libya who was a Specialist in Mediterranean Ceramics’, Independent, 25 June 2008. Mattingly, H., Sydenham, E., Sutherland, C., and Carson, R. (1923–​84) (eds). The Roman Imperial Coinage. Volume 1: From 31 BC to AD 69. London: Spink and Son. Merrifield, R. (1987). The Archaeology of Ritual and Magic. London: Batsford. Miller, E. (1973). That Noble Cabinet: A History of the British Museum. London: André Deutsch. Nash-​Williams, V. E. (1932). ‘The Roman Legionary Fortress at Caerleon:  Report on the Excavations Carried out in the Prysg Field 1927–​29. Part II:  The Finds’, Archaeologia Cambrensis, 87: 48–​104. Nurse, B. (2004a). ‘Way, Albert (1805–​ 1874)’, Oxford Dictionary of National Biography. Oxford: Oxford University Press (online edn). Nurse, B. (2004b). ‘Cox, John Charles (1843–​1919)’, Oxford Dictionary of National Biography. Oxford: Oxford University Press (online edn). O’Neil, J. (1935). ‘Coins and Archaeology’, Archaeological Journal, 92: 64–​80. Oswald, F. (1952). Excavation of a Traverse of Margidunum. Nottingham:  Nottingham University Press. Painter, K., and Thompson, H. (1994). ‘Obituary: Donald Harde’, Independent, 29 April 1994. PAS (Portable Antiquities Scheme) (2012). (accessed 22 May 2012). Peacock, D. P.  S. (1982). Pottery in the Roman World:  An Ethnoarchaeological Approach. London: Longman. Pevensey Excavation Committee (1908). Excavations on the Site of the Roman Fortress at Pevensey: First Report. Lewes/​Eastbourne/​East Grinstead: Privately printed for subscribers. Philpott, R. (1991). Burial Practices in Roman Britain:  A  Survey of Grave Treatment and Furnishing. BAR British Series 219. Oxford: Archaeopress. Piggott, S. (1989). Ancient Britons and the Antiquarian Imagination: Ideas from the Renaissance to the Regency. London: Thames and Hudson.

The Development of Artefact Studies    93 Pitt-​Rivers, A. H. L. F. (1887). Excavations in Cranborne Chase near Rushmore on the borders of Dorset and Wiltshire. Volume 1: Excavations in the Romano-​British Village of Woodcuts Common and Romano-​British Antiquities in Rushmore Park. Privately Printed. Pitts, M. (2010). ‘Artefact Suites and Social Practice: An Integrated Approach to Roman Finds Assemblages’, Facta: A Journal of Roman Material Culture Studies, 4: 125–​152. Quennell, M., and Quennell, C. H. B. (1937). Everyday Life in Roman Britain. 2nd edn. The Everyday Life Series. London: Batsford. Reece, R. (1973). ‘A Short Survey of the Roman Coins Found on Fourteen Sites in Britain’, Britannia, 3: 69–​76. Rees, H., Crummy, N., Ottaway, P. J., and Dunn, G. (2008). Artefacts and Society in Roman and Medieval Winchester:  Small Finds from the Suburbs and Defences, 1971–​ 1986. Winchester: Winchester Museums Service. Roach-​Smith, C. (1850). Antiquities of Richborough, Reculver and Lympne. London:  John Russell Smith. Roach-​Smith, C. (1859). Illustrations of Roman London. Unpublished: Private Subscription. Roach-​Smith, C. (1868a). Collectanea Antiqua:  Etchings and Notices of Ancient Remains, Illustrative of the Habits, Customs, and History of Past Ages. London: John Russell Smith. Roach-​Smith, C. (1868b). ‘On the Red-​Glazed Pottery of the Romans’, Journal of the British Archaeological Association, 4: 1–​19. Romilly Allen, J. (1901). ‘Romano-​British Fibulae Showing Late-​Celtic Influence’, Reliquary and Illustrated Archaeology, ns 7: 195–​199. Savory, H. N. (1956). ‘Some Sub-​Romano-​British Brooches from South Wales’, in D. B. Harden (ed.), Dark Age Britain. London: Methuen and Company, 40–​58. Serpell, M. (1982). ‘Obituary: Dorothy Charlesworth’, Antiquaries Journal, 62: 487–​504. Smiles, S. (2007). ‘Art and Antiquity in the Long Nineteenth Century’, in S. Pearce (ed.), Visions of Antiquity: The Society of Antiquaries of London 1707–​2007. London: Society of Antiquaries, 123–​145. Strong, D. and Brown, D. (1976) (eds.). Roman Crafts. London: Duckworth. Sutherland, C., and Carson, R. (1923) (eds). The Roman Imperial Coinage. Volume 1: From 31 BC to AD 69. London: Spink and Son. Swift, E. (2007). ‘Small Objects, Small Questions? Perceptions of Finds Research in the Academic Community’, in R. Hingley and S. Willis (eds), Roman Finds in Context and Theory. Oxford: Oxbow Books, 18–​28. Swift, E. (2010). ‘Identifying Migrant Communities:  A  Contextual Analysis of Grave Assemblages from Continental Late Roman Cemeteries’, Britannia, 41: 237–​282. Todd, M. (1989). Research on Roman Britain, 1960–​89. Britannia Monograph Series 11. London: Society for the Promotion of Roman Studies. Tomber, R., and Dore, J. (1998). The National Roman Fabric Reference Collection: A Handbook. London: Museum of London Archaeology Service (accessed 25 October 2011). Tonnochy, A. B., and Hawkes, C. F. C. (1931). ‘The Sacred Tree Motif on a Roman Bronze from Essex’, Antiquaries Journal, 11: 123–​128. Toynbee, J. (1962). Art in Roman Britain. London: Phaidon. Tyers, P. (1996). Roman Pottery in Britain. London: Batsford. van Driel Murray, C. (1999). ‘ “And did those feet in ancient time . . . ”: Feet and Shoes as a Material Projection of the Self ’, in P. Baker, C. Forcey, S. Jundi, and R. Witcher (eds), TRAC

94

94   Ellen Swift 98: Proceedings of the Eighth Annual Theoretical Roman Archaeology Conference: Leicester 1998. Oxford: Oxbow Books, 131–​140. Wacher, J. (1969). Excavations at Brough on Humber 1958–​1961. Research Report of the Society of Antiquaries of London, volume 25. Oxford: Oxford University Press for the Society of Antiquaries of London. Wallace, C. (2002). ‘Writing Disciplinary History:  Or why Romano-​British Archaeology Needs a Biographical Dictionary of its Own’, Oxford Journal of Archaeology, 21/​4: 381–​392. Walters, H. B. (1908). Catalogue of the Roman Pottery in the Department of Antiquities, British Museum. London: British Museum Department of Greek and Roman Antiquities. Ward, J. (1911). The Roman Era in Britain. London: Methuen and Co. Warner, R. (1797). An Illustration of the Roman Antiquities discovered in Bath. Bath: The Mayor and Corporation. Webster, G. (1977). ‘Reflections on Romano-​British Pottery Studies, Past, Present and Future’, in J. Dore and K. Greene (eds), Roman Pottery Studies in Britain and Beyond. BAR British Series 30. Oxford: Archaeopress, 317–​333. Wedlake, W. J. (1958). Excavations at Camerton, Somerset: A Record of Thirty Years’ Excavation Covering the Period from Neolithic to Saxon Times, 1926–​56. Bath: Privately published. Wheeler, R. E. M. (1955). ‘Anniversary Address’, Antiquaries Journal, 35: 153–​161. Wheeler, R. E.  M., and Wheeler, T. V. (1928). ‘The Roman Amphitheatre at Caerleon, Monmouthshire’, Archaeologia, 78: 111–​218. Wheeler, R. E. M., and Wheeler, T. V. (1932). Report on the Excavation of the Prehistoric, Roman and Post-​Roman Site in Lydney Park, Gloucestershire. Research Report of the Society of Antiquaries of London, volume 9. Oxford: Oxford University Press for the Society of Antiquaries. White, R. H. (1988). Roman and Celtic Objects from Anglo-​Saxon Graves: A Catalogue and an Interpretation of their Use. BAR British Series 191. Oxford: Archaeopress. Whiting, W., Hawley, W., and May, T. (1931). Report on the Excavation of the Roman Cemetery at Ospringe, Kent. Research Report of the Society of Antiquaries of London, volume 8. Oxford: Oxford University Press for the Society of Antiquaries. Willis, S. (2004). ‘The Study Group for Roman Pottery: Research Framework Document for the Study of Roman Pottery in Britain, 2003’, Journal of Roman Pottery Studies, 11: 1–​20. Wilson, D. M. (1997). ‘Augustus Wollaston Franks: Towards a Portrait’, in M. Caygill and J. Cherry (eds), Augustus Wollaston Franks:  Nineteenth-​Century Collecting and the British Museum. London: British Museum Press, 1–​5. Winbolt, S. E. (1925). Roman Folkestone: A Record of Excavation of Roman Villas at East Wear Bay. London: Methuen. Winbolt, S. E. (1935). ‘Report on Recent Finds at Chichester’, Sussex Archaeological Collections, 76: 136–​171. Wroth, W. W. (rev. Henig, M.) (2004). ‘King, Charles William (1818–​1888)’, Oxford Dictionary of National Biography. Oxford: Oxford University Press (online edn). Young, C. (2007). ‘Obituary: Joan Kirk’, Society of Antiquaries Online Newsletter, 176 (accessed 20 January ​2012).

Chapter 5

The Textua l a nd Archaeol o gica l Ev i de nc e Henry Hurst

Introduction Because we like stories, in common with all human beings, and also because of our logocentric culture, in which we communicate most information in the form of writing, texts are usually given a privileged status when they survive from a period of the past in which we are interested. For Roman Britain the longest-​known and, until recently, most accessible texts were those of classical authors who wrote about the province, in Latin or Greek, always from a distance and in passages that were secondary to the wider purpose of their works. The collective name for these passages, ‘sources’, reflects the way they were traditionally regarded as the fount of knowledge. They are not particularly abundant for Roman Britain. What survives most fully is coverage of the decades between the invasion of ad 43 and the governorship of Agricola (ad 78–​85), with a particular focus on military history. For later events shorter passages and spasmodic references survive, and there is a body of geographical information in ancient writers and road itineraries (Mann and Penman 1996; Ireland 2008; summarized in Mattingly 2006: 26–​38). These have been supplemented by ingenious scholarship, partly about texts found within Britain and partly about the range of other material evidence found by archaeology. Until recently most of the known texts from Britain were inscriptions on stone, of which 2,400 were published collectively as Volume I of the Roman Inscriptions of Britain (hereafter RIB: Collingwood and Wright 1965; indexes by Goodburn and Waugh 1983). These tended to skew the overall view even more towards military history, since the army had a stronger ‘epigraphic habit’ than civilians (neatly shown in the case of religion by Millett (2005: figures 88–​89)) and more of what it inscribed survives, because it spent most time in those parts of Britain where stone suitable for inscribing is freely available. The way in which archaeology was done until the 1980s, with its emphasis on military, urban, and

96

96   Henry Hurst villa sites and an excavation-​dominated approach to their study, also tended to reinforce this. A traditional format for histories of Roman Britain was, therefore, a military and political narrative with annexed chapters on ‘industry’, commerce, and cultural life (see Salway 1981; Frere 1987; Wacher 1998). As discussed more fully elsewhere in this volume (see chapters by Wilson, Millett, and Swift, this volume), a reaction set in from the 1980s, leading to a series of different emphases, mostly approaching the history more from a ‘native’ standpoint; these works sometimes identified themselves as ‘postcolonial’, in contrast with earlier scholarship, seen as linked with British imperialism, which had itself been partly influenced by Roman texts (see Millett 1990; Mattingly 1997, 2006; Hingley 2000, 2005; Gardner 2007). These approaches tended to emphasize the links between prehistoric and Roman Britain, through a whole range of archaeological studies. Such links were strengthened by the observation that the vast bulk of the Romano-​ British population was illiterate.

Types of Text from within Britain Meanwhile a quiet revolution was taking place, with the publication of large quantities of new textual evidence. As a result of the extensive recovery in excavations of anaerobically preserved material, we have learnt that a simple and cheap writing technology, using carbon ink on thin sheets of wood, was widespread in Roman Britain (as probably it was throughout the Roman world). The earliest publication of a text written in ink from Britain may be that by E. G. Turner (1956), of a deed of sale on a stylus tablet of larch wood from the well of a villa at Chew Stoke in Somerset; a letter of ink on wood, sent from Rochester to London, was published four years later (Turner and Skutsch 1960), but it was only with the discovery of the texts from the military site at Vindolanda just south of Hadrian’s Wall, from 1973 onwards, that the abundance of writing in these materials, and its significance, came to be appreciated (Figure 5.1; Bowman and Thomas (1983: 33–​37) give a list of the evidence known at that time). Our previous assumption, that most Roman letter-​and other writing for routine purposes was incised with a metal stylus on wax in a specially made tablet, was more restrictive to the practice of writing, since these materials required several skills and resources in their production. Stylus tablets and, more so, metal styli are nevertheless also fairly widespread finds in Roman Britain, but the tablets offer us written content only where the stylus has penetrated into the wood below the wax (Figure 5.2). The first collective volume giving the texts of the ink inscriptions on wood from Vindolanda was Bowman and Thomas (1983), two more volumes followed in 1994 (Bowman and Thomas 1994), and two further primary publications of texts in 2010 and 2011 (Bowman et al. 2010, 2011). These have been accompanied by many ancillary publications, of which Bowman (2003) is the most comprehensive. This site has produced about 400 substantial, intelligible texts (with fragments of hundreds of others), which form the largest body of written documentation of life in a military community from anywhere in the Roman

The Textual and Archaeological Evidence    97

Figure 5.1  Memorandum on wooden leaf tablet from Vindolanda, referring to the British use of cavalry (Tab. Vindol. II. 164): dimensions 78 × 186 millimetres. After a punctuation mark in l. 4, the text reads: ‘nec residunt | Brittunculi ut iaculos | mittant’ (‘nor do the Brits mount in order to throw javelins’). Source: full text and translation in Bowman and Thomas (1994: 107–​108). © The Trustees of the British Museum.

world. They were mostly discarded material from the commander’s house at the end of the first century ad, though others have also been found in a barrack block and the workshop of the fort (Pearce (2004: 48–​49) describes the complexities of their deposition) and are not so much formal documents about military organization, though such

98

98   Henry Hurst

Figure 5.2  Sale of a slave girl, Fortunata, recorded on a stylus tablet from London, drawn by R. Tomlin: dimensions c. 140 × 114 millimetres, probably one of three tablets. The writing survived as scratches in the wooden tablet, becoming illegible where a triangular patch of the original wax still survived; [p]‌uellam Fortunatam are the first two words of l. 3. Source: full text and translation in Tomlin (2003, 2011). © Roger Tomlin.

material is present, as a written reflection of life in this military community from the viewpoint of the commander’s household. They therefore contain domestic material ranging from accounts and arrangements for supplies to personal correspondence. To the extent that they offer a view of life in Britain from the standpoint of a military community in a frontier region and in one well-​known case refer to the native fighters in the diminutive as Brittunculi (‘Brits’: Figure 5.1), they might seem to bear out a stereotype of the Romans as colonialists. However, as will be seen, they are more remarkable for what they reveal of the complexity of communications and contact, and indeed of the colonial process itself. There are also major collections of writing tablets from London and Carlisle (estimated for stylus and ink tablets combined as 200+ and 150+ respectively by Pearce (2004:  48), although smaller numbers have been published). The London finds are mostly stylus tablets and from two main areas, the river bank and areas adjacent to the Walbrook stream; the small number with readable texts have included legal and

The Textual and Archaeological Evidence    99 business documents, including the sales of a wood in Kent (Tomlin 1996) and a slave girl (Figure 5.2). The military site at Carlisle has yielded tablets inscribed in ink, from Annetwell Street, which are mainly fragments of accounts and letters, published by Tomlin (1998); stylus tablets from two other sites have revealed only addresses. The date range of these finds is later first to mid-​second century (Pearce 2004: 49). A contrasting set of texts, also published from the 1980s on, are the curses incised on lead sheet or defixiones (Figure 5.3). The contents of only about ten of these were

Figure 5.3  Curse tablet no. 43 from the sanctuary of Mercury at Uley, Gloucestershire, drawn by R. Tomlin: dimensions c. 95 × 83 millimetres. Docilinus asks Mercury to drive Varianus, Peregrina, and Sabinianus to death for injuring his farm animal unless they redeem their action. Source: text and translation in Hassall and Tomlin (1989: 329–​331, no. 3) and Tomlin (2002: 172). © Roger Tomlin.

100

100   Henry Hurst known until the finds made at the shrine of Mercury at Uley in Gloucestershire, the first three of which were published in Britannia 10 for 1979, and the major collection of 130 from the sacred spring at Bath, whose texts were fully published in 1988 (Tomlin 1988). Uley produced over 100, but because of the difficulty of unrolling the corroded lead these texts have been published more slowly: five, including the three in Britannia 10, were published in the main excavation report (Tomlin 1993), and seven more since, in Britannia 20, 23, 26, 27, and 29. Since then others have come to light, and in all over 300 are now known (Mattingly 2006: 310; the contents of sixty-​nine are summarized in his Table 11, including thirty-​three from Bath and twelve from Uley). In response to an appeal by the Oxford Centre for the Study of Ancient Documents, thirty-​five sites are now known to have produced one or more defixiones. Many remain unread—​ where it is either difficult or impossible to unroll the corroded lead sheets—​but this is nevertheless a growing source of textual information. As will be discussed, their contents have a broadly uniform framework, where most frequently the help of a deity is invoked to punish one or more thieves of personal property, but, even if filtered through conventions of form or expression and the medium probably of a scribe, this is as near as we can ever come to hearing the voices of the non-​elite population of Roman Britain. A third major publication of writing on materials other than stone in the 1990s were the eight fascicules (and a ninth of indexes) belonging to RIB Volume II (Collingwood and Wright 1990–​5). This ranged from official tile stamps and indications of weights and contents of goods on containers, through commercial makers’ stamps and other marks on products from silver to leather to scratched graffiti (see Figure 5.4), many simply recording ownership, but also containing dedications and other information. RIB II performed a most valuable service by bringing this disparate material up to date in conveniently collected form. These fascicules have an eccentric title page, in that their collective heading of ‘Instrumentum domesticum’ (‘private documents’ in a free translation) under-​represents the range of material and the naming of R. G. Collingwood (died 1943)  and R.  P. Wright (retired 1976, died 1992)  as principal authors piously records their pioneering contribution, but sends a misleading signal about scholarly works of the 1990s. The fascicules can be seen as a long-​term follow-​ up to the design of the broader Corpus Inscriptionum Latinarum (CIL), which from an early stage published other types of written evidence as well as public inscriptions on stone: so, for example, Volume IV (1871) contains graffiti and painted inscriptions on the walls of Pompeii and Volume XV (1891) is Heinrich Dressel’s study of stamped bricks from Rome published under the heading of Instrumentum domesticum. The first volume of RIB (Collingwood and Wright 1965) had covered (mainly) inscriptions on stone found up to the start of 1955; and a further 550 inscriptions on stone found up to the end of 2006, RIB III, were published in 2009 (Tomlin et al. 2009), bringing their total to just short of 3,000. At the present time, then, epigraphic evidence of all sorts from Roman Britain is, to a greater extent than ever before, fully published in a convenient form; and, as has been the case for many years, new discoveries are added each

The Textual and Archaeological Evidence    101

Figure  5.4 Tile with personal-​name graffito scratched before firing, Candid[us]. From a fourth-​century deposit at Hayton, East Yorkshire. Source: Tomlin and Hassall (2001: 393). Photograph: M. Millett.

year in the journal Britannia under the heading ‘Roman Britain in [the year preceding the issue date]’. If we add to these classes of writing inscriptions on coins and potters’ stamps, especially on imported samian (Hartley 1969; Hartley and Dickinson 2008–​12)—​it becomes apparent that, illiterate as the great majority of people in Roman Britain undoubtedly were, they encountered writing at virtually every turn. Even in the countryside coins and stamped and graffito-​marked pots and other objects were common; main roads had inscribed milestones, cult centres contained inscribed dedications and defixiones and (see Beard 1991 for a general discussion of writing and Roman religion; for further discussion on inscriptions and language see Hope and Mullen, this volume). There is some evidence generally for an association between writing and votive settings in settlement sites (Pearce 2004: 51, with references). In urban and military settlements, the volume and range of writing became immense, from inscribed funerary monuments on the approaches and monumental inscriptions over entry gates or arches to formal and informal writing on buildings and then functional writing for legal, commercial, and administrative purposes.

102

102   Henry Hurst

Ancient Uses of, and Responses to, Writing A first challenge therefore lies in the abundance of written evidence. In a very straightforward way the mass of writing we now have challenges the old view that there are, on the whole, only some frustratingly short texts of classical authors, with little written evidence of note from within Roman Britain. ‘Our written sources are primarily external ones… comparatively little documentation survives that was produced in Britain itself,’ says Mattingly (2006: 21), even if in subsequent pages of the same volume he goes on to discuss the Vindolanda evidence and the defixiones in some detail. That statement is only true if ‘documentation’ is taken to mean historical narration in some form. Why, then, should writing be so common? Viewed from our culture of mass literacy, a natural inclination might be to assume that writing reflects some degree of functional literacy, using this term (and subsequently in this chapter) with a meaning as defined by the 1978 Unesco General Conference: ‘A person is functionally literate who can engage in all those activities in which literacy is required for effective functioning of his group and community and also for enabling him to continue to use reading, writing and calculation for his own and the community’s development’ (see also Harris 1989: ch. 1). But the reality in Roman Britain, as in the modern world, was clearly more nuanced than that. In the latter case, the ‘New Literacy Studies’ of the 1990s emphasized literacy and numeracy as social practices, embedded in social settings. ‘Literacy practices’ might then be exercised at ‘literacy events’, defined as ‘any occasion in which a piece of writing is integral to the nature of the participants’ interactions and their interpretative processes’ (Street 1984). The merit of this is that it provides a theoretical framework for all sorts of limited literacy. ‘The literacy as applied, practised and situated approach questions the validity of designations of individuals as “literate” or “illiterate”, as many who are labelled illiterate are found to make use of significant literacy practices for specific purposes in their everyday lives’ (Unesco 2006). We need not doubt the view that full functional literacy was confined to a small minority in the ancient world. In his wide-​ranging study, Harris (1989) argued that in very broad terms less than 10 per cent of the population of the Roman Empire is likely to have been literate. This broad order of magnitude has not been challenged since, and is thought to apply even where writing survives in such abundance as from Pompeii or Roman Egypt. For Roman Britain we need only note momentarily that, even if the entire resident army of some 40,00–​50,000 soldiers were fully literate (which would certainly not have been the case) and there was a comparable number of literate local elites, traders, and officials (which there would certainly not have been), these 80,00–​100,000 would be less than 5 per cent of a population estimated at between two million and six million (Millett 1990: 181–​186); the real figure was probably smaller. This is not altered by the discovery of short texts written in cursive script on tiles before they were fired (RIB 2491), including ‘a,b,c’ texts as of learners (2491.135–​145), metrical fragments, among them the phrase ‘conticuere omnes’ (‘all were silent’) which begins Aeneid book II, and evidence that the civitas we used to know as the Coritani may, more correctly, have been

The Textual and Archaeological Evidence    103 called the Corieltauvi (2491.150; Tomlin 1983), interesting as all of these are as anecdotal instances. The reality was that the great bulk of the population was not only untaught but had no general need to use reading or writing actively in their daily lives: a situation that in Britain was brought to an end only with the mass education of modern times, and in the world as a whole has been a major concern of Unesco up to the present time. The question of why writing should be so common in a largely illiterate society has been addressed for many ancient and early medieval societies, and answers have usually included the theme that it was an expression of power by mostly elite members of society: they shared written content, which was attainable only indirectly if at all by the majority; public displays of writing had symbolic meaning accentuated by various visual devices, from settings to the aesthetics of lettering (Goody 1977; McKitterick 1990; Bowman and Woolf 1994). All of this manifestly applied to Roman Britain (as, for example, Figure 5.5; an analysis of the lettering and layout of this inscription is given in

Figure 5.5  Tombstone of Classicianus from London: pieces as displayed in the British Museum with reconstruction of original monument drawn by Richard Grasby (Grasby and Tomlin 2002: figure 21), slightly adjusted. Reconstructed text (as in RIB 12): ‘Dis | [M]‌anibus | [G(ai) Iul(i) G(ai) f(ili) F]ab(ia tribu) Alpini Classiciani |… |… | proc(uratoris) provinc(iae) Brita[nniae] | Iulia Indi filia Pacata I[ndiana(?)] | uxor [f(ecit)]’ (‘To the spirits of the departed (and) of Gaius Julius Alpinus Classicianus, son of Gaius, of the Fabian voting-​tribe… procurator of the province of Britain, Julia Pacata I[ndiana], daughter of Indus, his wife, had this built’). Source: Photograph © The Trustees of the British Museum.

104

104   Henry Hurst Grasby 2012), but, before it is considered further, a methodological issue also arises—​ that of regarding texts as a type of artefact. This has been well argued by Moreland (2001) using examples of the early medieval period, but it applies equally to Roman Britain. We can see how writing was not always being used as we would use it. It is helpful, therefore, to think of it like, say, pottery or coins, and to try to reconstruct its ancient uses. Our understanding of Roman Britain would be limited by treating it as a ‘prehistoric’ (in the sense of unlettered) society, but it is almost equally limiting to focus just on the content of the surviving writing, without considering how its occurrence might have worked among the people who created and experienced it. The prime set of documents from Roman Britain indicating functional literacy are the Vindolanda texts, studied by Bowman and others: the following observations are dependent on their work (especially Bowman 2003). As Bowman has commented, a striking feature of this assemblage is the insight it gives into the existence of a literate ‘sub-​ elite’. Flavius Cerialis, the commander of a Batavian cohort by Hadrian’s Wall, and probably Lepidina his wife and their friends were not from the top level of Roman provincial society; they may only have been first-​or second-​generation ‘Romans’ originating from parts of modern Netherlands and north Germany. Yet he and Lepidina’s friend Claudia Severa seem to have been using written Latin for business and social purposes, as Cicero might have. As is seen from comparison of the hands in their personal correspondence, their circle often made use of scribes, adding only greetings at the beginning and end, but part of Cerialis’ correspondence in particular also seems to have been written in his own hand, including preliminary drafts. The variety of expression in the greetings of his and his wife’s group shows that individuals were able to express themselves directly in writing (Bowman 2003: 85). The geographical range of their contacts, from other military sites on Hadrian’s Wall to the garrison city of Carlisle to London and probably beyond, and the number of different hands—​‘several hundred’ (Bowman 2003: 85)—​suggest that the Vindolanda collection is the tip of a considerable iceberg of literate communication. The advantage of a locally available writing technology, of carbon ink on wood, has already been emphasized (stylus tablets were also present but in smaller numbers (Pearce 2004: figure 37)). There is, of course, a general military angle, that the Roman army looks to have been as paper-​bound as its modern successors, as shown by micromanagerial documents in the Vindolanda collection, complementing the military documents on papyrus from Dura Europus on the Eastern frontier (Welles et al. 1959), so that a career as an officer would develop an individual’s ability to express himself on paper. Illiterate ordinary soldiers also must have been steered in the direction of writing by belonging to an organization that used written communication so much. Retired auxiliary troops received written discharge certificates (RIB 2401), and veterans from all parts of the army probably also possessed other written documents. The more enterprising seem likely to have become functionally literate, so that when they joined civilian communities, not always close to where they had served (an extreme example is CIL XVI, 69, a discharge diploma of a British auxiliary found at Brigetio in Hungary), they may have contributed to the diffusion of literacy. The impact would have been strongest in the period of high military mobility in the first half-​century after the invasion of 43.

The Textual and Archaeological Evidence    105 Functional literacy would have also been spread among certain civilian professional people. The Vindolanda documents suggest this for traders doing significant business with the army (but see also Mairs 2012 on ‘interpretation’). Two deeds of sale from London (Figure 5.2; Tomlin 1996) and that from Somerset (Turner 1956) imply a class of functionally literate lawyers and their scribes, as might be expected. Literate individuals would also have been involved at all levels of the administration, from the governor and provincial procurator’s office, through civic units of coloniae, civitates, and so on, down to the organizers of craft and commercial guilds. But numbers here were almost certainly small: they need only have been hundreds overall. From the repertoire of figurative imagery on mosaics, especially of the fourth century, it seems likely that the Romano-​British elite shared the intellectual and spiritual tastes of other parts of the empire (cf. Scott 2000), even if the interpretation of individual cases is often arguable. The very limited direct evidence so far for their literary culture takes the form of inscriptions on mosaics and wall paintings. Writing at villa sites is shown by numerous finds of metal styli (Hanson and Conolly 2002; Mattingly 2006: 461), but we cannot say by whom or for what purpose. In a survey of the inscriptional evidence, Ling (2007) divided the content into signatures of artists and owners, good luck messages, labels for images portrayed, and ‘literary or pseudo-​literary quotations’. The last group included part of a quotation from the Aeneid on a wall painting in a villa at Otford, Kent (Aeneid 1. 313, or 12. 165) and the mosaics from the villas at Lullingstone, Kent, and Frampton, Somerset. In an earlier review of literary allusions from Roman Britain, Barrett (1978: 309–​313) argued that the Lullingstone floor, with the image of Europa and the bull and the inscriptional allusion to Juno’s jealousy in Aeneid 1. 50 (Figure 5.6), written in the style of Ovid, should be seen as the original expression of a literate man. Ling suggests that such a polished piece of writing may not have been original, while two possible ‘botched hexameters’ recorded by Lysons in the early nineteenth century on the Bellerophon mosaic at Frampton ‘are questionable in syntax and metre, and could have been made up by the patron or one of his peers’ (Ling 2007: 87). We cannot know, but this line of argument echoes a former view about Romano-​British art, that, if it was refined, it was probably imported, and, if it was blundering, as R. G. Collingwood once put it, it was British, and it seems unduly negative when placed in the wider context of fourth-​century classical culture in Roman Britain. One aspect of our appreciation of this has been the degree of content we try to read into the evidence, as illustrated by Henig’s suggestion (1997) that the Lullingstone inscription contained Christian cryptograms (cf. Neal and Cosh 2009: 385, on this) and the arguments for Gnostic interpretations of Frampton (Perring 2003; Tite 2010). In detail, these ideas often seem far-​fetched, but the principle, that complex thought-​processes are present in the subject matter that we struggle to understand, is surely right. RIB II illustrates many instances where the use and comprehension of written symbols were operating in a functional, but narrowly targeted, way, as, for example, the indications of contents, quantities, and sources of goods, in dipinti, stamps, and graffiti. Makers’ stamps on pottery and other goods, even leather (RIB 2445) and owners’ names scratched as graffiti, can be seen as combining this sort of limited functionality with

106

106   Henry Hurst

Figure 5.6  Europa and the bull with two cupids in front of the presumed semicircular couch in the apse of the dining room at Lullingstone villa, Kent. The inscription (RIB 2448.6) reads: ‘Invida si ta[uri] vidisset Iuno natatus | Iustius Aeolias isset adusque domos’ (‘If jealous Juno had seen the swimming of the bull, more justly would she have gone to the palace of Aeolus’). Source: © Alamy.

other important roles of writing: symbolic in both cases; and probably (to some degree) talismanic in the second case. The symbolic use of writing in Roman Britain can be seen to cover a spectrum from statements of power by the imperial authority to local applications. Lettering for Latin inscriptions can be regarded as one of the major original creations of Roman culture, with high levels of aesthetic subtlety (Gordon 1983): borrowing from a term applied to Roman narrative art (as by Brilliant 1984), it could be described as ‘imagistic’ in the sense that its appearance was intended to convey a message before its content (Figure 5.5). A striking example would have been the four-​way arch overlooking the coast at Richborough, Kent, at the entry to Britannia. The survival of a single large bronze letter A suggests that this monument to imperial power included a visually impressive inscription, with bronze letters glinting against their marble setting (Strong 1968: 68). Most of the 3,000 surviving inscriptions carved on stone of RIB I and III are also, to some extent, concerned with the visual impact of writing. Writing accompanying an image usually amplified the symbolism of the image but was not essential to its basic meaning. Thus,

The Textual and Archaeological Evidence    107 for example, the message of a triumphant Roman cavalryman is conveyed by the tombstone of Rufus Sita at Gloucester (RIB 121), irrespective of the accompanying inscription. The context of the writing could also send a message, as on building inscriptions, milestones, and many epitaphs: it was unnecessary to read the gateway inscriptions, distance slabs, and building stones of Hadrian’s Wall to grasp its meaning. However, for those who could read, formal inscriptions contained an extra layer of meaning, especially important for their authors, as Woolf (1996) argues. We can indeed see monumental writing being used to provide controlled access to meaning in a manner not unlike the stages of entry into a Roman house: a first stage is provided by its existence and the material used for it; a second stage by the context or any associated image; a third stage, of reading the content, was accessible only to a privileged minority. Their privilege would be enhanced by their ability to pass this information on to others, just as one might describe dining in a rich man’s house. This was a power game to be played by all who ‘had the epigraphic habit’. Their authorship is shown in RIB I and III: overwhelmingly the army or those in a military setting, with named individuals (predominantly with non-​Celtic names) and civic, commercial, and social entities following after them. Officially issued Roman coins give a miniaturized version of this, first as emblems of authority expressed in different materials (gold, silver, brass, and copper), then through imagery of the emperor’s head (easy to recognize) and reverse type (gradations of difficulty), inscriptions, and mintmark. A responsive use is also seen in local copies of the official coinage, as, for example, in the Claudian-​copy coins issued locally when there was a shortage of small change among coin-​using communities attached to military sites, in ad 48–​64 (Kenyon 1993). The emperor’s head remains recognizable if not in detail, while the reverse image and inscription are represented with varying degrees of abstractness. Similar tendencies can be seen with pre-​Roman Iron Age coins imitating Roman issues. Williams (2002) showed that in their style some inscriptions were influenced by, or ran parallel to, potters’ stamps on amphorae, as well as Arretine and Gallo-​Belgic wares. Whether a coin offers an illegible inscription or a cartouche in the manner of a potter’s stamp, the emphasis is towards the ‘imagistic’ or symbolical aspect of writing. Terra sigillata pottery stamps themselves develop from a significant minority presence of illiterate stamps or symbols in Arretine ware to more uniformly literate stamps of the Gallic potters of the mid-​first to late second centuries ad. This can be seen as a refinement of the top stage, of ‘content’, above the symbolism established previously. With the Gallic samian there is indeed a classic Romanizing process, where ‘recent Romans’, the potters, with Gallic rather than Roman names in many cases, carry the message of empire to those a step or two less Romanized than them (on the impact of this pottery, see Woolf 1998: ch. 7, especially pp. 193–​203, and, on linguistic aspects, Mullen 2013). Non-​literate stamps in British imitations of samian (Biddulph 2013) and in British-​produced mortaria (Hull 1963: 112–​114; Tyers 1996: 116–​135) also exemplify the symbolic role of potters’ stamps. Generally there is a huge volume of naming by makers and users on artefacts, from precious-​metal items down to coarse pottery: pre-​casting or pre-​firing graffiti as well as stamps can be distinguished from graffiti scratched on a finished product (Figure 5.4). Most of these names are of single individuals, roughly equally divided between Latin and

108

108   Henry Hurst Celtic names. There are also collective users as military units and some evidently commissioned objects, not necessarily of high value (for example, RIB 2496.2 and 3, mortaria, with graffiti cut before firing, the first for a century at Colchester, the second for a contubernium at Usk). A small minority of objects are marked with dedications to deities. An analysis of graffiti on pottery by Evans (1987) showed that about three-​quarters were personal names, followed by numbers and a collective ‘other’ category, and that names were more frequently scratched on fine tablewares. Naming on artefacts by users, as a mark of ownership or identity, develops the ‘responsive’ use of writing one stage further, for many of these individuals must have been illiterate, either capable just of scratching their name on the item or of getting somebody else to do that. That they chose to ‘borrow’ this medium is testimony to its power. It possibly also reflects a sense, exploited in a negative way in curse tablets, that a name embodied the person (cf. Gager 1992: 14), and thus a stronger bond would be set up between the object and person than would be the case for us writing our name in a book or sewing a nametape onto an article of clothing.

Curse Tablets The British examples of curse tablets or defixiones represent about a fifth of the total of 1,600 known throughout the Greco-​Roman world from the fifth century onwards (Gager 1992; Martin 2010). Their material is a first point of interest. Most surviving examples, including those from Britain, are on lead sheet or, in the case of those analysed from Bath, a pewter-​like alloy (Tomlin 1988: 82–​84). This might be thought to fit with their dark messages, invoking illness or death, and deposition in the ground, and occasionally the texts make such a connection—​‘cold lead and cold name’, cold and useless or heavy as the lead on which this is written (examples cited by Gager 1992: 3–​4). However, these texts also survive on other materials, from papyrus to potsherds, and lead had from early times a general use as a writing medium, so, while evidently favoured for this role, it was not used exclusively. As regards subject matter, various cursing situations are known: competitive, as in the theatre or circus; sexual; legal and political; commercial; and pleas for justice and revenge, especially relating to the theft or loss of possessions. As Tomlin (1988, 2002) has pointed out, the British examples almost exclusively belong to this last category (the contents of sixty-​nine, from twenty-​one possible sites, were summarized by Mattingly (2006: table 11); only one, from Old Harlow, possibly refers to losing in love). A common formula is to name the alleged culprit, offer the stolen article to the deity (as Sulis, Sulis Minerva, Mars or Mercury at Bath, or Mercury at Uley) and invite vengeance, paid in blood, through the loss of sleep and health or through death. As an example of their contents, Tomlin highlights an interesting pair, possibly of the same individual from Bath and from Uley: Docilianus, son of Brucetus to the most holy goddess Sulis. I devote him who has stolen my hooded cloak, whether man or woman, whether slave or free. May the

The Textual and Archaeological Evidence    109 goddess Sulis inflict him with the greatest death, and not allow him children now nor in the future, until he has brought my hooded cloak to the temple of her divinity. (Tabellae Sulis, no. 10: text and translation in Tomlin (2002: 167))

Uley no. 43 is from Docilinus to Mercury: ‘Varianus and Peregrina and Sabinianus who have brought evil on my farm animal… I ask that you drive them to the greatest death, and do not allow them health or sleep unless they redeem from you what they have administered to me’ (Figure 5.3). These are formulaic texts, following a preordained pattern, and there is the use often of judicial phraseology in them. It might be supposed that they were produced to order by scribes, and in certain non-​British examples this has been indicated by groups of texts with similar handwriting. In the British material, notably the large group from Bath, the variety not only of different hands, but of capacities to write, with elegant capitals in some cases but also blundering ones as if clumsily imitating a stone cutter, and of cursive from flowing to ‘pseudo-​texts’ whose writers may have been illiterates, has led to the suggestion of more direct involvement by the authors of the texts. There is also enough variation of expression and in some cases of language to suggest that they reflect to some degree the feelings of their authors (two seem to be rare examples of written Celtic (Mullen 2007)). The writing of some tablets could reflect something similar to a ‘literacy event’ in the modern world, but many were probably produced by scribes and literate ‘friends’. The Bath assemblage probably covered two centuries, so a variety of hands over this time is not so surprising (see Tomlin 1988: 98–​100). Behind their conventional expressions, these tablets record private conversations with the gods, since they were rolled up after being written and deposited in a shrine for the deity’s eyes only. With their belief in divine powers of supra-​human sight and capacity to inflict punishment, their theology fits within perceptions of divine power as expressed in classical literature. The unexceptional nature of many of the objects stolen—​an iron pan, a cloak, gloves, a small amount of money—​might be taken to reflect poverty. The brutality of the punishments proposed—​the shedding of blood, loss of health or death—​even if in itself conventional, combined with the use of pseudo-​ judicial language as if proposing a sentence, evokes the rough justice meted out to the poor of early modern times, where the theft of a sheep could result in a hanging. So it is tempting to relate these voices to the image given by the physical remains of Roman-​ Britons, as having lived mostly short, undernourished lives, and suggest that they are the thoughts of unprivileged members of a poorly policed society. Mattingly’s discussion (2006: 315) introduces an ‘us and them’ distinction by pointing out that only one of over 300 curse tablets known was from a military site: ‘soldiers enjoyed better protection in and above the law.’ But this seems likely to be an oversimplification. As Gager (1992: 177) points out, the depositional context of most of the surviving British tablets—​ in the sacred spring at Bath or buried at a sanctuary—​may give us a skewed view: we may simply lack those with different contents deposited in theatres or in the circus or in other contexts. As regards justice, as Martin (2010: 68) has said, these documents are unsubstantiated allegations. They show aggravation, then, but not necessarily injustice.

110

110   Henry Hurst With regard to imperialism, it is striking that these people used one of its prime tools—​ writing in Latin—​as a medium to express their thoughts. The content of these defixiones can perhaps be viewed in a similar way to letters to agony aunt columns in newspapers: they show people’s concerns, but give no balanced view of society; the modern concerns are, however, expressed publicly.

Texts Mainly Viewing Britain from  the Outside: Issues of History and Archaeology Writing about Britain includes a range of material from poetry to road itineraries, all of it fragmentary. At its core are the accounts of three historians: Tacitus, a Roman senator, writing around the turn of the first and second centuries, Dio Cassius, another senator, who came originally from Nicaea (modern Iznik in Turkey) and wrote in Greek a century after Tacitus, and Ammianus Marcellinus, from Antioch (Antakya in south-​ east Turkey), who lived c. 33–​395 and wrote (in Latin) a history of Rome from ad 96 in thirty-​one books. These and shorter passages in other texts, sometimes as short as a phrase, but derived from nearly 100 different authors, together with a substantial amount of geographical and place-​name information, have been studied for generations past. As noted, this material offers military and political history between ad 43 and the governorship of Agricola (ad 77–​85), slanted in his favour because of the surviving writing by his son-​in-​law Tacitus, including parts of his biography, the Agricola. Relatively little survives covering later events and nothing of any substance about, or resulting from, cultural life in Britain: no Latin works of British authors survive until the fifth-​century writings of Pelagius and St Patrick and the probably sixth-​century work of the monk Gildas (Dumville 1984). In the classical authors there are a handful of clichés (topoi) about Britain being across the ocean with backward peoples who lived off meat and milk and painted themselves blue in battle (as Caesar, Gallic War, V. 14), and, in Tacitus, some rhetorical swipes about Roman power, of which enslavement through comfort (Agricola 11) is perhaps one of the best known; there is also some information from Ptolemy’s Geography of the second century ad. As already commented, these meagre texts were supplemented by epigraphy and archaeology to make up ‘histories’ of Roman Britain, and a political history for the pre-​conquest period was also reconstructed from inscribed British coins. Military and political history will always be a part of the study of Roman Britain, because this is one way—​even if, emphatically, not the only one—​to place this province within a grand narrative of the Roman Empire. Textual and archaeological evidence will be combined within this, but this is often not straightforward. The directness of the tie-​up between the mention by Tacitus, in Annals 14, 38, of Classicianus, procurator of Roman Britain at the time of the Boudiccan uprising, and the tombstone found in London (Figure 5.5: cf. Grasby and Tomlin 2002), is exceptional.

The Textual and Archaeological Evidence    111 Combining textual and archaeological evidence requires a high command of both historical and archaeological thinking as well as data: its challenge tends to be underrated, because most of us lean towards one or other of these, and there is a tendency to oversimplify the methodology of the less familiar discipline. The problem is celebrated in obiter dicta, such as ‘historians who shop in the supermarket of archaeology’ (Pucci 1983: 106). In Roman Britain, because of the relative survival of evidence, it has been mostly archaeologists shopping in the supermarket of surviving texts. The flaw seems to be a certain wished-​for historical neatness, of tying archaeological evidence into randomly surviving documentation, with which it usually has no more than an indirect relationship. It is all too easy to find examples of misguided attempts at this, and it would be dispiriting to dwell too much on them, but we should nevertheless draw the lesson from one episode of military archaeology in early Roman Britain, of how it led to distortion of both the archaeological and documentary information and in the medium term harmed the subject. In the 1960s and 1970s there was an attempt to reconstruct the military activities of individual governors in the early years of Roman Britain by assigning to their term of office particular forts or fortresses (as Webster 1970). The dating and location of such sites were compatible with the brief documentation in Tacitus. However, archaeology was uncovering mostly semi-​permanent winter quarters, as opposed to marching camps, which did not reflect every twist and turn in campaigning; their dating was from finds of coins and pottery rather than from inscriptions, so, at best, fluid to a decade or so; finds from periods of maximum use were usually not abundant, and fewer than those from the moment of disuse; and overviews of dispositions had to assume a more-​or-​less textbook composition of units, whereas the reality was soon discovered to be more untidy. To be brutal, this choice of research target was unattainable, it led to inappropriate interpretations of evidence, and it diverted attention away from questions on which archaeological evidence could have thrown light—​as, for example, the interrelationships between military sites, the degree to which they were integrated locally, and so on. It also encouraged the discussion of non-​existent meanings in the text of Tacitus, as, for example, whether his rhetorical flourish about the Silures being ‘castrisque legionum premenda’ (‘suppressed [or oppressed] by legionary camps’) (Annals 12. 32), meant one or more forts or fortresses (cf. Hurst 1985: 121, for the details). For a while after the 1970s, for reasons partly of contemporary culture, but also because of disappointment engendered by the approach described, there was a reaction against research on the Roman army in Britain (cf. James 2002). Perhaps ironically, the subject was rescued by the discovery of new texts, for the Vindolanda documents forced us to take a more complex, and more rewarding, view of the Roman army in Britain and a more realistic view of the relationship between textual and archaeological evidence. A well-​known comment, made against the general background described, in a symposium on the forts of the Saxon Shore, was Richard Reece’s: ‘We must ask historical questions of historical sources and archaeological questions of archaeological sources, and never get the two mixed up’ (quoted by Johnston 1977: p. v).1 That is perfect advice at the basic methodological level when studying texts and/​or material remains, but, as I hope has been demonstrated, it is insufficient as a wider statement of the relationship

112

112   Henry Hurst between history and archaeology. Moreland’s argument that texts are also artefacts should be remembered, and it should be obvious that, for example, the Vindolanda tablets, the Bath and Uley defixiones, and the Classicianus tombstone belong to both fields of study. In any case, the aims of ‘archaeology’ and ‘history’ for Roman Britain are not fundamentally different. Archaeology might be said to explore human agency as well as wider socio-​economic or technological processes in order to understand what happened. So also does document-​based history, except that more names tend to be attached. That does not mean that there are not a host of special studies, where the two disciplines can and should be kept apart, but the results are ultimately to be synthesized. That will be fruitful as long as the conceptual frameworks and disciplinary requirements of both fields of study are respected.

Conclusion The whole field of textual and archaeological evidence for Roman Britain has seen a transformation, both of available material and of approaches over the last three decades. The texts most studied used to be those written at a distance about Britain, with—​ for the most part—​only formal inscriptions on stone from within the province, and they led to a Rome-​based, top-​down view, against which there were sharp, and varyingly focused reactions. Now there is a greater body of writing from within Roman Britain than about it, nearly all of it well published, and this writing is representative of many sectors of society from the imperial administrators and army to persons made desperate by the loss of a cloak. It is a rich assemblage, and it challenges us to think actively about how writing was used and to face the complexities of life in the province with, perhaps, fewer preconceptions than have been apparent in some earlier studies. As regards the synthesis of archaeology and history, the archaeology of Roman Britain is, inescapably, the study of people in whose lives writing was a major presence, even if most could not read; they are also a people and period often written about (by modern historians) within two broad frameworks, of the history of Britain and of the Roman Empire. It will be possible to advance on the archaeological front by developing better thinking, with the written evidence being given its due weight; and, on the historical front, by adjusting the type of narrative written about Roman Britain to the range of evidence that is now available.

Acknowledgements Professor Martin Millett and Dr Roger Tomlin kindly read earlier drafts and made suggestions for improvement; mistakes, however, remain my own. Thanks are also offered to those who helped in providing images: Chris Sutherns of the British Museum images

The Textual and Archaeological Evidence    113 for Figure 5.1 and the photograph in Figure 5.5; Dr Tomlin for Figures 5.2 and 5.3; Professor Millett for Figure 5.4; the staff of Alamy for Figure 5.6.

Abbreviations CIL  Corpus Inscriptionum Latinarum RIB  The Roman Inscriptions of Britain

Note 1. Professor Martin Millett kindly located the concealed setting of this well-​known remark.

References Barrett, A. A. (1978). ‘Knowledge of the Literary Classics in Roman Britain’, Britannia, 9: 307–​313. Beard, M. (1991). ‘Ancient Literacy and the Function of the Written Word in Roman Religion’, in J. H. Humphrey (ed.), Literacy in the Roman World. JRA Supplementary Series 3. Ann Arbor: Journal of Roman Archaeology, 35–​58. Biddulph, E. (2013). ‘The Blind Potter: The Evolution of Samian Ware and its Imitations’, in M. Fulford and E. Durham (eds), Seeing Red: New Economic and Social Perspectives on Terra Sigillata. London: Institute of Classical Studies, 368–​380. Bowman, A. K. (2003). Life and Letters on the Roman Frontier: Vindolanda and its People. 2nd edn. London: British Museum Press. Bowman, A. K., and Thomas, J. D. (1983). Vindolanda: The Latin Writing-​Tablets. Britannia Monograph Series 4. London: Society for the Promotion of Roman Studies. Bowman, A. K., and Thomas, J. D. (1994). The Vindolanda Writing-​ Tablets (Tabulae Vindolandenses). Volumes 2 and 3. London: British Museum Press. Bowman, A. K., and Woolf, G. (1994) (eds). Literacy and Power in the Ancient World. Cambridge: Cambridge University Press. Bowman, A. K., Thomas, J. D., and Tomlin, R. S. O. (2010). ‘The Vindolanda Writing-​Tablets (Tabulae Vindolandenses IV, Part 1)’, Britannia, 41: 187–​224. Bowman, A. K., Thomas, J. D., and Tomlin, R. S. O. (2011). ‘The Vindolanda Writing-​Tablets (Tabulae Vindolandenses IV, Part 2)’, Britannia, 42: 113–​144. Brilliant, R. (1984). Visual Narratives:  Story-​Telling in Etruscan and Roman Art. Ithaca, NY: Cornell University Press. Collingwood, R. G., and Wright, R. P. (1965). The Roman Inscriptions of Britain. Volume I: Inscriptions on Stone. Oxford: Clarendon Press. Collingwood, R. G., and Wright, R. P. (1991–​5). The Roman Inscriptions of Britain, Volume II: Instrumentum Domesticum. Fascicule 1: Gloucester: Alan Sutton. Fascicules 2 (1991), 3 (1991), 4 (1992), 5 (1993), 6 (1994), 7 (1995), 8 (1995), indexes (1995): Stroud: Alan Sutton for the Administrators of the Haverfield Bequest. Dumville, D. (1984). ‘The Chronology of De Excidio Britanniae, Book I’, in M. Lapidge and D. Dumville (eds), Gildas: New Approaches. Woodbridge: Boydell Press, 61–​84.

114

114   Henry Hurst Evans, J. (1987). ‘Graffiti and the Evidence of Literacy and Pottery Use in Roman Britain’, Archaeological Journal, 144: 191–​204. Frere, S. S. (1987). Britannia: A History of Roman Britain. 3rd edn. London: Routledge and Kegan Paul. Gager, J. G. (1992). Curse Tablets and Binding Spells from the Ancient World. New York and Oxford: Oxford University Press. Gardner, A. (2007). An Archaeology of Identity: Soldiers and Society in Late Roman Britain. Walnut Creek, CA: Left Coast Press. Goodburn, R., and Waugh, H. (1983). The Roman Inscriptions of Britain [edited] by R. G. Collingwood and R. P. Wright. I: Inscriptions on Stone: Epigraphic Indexes. Gloucester: Alan Sutton. Goody, J. (1977). The Domestication of the Savage Mind. Cambridge:  Cambridge University Press. Gordon, A. E. (1983). Illustrated Introduction to Latin Epigraphy. Berkeley and Los Angeles, and London: University of California Press. Grasby, R. D. (2012). Processes in the Making of Roman Inscriptions. Study 7: RIB 12. Oxford: The Centre for the Study of Ancient Documents. Grasby, R. D., and Tomlin, R. S.  O. (2002). ‘The Sepulchral Monument of the Procurator C. Julius Classicianus’, Britannia, 33: 43–​75. Hanson, W. S., and Conolly, R. (2002). ‘Language and Literacy in Roman Britain:  Some Archaeological Considerations’, in A. E. Cooley (ed.), Becoming Roman, Writing Latin? Literacy and Epigraphy in the Roman West. JRA Supplementary Series 48. Portsmouth, RI: Journal of Roman Archaeology, 151–​164. Harris, W. V. (1989). Ancient Literacy. Cambridge, MA: Harvard University Press. Hartley, B. R. (1969). ‘Samian Ware or Terra Sigillata’, in R. G. Collingwood and I. A. Richmond, The Archaeology of Roman Britain. London: Methuen, 235-​251. Hartley, B. R., and Dickinson, B. M. (2008–​12). Names on Terra Sigillata:  An Index of Makers’ Stamps and Signatures on Gallo-​Roman Terra Sigillata (Samian ware). 9  vols. London: Institute of Classical Studies. Hassall, M. W.  C., and Tomlin, R. S.  O. (1989). ‘Roman Britain in 1988. II. Inscriptions’, Britannia, 20: 327–​345. Henig, M. (1997). ‘The Lullingstone Mosaic: Art, Religion and Letters in a Fourth-​Century Villa’, Mosaic, 24: 4–​7. Hingley, R. (2000). Roman Officers and English Gentlemen: The Imperial Origins of Roman Archaeology. London: Routledge. Hingley, R. (2005). Globalizing Roman Culture:  Unity, Diversity and Empire. London: Routledge. Hull, M. R. (1963). The Roman Potters’ Kilns of Colchester. Reports of the Research Committee of the Society of Antiquaries of London, XXI. Oxford: Oxford University Press. Hurst, H. (1985). Kingsholm. Cambridge: Gloucester Archaeological Publications. Ireland, S. (2008). Roman Britain: A Sourcebook. 3rd edn. London: Routledge. James, S. (2002). ‘Writing the Legions: The Past, Present and Future of Roman Military Studies in Britain’, Archaeological Journal, 159: 1–​58. Johnston, D. E. (1977) (ed.). The Saxon Shore. CBA Research Report 18. London: Council for British Archaeology. Kenyon, R. F. E. (1993). ‘The Copying of Bronze Coins of Claudius I in Roman Britain’. Unpublished Ph.D. thesis, Institute of Archaeology, University College London.

The Textual and Archaeological Evidence    115 Ling, R. (2007). ‘Inscriptions on Romano-​British Mosaics and Wall-​Paintings’, Britannia, 38: 63–​91. Mairs, R. (2012). ‘“Interpreting” at Vindolanda: Commercial and Linguistic Mediation in the Roman Army’, Britannia, 43: 17–​28. Mann, J. C., and Penman, R. G. (1996) (eds). Literary Sources for Roman Britain. 3rd edn. London: London Association of Classical Teachers. Martin, M. (2010). Sois maudit! Malédictions et envoütements dans l’antiquité. Paris: Errance. Mattingly, D. J. (1997) (ed.). Dialogues in Roman Imperialism: Power, Discourse and Discrepant Experience in the Roman Empire. Journal of Roman Archaeology Supplementary Series 23. Portsmouth, RI: Journal of Roman Archaeology. Mattingly, D. (2006). An Imperial Possession:  Britain in the Roman Empire, 54 bc–​ad 409. London: Allen Lane (Penguin Books). McKitterick, R. (1990) (ed.). The Uses of Literacy in Early Medieval Europe. Cambridge: Cambridge University Press. Millett, M. (1990). The Romanization of Britain:  An Essay in Archaeological Interpretation. Cambridge: Cambridge University Press. Millett, M. (2005). Roman Britain. London: Batsford. Moreland, J. (2001). Archaeology and Text. London: Duckworth. Mullen, A. (2007). ‘Evidence for Written Celtic from Roman Britain: A Linguistic Analysis of Tabellae Sulis 14 and 18’, Studia Celtica, 41: 31–​45. Mullen, A. (2013). ‘The Language of the Potteries:  Communication in the Production and Trade of Gallo-​Roman terra sigillata’, in M. Fulford and E. Durham (eds), Seeing Red: New Economic and Social Perspectives on Terra sigillata. London: Institute of Classical Studies, 97–​110. Neal, D. S., and Cosh, S. R. (2009). Roman Mosaics of Britain. Volume III: South-​East Britain, Part 2. London: Society of Antiquaries. Pearce, J. (2004). ‘Archaeology, Writing Tablets and Literacy in Roman Britain’, Gallia, 61: 43–​51. Perring, D. (2003). ‘Gnosticism in Fourth-​ Century Britain:  The Frampton Mosaics Reconsidered’, Britannia, 34: 97–​127. Pucci, G. (1983). ‘Pottery and Trade in the Roman Period’, in P. Garnsey, K. Hopkins, and C. R. Whittaker (eds), Trade in the Ancient Economy. London: Chatto and Windus, 105–​117. Salway, P. (1981). Roman Britain:  The Oxford History of England, Volume 1A. Oxford: Clarendon Press. Scott, S. (2000). Art and Society in Fourth-​ Century Britain:  Villa Mosaics in Context. Oxford: Oxford University School of Archaeology. Street, B. V. (1984). Literacy in Theory and Practice. New York: Cambridge University Press. Strong, D. (1968). ‘The Monument’, in B. W. Cunliffe (ed.), Fifth Report on the Excavations of the Roman Fort at Richborough, Kent. Reports of the Research Committee of the Society of Antiquaries 23. Oxford: University Press for the Society of Antiquaries, 4–​73. Tite, P. L. (2010). ‘“Reading” and “Re-​Reading” the Frampton Mosaics: Religious Innovation and the Construction of Cultural Identity in Roman Britain’, Bolletino di Archeologia Online. Volume Speciale E/​E10/​4: 41–​55. Tomlin, R. S. O. (1983). ‘Non Coritani sed Corieltauvi’, Antiquarians Journal, 63: 353–​355. Tomlin, R. (1988). ‘The Curse Tablets’, in B. Cunliffe (ed.), The Temple of Sulis Minerva at Bath. Volume 2: The Finds from the Sacred Spring. Oxford Committee for Archaeology Monograph 16. Oxford:  Oxford Committee for Archaeology, 59–​277. [Also published separately as Tabellae Sulis: Roman Inscribed Tablets of Tin and Lead from the Sacred Spring at Bath.]

116

116   Henry Hurst Tomlin, R. S.  O. (1993). ‘The Inscribed Lead Tablets:  An Interim Report’, in A. Woodward and P. Leach (eds), The Uley Shrines: Excavation of a Ritual Complex on West Hill, Uley, Gloucestershire, 1977–​9. London: English Heritage, 113–​130. Tomlin, R. S.  O. (1996). ‘A Five-​Acre Wood in Roman Kent’, in J. Bird, M. Hassall, and H. Sheldon (eds), Interpreting Roman London:  Papers in Memory of Hugh Chapman. Oxford: Oxbow Books, 209–​216. Tomlin, R. S. O. (1998). ‘Roman Manuscripts from Carlisle: The Ink-​Written Tablets’, Britannia, 29: 31–​84. Tomlin, R. S. O. (2002). ‘Writing to the Gods in Britain’, in A. E. Cooley (ed.), Becoming Roman, Writing Latin? Literacy and Epigraphy in the Roman West. JRA Supplementary Series 48. Portsmouth, RI: Journal of Roman Archaeology, 165–​179. Tomlin, R. S. O. (2003). ‘“The Girl in Question”: A New Text from Roman London’, Britannia, 34: 41–​51. Tomlin, R. S. O. (2011). ‘Stylus Writing Tablets’, in J. Hill and P. Rowsome (ed.), Roman London and the Walbrook Stream Crossing: Excavations at 1, Poultry and Vicinity. MOLA Monograph 37. London: Museum of London Archaeology, 514–​517. Tomlin, R. S.  O., and Hassall, M. W.  C. (2001). ‘Roman Britain in 2000. II. Inscriptions’, Britannia, 32: 387–​400. Tomlin, R. S. O., Wright, R. P., and Hassall, M. W. C. (2009). The Roman Inscriptions of Britain. Volume 3: Inscriptions on Stone Found or Notified between 1 January 1955 and 31 December 2006. Oxford: Oxbow Books. Turner, E. G. (1956). ‘A Roman Writing-​Tablet from Somerset’, Journal of Roman Studies, 46: 115–​118. Turner, E. G., and Skutsch, O. (1960). ‘A Roman Writing-​Tablet from London’, Journal of Roman Studies, 50: 108–​111. Tyers, P. (1996). Roman Pottery in Britain. London: Batsford. Unesco (2006). ‘ Understandings of Literacy’, in Education for All. Unesco Global Monitoring Report 2006, ch. 6. Wacher, J. S. (1998). Roman Britain. 2nd edn. Stroud: Sutton. Webster, G. (1970). ‘The Military Situations in Britain between ad 43 and 71’, Britannia, 1: 179–​199. Welles, C. B., Fink, R. O., and Gilliam, J. F. (1959). The Parchments and Papyri: The Excavations at Dura-​Europos. Final Report 5, Part 1. New Haven: Yale University Press. Williams, J. H. C. (2002). ‘Pottery Stamps, Coin Designs and Writing in Late Iron Age Britain’, in A. E. Cooley (ed.), Becoming Roman, Writing Latin? Literacy and Epigraphy in the Roman West. JRA Supplementary Series 48. Portsmouth, RI:  Journal of Roman Archaeology, 135–​150. Woolf, G. (1996). ‘Monumental Writing and the Expansion of Roman Society in the Early Empire’, Journal of Roman Studies, 86: 22–​39. Woolf, G. (1998). Becoming Roman: The Origins of Provincial Civilization in Gaul. Cambridge: Cambridge University Press.

Chapter 6

The Early Rom a n H ori z on Lacey Wallace

Introduction The early Roman horizon in Britain refers both to archaeological deposits from a particular time period (in the south-​east of Britain, it corresponds roughly to the generation who lived during and after the Claudian invasion in ad 43) as well as to the material remains of an emerging Romano-​British culture. To question what happened to the lives of the indigenous population in Britain when their land was incorporated into the Roman Empire requires not only the scrutiny of this evidence to investigate how people survived, adapted, and resisted in a period of rapid change, but also an elusive chronological specificity. Isolating the ‘early Roman’ strata for examination, exploring their relationship to the late pre-​Roman Iron Age (LPRIA), and drawing meaning and significance from the study of artefacts, features, and deposits is a complex feat, and one that can be touched upon only briefly here (see also Creighton 2006). Bearing this caveat in mind, this chapter will explore a few of the problems and biases encountered by archaeologists studying the ‘early Roman horizon’ and examine the nature of the surviving archaeological evidence for change and continuity, particularly within the early urban centres destroyed, and thereby preserved, in the Boudiccan Revolt of ad 60/61.

Problems with the Evidence and Overcoming our Biases One of the main obstacles one encounters when exploring the ‘early Roman horizon’ is the artificial distinction of the LPRIA and the Roman period, divided by the Claudian invasion of ad 43. The temptation to put weighty emphasis on ad 43 ought to be avoided, as few sites bear evidence with which an archaeologist could identify changes resulting from the invasion.

118

118   Lacey Wallace The academic separation of pre-​Roman and Roman archaeological remains stems, simply put, from the classical education prevalent in the modern period, which led to British archaeology of the Roman period being interpreted within the literary and historical framework of classical studies (See also chapters by Millett and Wilson, this volume). Traditional syntheses of Roman Britain often use the histories (e.g. Tacitus’ Annals and Agricola, Caesar’s Gallic Wars, and Cassius Dio’s Histories) to create a narrative for the archaeological evidence, which does a disservice to both disciplines by oversimplifying the relationship between them (see Creighton 2001). Archaeological investigations can include analysis of ancient texts as part of the research arsenal, but ought not to use this extremely limited and incomplete body of evidence to create an analytical framework. The LPRIA archaeological record cannot be ‘matched up’ to military incursions and administrative changes, such as the invasions of Caesar and possible installation of loyal vassal kings (Creighton 2000). Similarly, the movements of troops described by ancient historians ought not to be linked directly to the building of roads and founding of towns; a strong bond ties those scholars who have emphasized the historical accounts to those who would see the Roman military as the mechanism by which social and economic change occurred in Britain. An academic tradition of linking archaeological remains to events described by ancient authors—​most of whom had never travelled to Britain and who wrote with their own agendas and intended audiences—​is entrenched in our field to this day. This connection has led to assumptions and outmoded theories, which can take decades to unseat and must be challenged and confronted with rigorous studies of evidence. We can add to these pitfalls the regional diversity of the effects of contact with the Roman Empire: it is a certainty that the ‘early Roman period’ was experienced differently in many parts of Britain as ideas, objects, and people moved through the new province. The experience of people living in east Kent, for example, who had been in direct contact with the Roman Empire for at least a century before the Claudian invasion, would have differed greatly from people living in locations distant from the trade routes bringing continental goods to Britain. No two sites exhibit the same ‘early Roman horizon’, and generalizing would result, at best, in a vague account. If we can agree to call the ‘early Roman period’ in the south-​east of Britain that which it is possible to date to the decades immediately following the Claudian invasion, we will immediately see that it was experienced drastically differently in the countryside from the way it was in the oppida, emerging urban centres, and military installations. Finally, one must also consider the advantages and issues of studying a particular time period as brief as the ‘early Roman horizon’, as well as those types of evidence that can offer the possibility of assigning dates. Although this chapter avoids the vagaries of archaeological dating, including the sometimes-​circular logic for the assignation of artefact dates—​for example, historical events used to identity strata, which contain object types that are then used to link strata elsewhere with historical dates, eventually creating a corpus of artefacts assigned a particular date range—​a healthy scepticism of the precision of such techniques ought to be maintained.

The Early Roman Horizon    119

Overcoming these Issues in Londinium Few sites offer robust evidence to isolate archaeological strata from the early Roman period; Londinium, however, is one such site. Because there was no significant LPRIA occupation in the area that would become the Roman town, the issue of identifying the early Roman transition is avoidable. Overcoming the historical narrative remains a significant obstacle to interpreting early Londinium, but it is one that can be explored with a huge quantity and variety of evidence. Being placed in the south-​east at the centre of transport, trade, and communication routes both over land and by sea, London can be seen as an ideal site for exploring the urban experiences of the generation who lived through the Claudian invasion. Moreover, the ad 60/​61 Boudiccan Revolt burnt deposits provide a clear archaeological cut-​off point that can be used to isolate the first decade (or so) of occupation there. In London, as elsewhere, scholars have made hypotheses (and assumptions) about the early Roman levels based on existing historical, toponym, and anecdotal evidence before many sites had been excavated. By the 1950s, the notion of a pre-​Roman London was out of favour, and a major compilation of the evidence (Merrifield 1965) put archaeology at the forefront of future investigations, although there were few sites then known representing the earliest levels of Roman London. After this time, the academic context was characterized by the popularity of administration-​driven urbanization following military installations (the fort-​to-​town hypothesis—​see Webster 1966 and Frere 1967). So it was suggested by many that a camp, fort, or supply base provided the location, impetus, and form for the earliest urban settlement, linked to the military movements of Plautius as described by Cassius Dio (a third-​century historian who wrote in Greek and had never been to Britain). Hanging the interpretations on the historical evidence, many thought that, because Plautius crossed the Thames somewhere, it was most likely that he crossed at London and set up a camp for his troops while they waited for the arrival of Claudius. The ancient literature was not often subject to critical evaluation, but rather taken to be a relatively unbiased record of events, as understood by their authors. The ubiquitous use of the word ‘conquest’, until relatively recently, employed to describe the result of Claudius’ efforts in ad 43, is a reflection of how scholars have allowed their interpretations to be skewed by the propaganda of the time. The effort to ‘match up’ the events in the surviving accounts—​for example, to link the construction of roads to the movements of troops described—​led to a situation where archaeologists decided what they were going to find before they dug it up. In London, historians and archaeologists expected to find the military installation of regular streets crossing at a principia (underlying the later forum) with rows of barracks and store buildings, surrounded by defensive ditches, all thought to have given form to the later town. The effect of London and its bridge was seen to have been caused by the presence of a Claudian base of some kind.

120

120   Lacey Wallace One example demonstrates how otherwise-​ambiguous evidence could be forced into the narrative derived from the scraps of ancient histories. In 1968–​9, an excavation north of Fenchurch Street revealed a small amount of occupation debris, burnt sediment, a copy of a Claudian coin, and some coarseware pottery (Philp et al. 1977: 7). No contemporaneous structures were found. The interpretation of this meagre evidence was described thus: All of the material recovered from these primary deposits is of mid-​first-​century date and thus in accord with the long-​established view that permanent settlement began in London in 43, at the time of the Roman conquest. The probability is that this primary occupation was of a purely military character, perhaps reflecting either a temporary camp on this site or activity by troops living nearby. (Philp et al. 1977: 7)

There is no way that this confident interpretation, slotted uncomfortably into a historical paradigm, would have been made without the ancient texts; it is entirely dependent on them and acts as an illustration of what is already ‘known’, not as a search for new evidence. Certainly there are examples where towns founded in the 40s and 50s occupy earlier military settlements (for example, the veteran settlement at Colchester), but the assumption that urbanization could not be a force driven by civilians robs the vast majority of the population of their agency in forming the landscape of the new province. At the time, only a handful of sites with possible early Roman levels had been found, and most of these were merely observed as they were destroyed by construction, so hypotheses could not be formed by placing evidence within the context of wider patterns. Over the next four decades, another ninety sites with features pre-​dating the Boudiccan fire would come to light. These 118 sites have produced more than 14,000 recorded archaeological contexts (although many more have been destroyed, lost, or never recorded) (see Wallace 2014). We now have sufficient evidence to begin to discern patterns in the early Roman levels and to produce interpretations that are separate from the historical narrative. Where a site has not been investigated, allowing assumptions and the written histories to compensate for lack of evidence may be valid. In the case of Londinium, however, some (e.g. Perring 2011) have allowed their passion for the historical narrative to overcome the robust data, which are among the finest and most complete of any Roman town in the empire.

Nature of Change One of the best sources of evidence for the ‘early Roman horizon’ comes from urban contexts, like Londinium. Two other towns in the south-​east that were occupied in the decade following the Claudian invasion can also be compared: Colchester (Camulodunum/​ Colonia Claudia Victricensis) and St Alban’s (Verulamium). Each of these three also offers a unique opportunity to investigate a distinct early Roman time period because

The Early Roman Horizon    121 of the fate these settlements suffered during the Boudiccan Revolt of ad 60/​61. The destruction of these towns during the Boudiccan Revolt and subsequent rebuilding has sealed the archaeological deposits within and immediately beneath the fire debris associated with the sacking of these towns, allowing us to discuss a cohesive and comparable period relating to the experiences of the generation who lived through the invasion. Although their foundations have been linked to the pacification of the south-​east and the movement of the army north and west during the 40s ad (e.g. Frere 1972) and although affected by the revolt at approximately the same (archaeologically indistinguishable) time, the reasons for the foundation of these early urban sites up to ad 60/​61 were driven by different groups in dramatically different ways. Colchester, founded by the Roman provincial administration as a veteran colony (Colonia Claudia Victricensis) in ad 49 atop the earlier legionary fortress, was sited in an area that had not previously been used for settlement but that lay within a larger oppidum (Camulodunon) area delineated by major dykes and occupied by scattered ‘hamlets’ as well as high-​status enclosures and agricultural/​livestock areas (Gascoyne and Radford 2013). In contrast, a strong ruling elite and a group of people (tribe?) who had occupied a relatively large and well-​populated oppidum area (Verulamion) in the Ver river valley for some time then chose to begin to construct an urban settlement (Verulamium) in a symbolic and significant location where some people would live and work in much closer proximity to one another than they ever had before (Niblett 2001; Niblett and Thompson 2005). The earliest phases of development at Londinium also suggest a different agenda. Here, an implanted population of indigenous Britons and continental merchants and craftsmen not local to that place came to build a ‘town’ on an unoccupied site, and organized labour and communal development, interpreting what it meant to live in a town in different ways (Wallace 2014).

Verulamium That Verulamium was already established and linked into the provincial road networks prior to the Boudiccan Revolt is attested to, for example, by the burnt remains of timber workshops in Insula XIV (Frere 1972: 19), which overlay the road onto which they faced and also overlay evidence of Claudian-​period occupation. Frere, therefore, made the link between the ‘foundation’ of Verulamium as an urban centre (which he erroneously believed to overlay an earlier fort) and the creation of the colonia at Colchester as both being part of an imperial policy of urbanization. In the pre-​Boudiccan period, the nascent town of Verulamium lay at the centre of the later urban core, but was a small area (less than 10 hectares) enclosed by ditches and banks. A quarter of its area was taken up by the pre-​Claudian ‘Central Enclosure’, which pre-​dated the town and was a location where blanks for British coins were made (interpreted through the presence of coin moulds) and may have been a site of high-​status occupation perhaps constructed and built by the same high-​status family/​individuals as the ‘proto-​villa’ at Gorhambury (see Neal 1990) and the ceremonial site at Folly

122

122   Lacey Wallace Lane (see Niblett 1999). The ‘Central Enclosure’, although the course and nature of its ditches are imperfectly known, formed an area of c. 2.4 hectares underneath the later forum/​basilica complex (Niblett 2005). This feature, along with a 6-​metre-​wide bank (possibly revetted with timbers), a pile-​supported timber causeway, and a major north-​ east–​south-​west ditch, pre-​dated the Flavian grid of streets and formed the planning of the early ‘town’. The sub-​rectangular area formed by these ditches contained within it the crossing of Watling Street (prior to its rerouting) and the main south-​west–​north-​ east road focused on the high-​status Folly Lane funerary/​ceremonial enclosure atop the ridge across the stream. The four known roads within the area were lightly metalled, and, although they were at roughly right angles, they did not enclose insulae in the way of later roads and did not extend beyond the ditched area.

Colonia Claudia In Colchester, the fort (which would form the basis for the colonia) was built within a larger oppidum area delineated by major dykes. The relationship between the ‘town’ at Colchester and the indigenous local group was, therefore, entirely different from that at Verulamium. The area of the fort and additional areas constructed when it was transformed into a colonia covered c. 36 hectares, forming a grid-​planned area far larger than that of Londinium. We expect that those allocated land in the colonia were veterans, but their families and households could well have been of indigenous British descent. Labelling Colonia Claudia as a ‘military’ community is therefore difficult, although it is likely that those British people who chose to live and work in the town were considered ‘sympathizers’ by traditionalists like Boudica and her army of confederated tribes.

Londinium The earliest date for the construction of Watling Street through Londinium is ad 47/​48, derived from the felling date of a tree used for its timber to line a drain underneath the road (Tyers 2011). That date would be consistent with the beginning of development in the town c. ad 49 (see also Wallace 2014). In Londinium, the core Cornhill settlement was delineated with a grid of streets covering c. 18 hectares, containing six insulae, a gravelled ‘proto-​forum’ area, as well as relatively ephemeral waterfront/​port structures, all of which were constructed slowly over the first decade (see Wallace 2014). The other areas of the settlement were not ‘planned’ in this way, but rather formed in a linear development along the roadsides outside of the centre; this other occupation covered a further 22 hectares and may have actually housed more people than the central part of the town. The settlement was, therefore, divided into three areas separated by streams: west of the Walbrook stream lay Ludgate Hill, where indigenous families lived, probably engaging in scavenging and reclamation to survive in the early days, which rapidly became a densely occupied mercantile

The Early Roman Horizon    123 and production area focused on the London–​Verulamium road; east of the Walbrook on Cornhill lay the planned core, probably created and inhabited by immigrant citizens and merchants; south of the Thames, the London–​Canterbury stretch of Watling Street crossing two low-​lying eyots was lined with smaller houses, shops, and workshops inhabited by those immigrants who were perhaps slightly later in coming to Londinium, of lower status than those on Cornhill, and probably culturally distinct from the population on Ludgate Hill (Wallace 2014). Analysis of these data shows that the town was not just founded, planned and expanding in a simple pattern of growth, even in this short time period. The processes and decisions that affected its earliest years were mutable and piecemeal. For many different reasons (availability of materials, need to build rapidly, rejection or devaluation of certain cultural values), nearly all of the early buildings in Londinium were constructed of ephemeral materials. Buildings with posts and timbers in direct contact with the ground would be susceptible to rot and would have to be repaired and replaced at intervals. Even in the short period before the Boudiccan Revolt fire, many buildings were torn down and replaced. Equally, the planning and construction of roads were not even across the town—​buildings were destroyed to make way for roads in some places, while existing roads were blocked by new buildings in others. This dynamic and changing urban landscape makes identifying broad patterns a complex endeavour.

‘Things Called Towns’ The most significant change in settlement patterns at these three early urban centres is, of course, that a condensed area was chosen in which to construct houses, workshops, and public spaces. Although these new ‘towns’ (see Reece 1988) did not replace the settlements surrounding Verulamium and Colonia Claudia, their existence suggests that people did alter their lifestyles in either or both work and domestic activities. At the outset of ‘planning’ of these towns, a significantly smaller area was selected than would eventually be occupied at the height of their expansion.

Lack of Public Buildings The small scale of each of these towns is not just in their restricted area, but also in the general lack of public buildings in each. Neither Londinium nor Verulamium had a monumental forum complex until the Flavian period (Verulamium’s is dated by a dedicatory inscription of ad 79 (see Frere 1983: 55, 69); and London’s, less specifically, was constructed c. ad 80 (see Marsden 1987: 22)), despite the fact that fora arguably formed the heart of any ‘Roman’ city and made it navigable and recognizable to those familiar with urban space on the continent. The space where the forum would later be built in each was a reserved central location in the pre-​Boudiccan period. In Verulamium,

124

124   Lacey Wallace the forum overlay the LPRIA ‘Central Enclosure’; in Londinium, the gravelled ‘proto-​ forum’ and its large storage/​distribution/​administrative buildings may have served a very different function from the later structures in the same area, but clearly show that there was a power structure in place that could organize the construction of a large public space. During the construction of Colonia Claudia, the area of the military principia was possibly cleared to make way for the forum, but it is not proven. The Temple of the Divine Claudius in Colonia Claudia is, however, an example of a major early urban monument unparalleled elsewhere in Britain in this period—​and may itself have been placed at the head of the forum of Colonia Claudia. In Verulamium, the macellum was comprised of two wings of rooms (workshops? shops?) enclosing a hall or courtyard with a conduit supplying water and may be the only pre-​Boudiccan ‘public’ building from Verulamium, but its date, as well as its function, are unclear. A bathhouse comprising a mortar-​lined sunken ‘pool’ surrounded by limestone columns with evidence for painted wall plaster need not have been public, but may, rather, have been associated with elite occupation (that is, the so-​called proto-​forum within the ‘Central Enclosure’) (Niblett 2005; Niblett et al. 2006). In Londinium, the large buildings associated with the proto-​forum (one of which may have been a grain storage and distribution centre) and a masonry-​founded aisled hall across the main street represent the only probable public buildings, but are not of the form or function typical of a Roman town (Wallace 2014). None, however, had the full suite of public buildings that we traditionally associate with urban centres in Roman provinces on the continent (baths, theatre, forum, basilica, temples). The absence of such public buildings is surprising in the light of how we generally think of the development of planned ‘Roman’ towns. For such crucial elements of the urban landscape not to have been extant during the first decade (or so) of occupation in well-​dated urban centres in the south-​east—​well situated on the road and communication/​trade networks, with excellent access to continental goods and ideas and bearing the ‘trademark’ of at least a limited area of orthogonal streets (see, e.g., MacDonald 1986: 17)—​presents a stark contrast to the foundation stories of cities at the centre of the empire. The idea that a grid of streets only served to provide a shape to the urban space and was not connected to what we might consider ‘urban planning’ in a modern sense (Laurence et al. 2011: 116) certainly applies to at least Verulamium and Londinium, but underplays the significance of how people would have made the decision to build a grid and organize labourers and space, which would have been a significant and symbolic resolution integrally related to their dedication to urbanize and plan for a town.

Adoption of New Material Culture The early Roman levels of each of these settlements are characterized by a sharp increase in the adoption of imported goods and objects, which also suggest the adoption of new behaviours. In Londinium, the proportion of imported ceramics was at its height in the

The Early Roman Horizon    125 pre-​Boudiccan period (Davies et al. 1994; Wallace 2014), which demonstrates the weak ties the earliest population there had to local trade networks and craftsmen (This can be compared to Waldgirmes, a town probably constructed by the military for civilian use that was strongly situated within local trade networks (see Schnurbein 2003)). Coupled with the cultural package of writing equipment, glass vessels, coins, lamps, toilet equipment, and continental-​style adornment and body-​preparation objects, it is clear that the population either arrived in Londinium with the knowledge and habits common to other Roman territories on the continent or rapidly adapted to the use of new material culture. Certain of these objects—​lamps and lamp-​holders, for instance—​are found in far smaller numbers in pre-​Boudiccan Londinium than in Colonia Claudia (see Eckardt 2002), which may be related to the more widespread use of lamps by veterans than civilians. Indeed, the relative quantities of military equipment bear the same pattern as lamps in these two settlements (see Wallace 2014). Only a statistically sound methodology for comparing quantities of artefacts in the three towns (for example, by density of artefacts per volume of excavated sediment) could truly clarify the differences in the adoption of new forms of material culture among the different groups of people who chose to live in these early urban centres.

So Why Did People Choose to Urbanize? Why did indigenous British people choose to move to these new settlements, so unlike those they had lived in before? Why did people from the continent choose to come to such a new and unstable, unproven place as Britannia? Recent work by Creighton (2006), Britannia: The Creation of a Roman Province, explored the LPRIA-​to-​Roman transition in several towns and proposed a variety of motivations for urbanization and factors that affected the forms it took, such as memory, symbolic landscapes, and re-​ creation of the familiar. He provides an insightful and multifaceted view that draws directly on the evidence and presents robust hypotheses. Much of the evidence suggests that these very different people may have had similar motivations—​namely, opportunity for social and economic improvement. Diverse circumstances undoubtedly drove people to make the leap; some were probably dispossessed of land and status by war, some may have found that their familiar systems of exchange and production were so disrupted that they could not continue to survive without choosing to change, others sought out the military life of stable pay and excitement followed by the reward of land, while there were undoubtedly those who seized upon a new market and made wide-​ reaching arrangements for developing their hold in it. Change certainly was more rapid (and presumably also experienced as accelerated) in urban centres where trade, communication routes, and people came together intensely. It is clear that certain groups of people chose to adopt new forms of material culture and to take advantage of the opportunities provided by a closer relationship to the economy and society of the continent, while others (especially those living in rural areas) continued life in a relatively unaltered fashion. The elite families that had made use of imported

126

126   Lacey Wallace goods and continental-​style symbols (for example, wine, coinage, new burial rites, and so on) to construct their identities in the LPRIA may have been at the centre of driving change in some instances (for example, at Verulamium), while the early Roman administrators and military did, in one instance (Colonia Claudia), set up a group of veterans to live in a town, and in others it appears that ‘regular’ people were on board with the notion of moving up the social ladder by taking up a new place in a dynamic hierarchy, especially that afforded by urban centres (for example, at Londinium).

Recognizing Continuity The ‘early Roman’ period cannot be studied without understanding the preceding period, the LPRIA (see, for example, the classic works of Cunliffe 1974 and Haselgrove 1984)—​the artificial division can illogically dissociate the archaeological record of the two. The arrival of the Roman army and administrators did not put an end to kinship, kingship, and indigenous culture; the significance of ‘tribal’ groups in Britain and the systems of exchange, political alliance, and power in the LPRIA and early Roman period have been explored through coinage and extrapolation/​comparison with the histories of other provinces (Creighton 2001). Among those people who lived under the rule of a British ‘friendly king’, the changes associated with the Claudian invasion may have affected them little more than had previous treaties with the Roman Empire (for example, those with Caesar and Augustus) and may have seemed more in the way of another step in an already-​extant process, rather than the commencement of a new era. There are several examples of how assumptions based upon historical frameworks affect archaeological interpretations. Separating archaeological features into LPRIA and Roman-​period on the basis of morphology can be misleading and both devalue the agency of the indigenous groups as catalysts of change as well as mask possible continuities. For example, at Skeleton Green (Partridge 1981), an Iron Age and Roman roadside settlement compound in Hertfordshire, a stratum of alluvial sediment covered the post-​built structures of the LPRIA levels and contained material dating it to c. ad 30–​40; the subsequent rebuilding was of a timber-​framed construction type and featured right-​ angled roads. Although no artefacts were found in the post-​flood buildings and the dating evidence suggests that the structures from this phase went out of use before c. ad 50 (Partridge 1981: 49, 51–​52), the author still felt confident to put the early date of ad 43 on this phase and to consider it ‘Roman’: the relative neatness of the buildings and the regularity of the road-​system seems to imply the presence of more than just Natives. The few scraps of military metalwork may be significant, in that they point to a military presence in the early years following the conquest in A.D. 43. (Partridge 1981: 52)

The Early Roman Horizon    127 A similar dating interpretation was presented a decade earlier, in the assignment of ad 43 to the earliest structures at Fishbourne (Cunliffe 1974). At Fishbourne, however, the incongruity between the dating evidence (pottery from c. 10 bc–​ad 25) and the improbability that the earliest sturdy and substantial ‘military supply-​base’ granary was standing for just a few years before being razed and replaced was swept under the rug (see Manley et al. 2005; also Creighton 2006: 54–​58). Cunliffe’s novel interpretation of ‘supply bases’, at both Fishbourne and Richborough, fit into a historical framework that explained how the military organized the invasion and eventual control of Britain. Once this historical hypothesis had been decided upon, the dating problems were explained under the same rubric—​that is, that the pottery was from an old store that the Claudian-​era military had brought with them. If we follow the notion of Occam’s razor, this explanation appears inventive and certainly worth questioning. Most archaeologists would now agree that we cannot point to an archaeological feature of broadly mid-​first-​century-​ad date and assume it has anything to do with an event as specific as the Claudian invasion. Equally important to recognizing that major changes were already happening in the LPRIA is the acceptance that other changes that occurred after the Claudian invasion did not necessarily have a direct relationship to it (although, of course, it is true that some archaeological remains, especially rural sites with little datable material culture, are notoriously difficult to date with any precision). In rural areas especially, but also in suburban areas near to early Roman towns, continuity is characteristic of the early Roman period. Niblett (2001: 60), for example, noted ‘there is little evidence from the Verulamium area as a whole for disruption at the time of the conquest. Indeed, as far as Verulamium is concerned, were it not for references in classical texts, we might not realize that Claudius conquered the area at all.’ Such continuity, especially in close proximity to a municipium, demonstrates that the experience of empire in the early years after the Claudian invasion was vastly different and dependent on economic status, place in the social hierarchy, occupation, motivations, and needs. The most conservative human activities—​such as gastronomy and language—​might have taken at least a generation (that is, more than twenty years) to begin to change among certain groups, even where their access to and potential benefit from choosing to change was strong. The use (and loss/​disposal) of ‘typical’ Roman-​period objects, such as writing implements, coins, imported ceramics, glass vessels, lamps, brooches, horse-​ tack ornaments, toilet equipment, and so on, may not have formed part of the lives of ordinary farmers, shepherds, and rural craftsmen. People did not immediately abandon their homes and dash for the new towns or raze their round-house farmsteads to the ground after the Claudian invasion. Life continued much as it had previously for a huge proportion of the population. Buildings would not be replaced just to suit the fashion where surplus money was not rattling around, although when people did finally rebuild, they may well have chosen a type of building and techniques of construction that were fashionable—​and continental—​in origin.

128

128   Lacey Wallace

Exploring Continuity at Colchester Within the wider oppidum area of Camulodunum/​Colonia Claudia (meaning the c. 28-​ square-​mile Camulodunon area, not a condensed continental-​style, proto-​urban oppidum), LPRIA and early Roman-​period features can be difficult to differentiate where continuity in form exists. The overall picture at many sites across the oppidum is one of little change in the early Roman period; the scattered settlements continued to be inhabited in much the same way after the construction of the legionary fortress. There is dating evidence that demonstrates that some dykes (large ditches with U-​or V-​shaped profiles with the excavated material used to create a bank) and non-​rectilinear enclosures, which would traditionally be considered pre-​Roman based on their morphologies, were constructed across the ad 43 divide. It is worth noting the limitations of the dating evidence: ‘it is generally difficult to distinguish between pre-​and post-​conquest occupation because the late Iron Age grog-​tempered pottery continued in use into the Flavian period’ (Radford 2013: 45). Continuity, however, has been identified into the late first century in the system of dykes, at the Gosbecks farmstead and sanctuary, at the Sheepen industrial/​production settlement, and in the burial area at Stanway (Niblett 1985; Crummy et al. 2007). The material found in the dykes of Camulodunum has been difficult to date, and it is evident that they were built over a period of time; for example, of the two dykes with secure termini post quem, the Sheepen Dyke produced only two types of LPRIA ceramics and the assemblage as a whole dates to c. ad 5–​61 (Hawkes and Hull 1947) and the Gryme’s Dyke is thought to be post-​ad 43 (Hawkes and Crummy 1995: 115). The securely Roman-​period changes to the dyke system included the construction of a new one in the form of a triple ditch, while certain other dykes were filled in (Gascoyne 2013: 64, 73). The notion that these dykes were defensive in nature, to create obstacles in open land rather than to enclose a defensive site, has been the traditional interpretation (e.g. Hawkes and Hull 1947: 15) and is linked to a reading of Caesar’s Gallic Wars (V.21) as well as the etymology of the name Camulodunon. Clearly, they demonstrate a stratified society where labourers could be organized to undertake this monumental work, and the way that they represent the power of those in authority is a far more poignant aspect than their possible defensive or deterrent properties. Sheepen, in the LPRIA and early Roman period (say, c. 15 bc–​Claudian period and continuing beyond), was a trading and industrial settlement and mint. The Sheepen settlement site was located adjacent to the Sheepen Dyke, between the River Colne and a small stream, west of the later Roman fortress and town. Within it, evidence of social stratification and craft specialization was found: round and sub-​rectangular structures, one of which was larger and possibly of higher status, evidence of industrial/​work areas (potteries, metal-​working areas) and storage, as well as two coin-​mould sites (of Cunobelin?) (Hawkes and Hull 1947; Niblett 1985). This settlement was probably controlled and/​or overseen by the ruling elite (who would have controlled coin production and, possibly, other industries), and that it was not affected by the construction of the fort and colonia demonstrates that the local elite held the power to continue to administer their holdings.

The Early Roman Horizon    129 The best hypothesis as to where this elite family was based is at Gosbecks, where there is a large trapezoidal farmstead enclosure (Hull 1958:  259), very different in character from Sheepen, with associated field systems around it visible as cropmarks (Hawkes and Crummy 1995). It is usually considered to be a ‘royal’ farmstead occupied from the LPRIA into the Roman period. A monumental enclosure with enormous ditches, adjacent to the Gosbecks settlement enclosure, may have been the location of an LPRIA shrine or mortuary enclosure (similar to Folly Lane, Verulamium?), which later became a Romano-​British religious site, indicating that its significance and sanctity continued, possibly under the patronage of the ruling family. Abutting a nearby LPRIA dyke, a fort thought to be early Roman has been identified in aerial photographs near to the settlement enclosure. Based on this fort’s apparent similarity to other mid-​first-​century forts, it has been suggested that it either immediately pre-​dates or post-​dates the main fortress that underlay the town (because duplication of military establishments in such close proximity would not be sensible or beneficial) (Hawkes and Crummy 1995: 101). The location of the fort could be interpreted as a way that the invaders (whether under Caligula or Claudius) exerted control over the indigenous population by placing the installation adjacent to the Gosbeck’s enclosure or it could be interpreted as a British-​built fort modelled after Roman plans—​a not-​inconceivable idea that is supported by continental/​Roman-​style military artefacts in high-​status indigenous graves, such as the Lexden tumulus (Creighton 2001: 9). Continued power probably held by the same people in Camulodunon/​Camulodunum from the LPRIA through the Roman period can also be found in the elite burial area at Stanway. The earliest enclosure there dates to the middle Iron Age, the second and largest was late first century bc, the third enclosure was built c. ad 25–​50, the fourth and fifth were securely post-​invasion (Hawkes and Crummy 1995: 170; Crummy et al. 2007). Continuity in burial rites, location, and status at Stanway suggests either that there was no significant break in the power structure, or that the outward expression of status and power was important to communicate to the local population despite the changes brought about by the construction of the fort and colonia. Camulodunon/​Camulodunum is perhaps better understood as an elite-​controlled agricultural holding than as an oppidum (Radford 2013: 42). Other stock enclosures, droveways, and field systems within the area enclosed by dykes and streams were constructed and used across the LPRIA and Roman periods, indicating a dynamic landscape of agriculture and animal husbandry. Various fields, droveways and trackways, and burial areas might have been associated with the dispersed settlement areas. These other settlements and land uses did not come to an abrupt end with the construction of the fort and, later, of the colonia. Continued agriculture and animal husbandry in the oppidum would have been necessary to support the population, but it is likely that unskilled and lower-​ranking veterans were given land to farm that was taken from the indigenous population, although there is evidence from crop marks that LPRIA field systems remained in use (there was no ‘centuriation’, which is uncommon in Britain anyway (see Esmonde Cleary 1987); for further discussion of agriculture and animal husbandry see also Millett and Van der Veen, this volume). Confiscated land for veteran use would have been the most significant change to indigenous farming and, although

130

130   Lacey Wallace that is how we understand the benefit to veterans to have taken shape, there is little archaeological evidence for it.

Conclusions At the time of the Claudian invasion of Britain, the vast majority of people in Britain were living in small rural farmsteads or compounds constructed of ephemeral materials, which are difficult for archaeologists to locate, identify, and study. Many of these rural (often poor) families possessed neither the means nor the motivation to change their home lives in any material way in the decade or two following the Claudian invasion. Many more people did, of course, embrace the changing power structures, new fashions, and connection to the imperial network—​no doubt some with a positive outlook and others with ambition and cynicism—​and a great deal of rebuilding, redesigning, expansion, and taking advantage of new opportunities for social and economic advancement occurred. New urban centres and military settlements had their effects on the surrounding territory, including supply demands. A sliding scale of the degree and velocity of change in the Claudian/​Neronian period can be observed, with greater changes visible in urban centres and high-​status sites and lesser changes in rural areas disconnected from the network of trade and communication. An approach that singles out the Julio-​Claudian/​Neronian periods for examination would probably find little change or adoption of new material culture on most rural sites. Nonetheless, it would be of great interest to determine the difference between rural sites that do produce material evidence of these periods and those that do not. From the theoretical paradigms (albeit explicitly not discussed here), we must move forward to robust investigations of the evidence, especially the ‘grey literature’ (e.g. that from London, which is incorporated into Wallace 2014) that is far more data-​rich than traditional publications. The post-​excavation work of commercial archaeological specialists is often funded, and their reports can be used to identify and analyse the early Roman levels, even where they are not combined into a finished piece of research. Rigorously comparing the features, deposits, and artefacts from this decade-​or-​so of early urban development in Londinium to that of a contemporaneous site with a large and powerful LPRIA society (for example, Verulamium) or a known military population (for example, Colonia Claudia) would also be of great import. Such work is generally difficult, because such temporal specificity is so rare. Such comparisons would not be unproblematic, of course. For example, Londinium had no local coin production, cemeteries, elite burials, enclosures, or high-​status farmsteads dating to the earlier first century ad; Verulamion had all of these. Although the people of Verulamion did not adopt imported pottery styles, coin use, and burial behaviours as early as indigenous people further north and east (for example, in east Hertfordshire), the explanation probably lies in cultural differences between indigenous groups, rather than in supply or economic obstacles. The society of Verulamion/​Verulamium was socially stratified, and

The Early Roman Horizon    131 resources were directed to large ‘public’ works, such as the dykes and ‘Central Enclosure’. Londinium had no such indigenous population—​indigenous British residents of the town would have had to move there. Any comparison with a site like Verulamion/​ Verulamium would have to encompass the LPRIA levels. Despite Verulamium’s pre-​Roman credentials, it—​like Londinium—​was included in the wide-​sweeping fort-​to-​town urbanization hypothesis. Verulamium has the added benefit of a recent reappraisal and broad reinterpretation (Niblett and Thompson 2005; Niblett et al. 2006). Comparing and contrasting the patterns of structural development and activities between these sites would produce an excellent context within which to situate future discussions of the early Roman horizon. Changes in Verulamium and Londinium were driven by indigenous and immigrant groups, but very little by military or imperial administration. Showing how this process occurred in such different cultural contexts would be illuminating. We have a tendency to identify with the ‘Romans’ in a way that we do not with indigenous Britons—​we think we understand the Romans, that they are not mysterious and superstitious, but rather that they acted with a dispassionate precision and planning for the control of the province. Even if one can presume to understand the character and decisions of the governor and his staff, it was the ‘regular’ people who made up the bulk of the population and who were responsible for most of the archaeological record. When studying humans, we must remember that pride, personalities, stubbornness, greed, laziness, ambition, weakness, devotion, sentimentality, and a whole host of other factors affect the decisions that people make. There is no reason to assume that the ‘obvious’ or most socially or economically ‘beneficial’ reactions to the Roman invasion were observed. Far better for us to begin with the evidence and challenge hypotheses until we find one that cannot be disputed easily. The re-​examinations of evidence (e.g. Creighton 2006; Niblett et al. 2006; Wallace 2014) and narratives constructed around more recent early Roman finds (e.g. Manley et al. 2005) show that large compilations of data, examined without predetermined patterns in mind, reveal that people of all kinds drove change for a variety of reasons, producing complex relationships between authority, power, and the archaeological record.

References Primary Sources Caesar, The Gallic War, trans. H. J. Edwards (London: Heinemann, 1917). Cassius Dio, Roman Histories, trans. E. Cary et al. (London: Heinemann, 1960–​70). Tacitus, Agricola, trans. M. Hutton, rev. R. M. Ogilvie (London: Heinemann, 1970). Tacitus, Annals, trans. J. Jackson (London: Heinemann, 1925–​37).

Secondary Sources Creighton, J. (2000). Coins and Power in Late Iron Age Britain. Cambridge:  Cambridge University Press.

132

132   Lacey Wallace Creighton, J. (2001). ‘The Iron Age–​Roman Transition’, in S. James and M. Millet (eds), Britons and Romans: Advancing an Archaeological Agenda. CBA Research Report 125. York: Council for British Archaeology, 4–​11. Creighton, J. (2006). Britannia:  The Creation of a Roman Province. London and New York: Routledge. Crummy, P., Benfield, S., Crummy, N., Rigby, V., and Shimmin, D. (2007). Stanway: An Élite Burial Site at Camulodunum. SPRS Volume 24. London: Society for the Promotion of Roman Studies. Cunliffe, B. (1974). Iron Age Communities in Britain:  An Account of England, Scotland, and Wales from the Seventh Century bc until the Roman Conquest. London: Routledge. Davies, B., Richardson, B., and Tomber, R. (1994). A Dated Corpus of Early Roman Pottery from the City of London. CBA Research Report 5. York: Council for British Archaeology. Eckardt, H. (2002). Illuminating Roman Britain. Montagnac: Editions Monique Mergoil. Esmonde Cleary, A. S. (1987). Extra-​Mural Areas of Romano-​British Towns. BAR British Series 169. Oxford: Archaeopress. Frere, S. (1967). Britannia: A History of Roman Britain. London: Routledge & Kegan Paul. Frere, S. (1972). Verulamium Excavations: Volume I. London: Society of Antiquaries of London. Frere, S. (1983). Verulamium Excavations:  Volume II. London:  Society of Antiquaries of London. Gascoyne, A. (2013). ‘The Roman Legionary Fortress, ad 43–​49’, in A. Gascoyne and D. Radford (eds), Colchester: Fortress of the War God. An Archaeological Assessment. Oxford: Oxbow Books, 59–​76. Gascoyne, A., and Radford, D. (2013). Colchester: Fortress of the War God. An Archaeological Assessment. Oxford: Oxbow Books. Haselgrove, C. (1984). ‘Romanisation before the Conquest’, in T. F. C. Blagg and A. C. King (eds), Military and Civilian in Roman Britain. BAR British Series 136. Oxford: Archaeopress, 5–​63. Hawkes, C. F.  C., and Crummy, P. (1995). Camulodunum 2. Colchester:  Colchester Archaeological Trust. Hawkes, C. F. C., and Hull, M. R. (1947). Camulodunum: First Report on the Excavations at Colchester 1930–​1939. London: Society of Antiquaries. Hull, M. R. (1958). Roman Colchester. London: Society of Antiquaries of London. Laurence, R., Esmonde Cleary, A. S., and Sears, G. (2011). The City in the Roman West: c.250 bc–​c.ad 250. Cambridge: Cambridge University Press. MacDonald, W. (1986). The Architecture of the Roman Empire. New Haven:  Yale University Press. Manley, J., Rudkin, D., Sykes, N., Lyne, M., Dannell, G., Scaife, R., Somerville, E., Barber, L., Allen, D., Williams, D., Pelling, R., and Clegg, S. (2005). ‘A Pre-​ad 43 Ditch at Fishbourne Roman Palace, Chichester’, Britannia, 36: 55–​99. Marsden, P. (1987). The Roman Forum Site in London: Discoveries before 1985. London: Museum of London. Merrifield, R. (1965). The Roman City of London. London: Ernest Benn. Neal, D. S. (1990). Excavation of the Iron Age, Roman, and Medieval Settlement at Gorhambury, St Albans. London: Historic Buildings and Monuments Commission for England. Niblett, R. (1985). Sheepen: An Early Roman Industrial Site at Camulodunum. CBA Research Report 57. London: Council for British Archaeology. Niblett, R. (1999). The Excavation of a Ceremonial Site at Folly Lane, Verulamium. London: Society for the Promotion of Roman Studies.

The Early Roman Horizon    133 Niblett, R. (2001). Verulamium: The Roman City of St Albans. Stroud: Tempus. Niblett, R. (2005). ‘Roman Verulamium’, in R. Niblett and I. Thompson (eds), Alban’s Buried Town: An Assessment of St Alban’s Archaeology up to ad 1600. Oxford: Oxbow Books, 41–​165. Niblett, R., Manning, W., and Saunders, C. (2006). ‘Verulamium:  Excavations within the Roman Town 1986–​88’, Britannia, 37: 53–​188. Niblett, R., and Thompson, I. (2005). Alban’s Buried Town:  An Assessment of St Alban’s Archaeology up to ad 1600. Oxford: Oxbow Books. Partridge, C. (1981). Skeleton Green: A Late Iron Age and Romano-​British Site. London: Society for the Promotion of Roman Studies. Perring, D. (2011). ‘Two Studies on Roman London: A. London’s Military Origins—​B. Population Decline and Ritual Landscapes in Antonine London’, Journal of Roman Archaeology, 24/​1: 249–​282. Philp, B. J., Merrifield, R., Dannell, G., Couldrey, P., Stant, M. Y., and Williams, W. (1977). ‘The Forum of Roman London: Excavations of 1968–​9’, Britannia, 8: 1–​64. Radford, D. (2013). ‘Camulodunon in the Late Iron Age, c.50 bc–​ad 43’, in A. Gascoyne and D. Radford (eds), Colchester:  Fortress of the War God, An Archaeological Assessment. Oxford: Oxbow Books, 33–​57. Reece, R. (1988). My Roman Britain. Cotswolds Studies Volume 3.  Cirencester:  Cotswold Studies. Schnurbein, S. von (2003). ‘Augustus in Germania and his New “Town” at Waldgirmes East of the Rhine’, Journal of Roman Archaeology, 16: 93–​108. Tyers, I. (2011). ‘Tree-​Ring Analysis’, in J. Hill and P. Rowsome (eds), Roman London and the Walbrook Stream Crossing:  Excavations at 1 Poultry and Vicinity, City of London. London: Museum of London, 562–​567. Wallace, L. (2014). The Origin of Roman London. Cambridge: Cambridge University Press. Webster, G. (1966). ‘Fort and Town in Early Roman Britain’, in J. Wacher (ed.), The Civitas Capitals of Roman Britain: Papers Given at a Conference Held at the University of Leicester, 13–​15 December 1963. Leicester: Leicester University Press, 31–​45.

134

Chapter 7

Britain at t h e E nd of Empi re Simon Esmonde Cleary

Introduction This section of the Handbook is concerned with Britain in the Roman Empire, helping us to combat the tendency to scholarly insularity that is particularly pernicious when dealing with the Roman period, since Britain formed part of a major empire and thus was directly influenced by decisions or trends originating outside the island. This chapter will seek to counter that risk effectively, by looking at Britain from across the Channel from Gaul, to assess to what extent late Roman Britain was part of and influenced by wider developments. This is not to say these wider trends and developments explain all that was going on in Britain; they do not, since Britain was a region (or regions) within the wider western provinces with its (or their) own particularities. But if we can identify and factor in these more general developments, then it may be easier to look in new ways at some of what was going on in Britain while at the same time defining features of the archaeology that do need to be looked at on the insular level. I want to look at four major sets of interrelated developments occurring on the continent in the fourth and fifth centuries of which Britain formed part, and which may help to explain why aspects of the archaeology of the island were the way they were. These are: the withdrawal of imperial power; the effects of this on the economic structures of the west; the effects of this on aspects of elite culture and auto-​representation for the ‘civilian’ aristocracies; the development of new forms of elite display linked to military-​style identities along with changes in settlement types.

Britain at the End of Empire    135

The Withdrawal of the Imperial Presence One of the defining features of the later fourth and fifth centuries across the west was the withdrawal of the imperial presence and of the exercise of imperial power from much of Gaul, the Rhineland, and Britain. Since the mid-​third century, the provinces north of the Alps had more often than not had an emperor (legitimate or ‘usurper’) resident, from the end of the third century mostly at Trier. This meant that northern Gaul, the Rhineland, and Britain were firmly on the agenda of a member of the imperial college. After the suppression of Magnus Maximus in 388 ad, Milan then Ravenna became the principal imperial seat in the west. Trier’s manifold administrative functions likewise moved south at the turn of the century, with the Praetorian Prefecture transferred to Arles. To emperors from the end of fourth century, northern Gaul, the Rhineland, and Britain were no longer close by and high priorities; they were far away and lower priorities. The Mediterranean mattered more. From the third century the region between the Rhine and the Loire had become increasingly militarized (Esmonde Cleary 2013a: ch. 2). In part this was militarization in the strict sense, with the presence of military garrisons and installations not just on the old frontier of the Rhine but also along the new maritime frontier of the southern North Sea and Channel. But the military were far more present than they had been in the interior at a range of new fortifications, including the fortified road linking Cologne and Boulogne and thus Britain, and with the settling of non-​Roman formations (laeti, foederati) in the interior. As well as garrisons, there were elements of the state ‘infrastructure’, such as the fabricae across northern Gaul. The Rhine–​Loire region was also the heartland of the heavily fortified ‘fortress cities’ behind their thick, high walls bristling with towers. Excavation and re-​evaluation have led to the old idea of these being hasty responses to the ‘barbarian invasions’ of the later third century being abandoned; the excavated dates range from the later third, through the early fourth, and down to the second half of the fourth century, and the volume of material that went into them and the care with which they were constructed makes it clear these were major monumental undertakings. Since these were sited at what under the High Empire had been cities, civitas-​capitals, the assumption is often that they remained urban in the late period. For some this is demonstrably untrue: sites such as Bavai or Jublains had become small fortified strong-​points. Even sites where the defences enclosed a larger area, such as Amiens, exhibit little evidence that could not be interpreted in a military sense. Another class of fortifications was the regional tradition of hilltop fortifications in north-​eastern Gaul between the Meuse and the middle Rhine. Because excavated examples have sometimes been associated with burials containing weapons and ‘Germanic’ objects, these have been seen as strong-​points with ‘barbarian’ garrisons. The ‘Germanic’ objects form part of a larger complex of late Roman dress fittings: for men primarily belt fittings and brooches (Böhme 1974; Swift 2000: chs 2, 5, and refs.), for women principally brooches

136

136   Simon Esmonde Cleary along with beads and bracelets (Swift 2000: chs 3, 4). Long considered as markers of ethnicity (e.g. Böhme 1974), much of this material is now recognized to be the markers of late Roman officialdom, military or civilian, and thus to say more about the relationship of the wearer to the Roman state (see also Allason-​Jones, this volume). The latter part of the title of Böhme (1974), ‘zwischen Elbe und Loire’ (between Elbe and Loire) makes the point that this material is also a phenomenon of northern Gaul—​indeed concentrated in the Seine–​Rhine region. Moreover, traditional markers of civilian elites over much of northern Gaul, in particular the villa but also urban residences, had shrunk greatly in both numbers and elaboration (Van Ossel 1992). It can be argued that the ‘militarization’ of northern Gaul went wider, with the vocabulary of military self-​representation having been taken up by the local elites instead of the civilian markers of the High Empire. Another way in which the region may have been ‘militarized’ was in the increasing dominance of state and military demands on the economy in order to support the garrison units, the fabricae, and their personnel. What happened to the Roman military complex in the fifth century is hard to read from the archaeology, partly because of the lack of good archaeological sequences. A key site on the middle Rhine is that of the fort of Alzey, constructed under Valentinian I (364–​75 ad), partially burnt down early in the fifth century, but then in part rebuilt before being destroyed by fire in the middle of the century (Oldenstein 1986). The material culture from the first half of the fifth century is more ‘Germanic’ in that its origins lie east of the Rhine. Another key site lies lower down the Rhine at Krefeld-​Gellep, where the extensive excavations of the cemeteries dating from the second century onwards (Pirling 1986; Pirling et al. 2000) have shown that from the turn of the fourth and fifth centuries male and female burials grouped in discrete areas contained increasing amounts of ‘Germanic’ material, not just dress accoutrements but also pottery of non-​ Roman fabrics and forms. A similar situation seems to obtain on the upper Rhine at Kaiseraugst (Drack and Fellmann 1988: 3–​42, 411–​414); again extensive excavation in the cemeteries has shown that from the end of the fourth century ‘Germanic’ grave-​ goods increasingly make their appearance. Because of the textual master narrative for the period of the replacement of Roman imperial power by the nascent ‘Germanic’ successor states, this evidence has been assimilated to that framework. Even if one were to accept (which I do not) an easy equivalence between object and ethnicity, the presence of ‘German’ men with their womenfolk does not of itself mean the garrisons were no longer part of a standing Roman army—​the Romans had employed non-​Roman troops since the early Republic. Moreover, one must beware of decontextualizing ‘Germanic’ burials from the larger majority of burials in the continuing Roman-​provincial rite. Apart from these key sites, there are few places where the Roman army of the fifth century in northern Gaul and the Rhineland can be pursued by archaeology. The lesson of Krefeld and Kaiseraugst seems to be that cemeteries may be the most revealing sites to excavate. If the fate of the standing Roman army in northern Gaul and the Rhineland is difficult to apprehend through archaeology, there is another class of material that has important lessons: the coinage. The current model for the striking and distribution of

Britain at the End of Empire    137 the tri-​metallic coinage of the fourth century is closely linked to the army in that its primary purpose is seen as the discharge of state obligations through gold and silver and the recapturing of the precious metals in part through the bronze coinage. In the west, army pay was probably the largest single head of expenditure by the fiscal authorities. Of course, pay in coinage was not the only benefit the soldiery received, and we shall look at some effects of the supply of materiel to the troops. At the turn of the fourth and fifth centuries the part of the fiscal cycle involving the distribution of coin seems to break down (Kent 1994; cf. Guest 2005). Put briefly: at this time gold was struck only sporadically in the west; silver was struck in quantity (if not consistently), with a major issue struck at Milan from 397–​402 ad, but the later issues of 408–​25 ad from Milan and Ravenna remained largely distributed within Italy; bronze was issued in declining quantity after the reorganization of the mints in 395 ad, until 402 ad, after which the issues from Rome were also rare north of the Alps. From 402 ad there was little coin penetrating north of the Alps, and the occasional striking at Trier did not correct this. The logic of this would seem to have to be that the Roman state no longer remunerated its troops with elements of pay in precious metals, even if this was an unforeseen consequence of the invasions of Gaul and Italy at the time. What happened to the proportion of remuneration in kind—​foodstuffs, animals and clothing, equipment and weaponry supplied from the fabricae—​is very difficult to tell, though the dating of some of the ‘official-​issue’ brooches and belt fittings suggests they were being produced into the first half of the fifth century (Böhme 1986). But by the second half of the century the regional variations indicate local rather than centralized production, suggesting that, though such metalwork retained ideological significance, official production of it had by then ceased. If the Roman state had de facto ceased to pay and increasingly not to supply its army from the start of the fifth century, then the likelihood must be that a standing Roman army of the type with which we are familiar ceased to exist, or transmuted into something else. The literary sources can be read in this way, with the military history of the west in the first half of the fifth century increasingly becoming a matter of strong military commanders, with personal followings (bucellarii) putting together coalitions in order to fight campaigns (Whittaker 1994: ch. 7). Some of these commanders were of Roman origin, but the army assembled by Aetius to confront Attila at the Catalaunian Fields in 451 ad was a coalition of a range of ethnicities, with very few troops that could be described as ‘Roman’ save that they were in the military retinue of Aetius himself. Military force seems effectively to have passed from the public power of the state to the private power of war-​leaders; some of these fought for the emperor, some were ‘Germanic’ kings. Such a series of developments would accord with the premiss put forward at the start of this section that the withdrawal of the imperial presence meant that northern Gaul, the Rhineland, and Britain were now low on the list of imperial priorities. The picture proposed here is one that is not unlike that proposed for the military zones of Britain. Fourth-​century Britain had two major defensive systems, Hadrian’s Wall and its hinterland (Collins 2012: ch. 3) and the Saxon Shore (Pearson 2002). The latter related to the fortifications on the other side of the Channel. It has also yielded a very considerable number of brooches and belt-​fittings of ‘official-​issue’ type (Swift

138

138   Simon Esmonde Cleary 2000). Interestingly, the distribution of these latter is almost entirely southern, with a marked absence from Trent to Tees and west of the Wye, the supposed ‘military zone’, but a presence along the Wall. On the analogy with Gaul it could, therefore, be argued that southern Britain also demonstrates a degree of ‘militarization’, though, as we shall see, to nothing like the same extent. In fact the distribution of this material shows a preference for the Saxon Shore forts and urban sites, so may be related to the garrisoning of such sites, or at least some form of military or official presence at the cities. From the turn of the fourth and fifth centuries the archaeology at military sites starts to change significantly. At the forts of Hadrian’s Wall and its hinterland there is a series of changes, suggesting a transition from a relatively coherent and organized situation to one where individual forts were adapted piecemeal, suggesting a transition to something much less regular than a ‘Roman’ army (Collins 2012: ch. 4). The suggestion is that instead we are seeing a series of smaller-​scale military groups becoming increasingly self-​reliant, both in terms of their identities and in terms of their economic base. This pattern would fit the wider evidence from the west already outlined, aligning the situation in Britain with the continent. It is perhaps worth noting that one of the building types regularly modified in the Wall zone was the horreum (the store building); that these buildings were no longer needed would also fit with the idea that the state had withdrawn from provisioning its troops. The evidence from the Saxon Shore is less good because of lack of excavation, but it is worth noting that the occupation sequence at Portchester runs through from the fourth century, when the site is supposedly a fort of the regular Roman army, into the fifth, when the material culture starts to change from Roman-​provincial to ‘Germanic’ (Cunliffe 1976: 121–​127, 301), echoing sites such as Alzey. It would be fascinating to see excavated a large late cemetery attached to a late fort in Britain. The wider pattern of the withering of coin supply north of the Alps is reflected in Britain with the likely impacts of this on the nature of the army; the smattering of coins post-​388–​402 ad from Britain would also reflect wider patterns (e.g. Collins 2008).

The Disintegration of the Economic Systems The disengagement of the state from the military sphere at the turn of the fourth and fifth centuries had implications not just for the army itself but also for the systems of supply that the state had engendered in order to sustain the army and thus of the other economic activities that had depended from these systems (Wickham 2005: ch. 11). It is clear that the ‘imperial’ economy, a system of redistribution by the state from the taxpayer to politically crucial groupings (the army, the administration, the city of Rome), although powered by taxation, involved the movement of material at state expense, allowing a range of other goods to travel ‘piggy back’, with their movement effectively subsidized by the state. This related to the operations of the ‘market’ economy, the movement

Britain at the End of Empire    139 and exchange of goods for economic profit, where the general stability of the empire encouraged longer-​distance links, and the ‘imperial’ economy promoted the distribution of some goods. Reciprocity, the mutual provision of goods and services for social reasons and profit, remained the bedrock of much economic activity in the late Roman world, most usually at the level of the ‘peasant’ economy, with goods moving over relatively short distances, but there was also the ‘prestige’ economy of reciprocity, causing high-​value items to move long distances. None of these forms of economic activity or the participants in them was sealed off from the others; the operations of the ‘imperial’ economy impacted on the ‘market’ economy, and individuals might swap between the various forms either as suppliers or as consumers. The discussion here will focus on the ‘imperial’ and the ‘market’ economies, partly because they are the most identifiable in the archaeological record, partly because they were clearly profoundly affected by the events of the fifth century. It will be based on the ceramic evidence because of its ubiquity and the amount of study it has attracted—​as so often, ceramics that survive must stand proxy for perishables. The industries with a regional or supra-​regional reach are most clearly exemplified by the ‘Argonne’ wares produced in north-​eastern Gaul and the Paris basin and distributed over the northern half of Gaul, with lesser amounts in southern Gaul, the upper Danube provinces, and Britain (Van Ossel 2011). It is found in quantity at military sites and the ‘fortress cities’, suggesting it may have been travelling in part ‘piggy-​back’ on the movement of supplies for the army and state servants. It was probably also being marketed, with the transport network and privileges of state supply allowing it to compete on favourable terms over long distances. Another industry that seems to have been attuned to military supply was Eifelkeramik/​Mayen ware, with its distribution in the Rhineland and south-​eastern Britain. There were other industries with a regional distribution that cannot be seen to be so dependent on the state nexus, such as the developing céramiques rugueuse and grumuleuse centres of the Île de France and or the Jaulges/​Villiers-​Vineux complex in northern Burgundy, which seem to have served more civilian markets. Other industries such as the céramique craquelée bleutée of the Champagne achieved subregional distribution (for these industries, cf. Brulet et al. 2010: §§29, 9, 28) Overall, the existence of the late Roman state seems to have promoted the levels of integration represented by the ‘imperial’ and ‘market’ economies. Turning to Britain, we have already seen that some continental products such as Argonne or Mayen entered Britain; equally, some British products went the other way, with surprising quantities of Black-​Burnished I now recognized along the Channel coasts and penetrating inland into Picardy and the Paris basin, and some Hadham, Oxfordshire, and New Forest products having made the crossing (Adrian 2006). Within Britain, it is possible to recognize the effects of the levels of integration already discussed for Gaul. Some industries seem to have achieved wide distributions partly on the back of ‘official’ requirements such as foodstuffs. This would seem to be the case in the north for the products of the Crambeck/​Malton kilns, with their wide distribution in the ‘military zone’ up to and including the Wall (cf. Evans 2000; see also Evans, this volume). Further south, the Oxfordshire industry could almost be seen as the British Argonne in terms of products and spread, with its distribution down the

140

140   Simon Esmonde Cleary Thames seemingly preferentially linked to military sites and London (Fulford and Hodder 1974). London was also a major destination for the Alice Holt industry. Some industries may have existed primarily to supply particular installations. Equally, many other aspects of the Oxfordshire distributions can be explained more in ‘market’ terms, as can industries such as the Nene Valley or the New Forest. Other, smaller-​scale industries probably served small catchments, perhaps through cities or ‘small towns’, or else may have been linked to major rural landholdings. One may argue for Britain, as for the west more generally, that its membership of the Roman Empire at this date had profound consequences for the way the island’s economy was structured and the degree and levels of integration. What then of developments on either side of the Channel in the fifth century? In northern Gaul in particular the fifth century sees continuing evidence from the ceramics (also other classes of material such as glass and metalwork) for continued production and distribution at regional, subregional, and local levels. At the regional level, Argonne wares continued to be distributed across much of the northern half of Gaul, with a diminution in quantity and in reach over time, so that by the end of the century it was at best a subregional industry. Other industries such as the céramiques grumuleuse/​rugueuse also continued to have significant distributions in the Paris basin and neighbouring areas in the first half of the century, less so thereafter; interestingly at the end of the fifth century vessels in ‘Germanic’ forms and decoration start to be produced in these fabrics. The Jaulges/​Villiers-​Vineux industry seems to have continued in production through the first half of the century, though probably on a diminishing scale. South of the Loire, ceramic productions continued uninterrupted, with some of them such as DSP (dérivées des sigillées paléochrétiennes—​a stamp-​ decorated repertoire) perhaps attaining their maximal spread in the first half of the century (Rigoir 1968). It is clear that the scale and reach of production, particularly in the northern half of Gaul, diminished as the fifth century advanced, with even the few remaining major industries, such as Argonne, a shadow of their fourth-​century selves by the end of the fifth century. By then other industries had either ceased production or become small-​scale local enterprises. The economic integration of the fourth century had undergone significant disintegration; nevertheless, there were still functioning ceramic industries, and more desirable commodities such as glass continued to be traded over considerable distances (Foy 1995). The contrast with Britain is clear: there the Roman-​style pottery industries seem to have ceased production and distribution in the first half of the fifth century. Indeed, some industries seem already to have been in decline in the latter part of the fourth century. The dating of this is debated, because of the dependence of pottery dating on the availability of coin as a dating medium. Nevertheless, the debate centres on how far into the first half of the fifth century pottery production continued rather than whether it continued to the end of that century. If we accept ceramics as a proxy of the levels of economic activity and integration, then the island suffered a breakdown and disintegration in economic activity of a depth and geographical spread that Gaul simply did not experience (cf. Esmonde Cleary 2013b).

Britain at the End of Empire    141

Elite Auto-​R epresentation: The ‘Civilian’ Tradition The consideration of Gaul north of the Loire in the fourth century proposed that it became a ‘militarized’ region in the later Roman period. But Gaul south of the Loire was a very different world, one that in many ways is more easily related to southern Britain. Apart from some cities defended under the High Empire and a small number of the ‘fortress cities’ that characterized the north of Gaul, the cities of the southern half of Gaul remained undefended. So far as can be seen, the street grids continued to articulate the urban landscapes, although at many cities where there has been sufficient excavation it is clear that the area occupied had contracted from its late-​second-​century peak to more restricted areas in the fourth century. The public buildings and monuments provided under the High Empire do not seem to have been to be subjected to any general programme of disuse or suppression, though there was variation by monument type (Heijmans 2006a). At these south Gaulish cities, therefore, the layout and monumental buildings inherited from the earlier empire seem to have been preserved into the later period. It can be argued that this was a question not just of the passive reception of these earlier buildings but of their active curation under the late empire and thus self-​ representation that perpetuated civic ideologies inherited from the earlier empire. Another major difference between Gaul north and south of the Loire was the presence of large numbers of villas, many of them extensive and luxuriously appointed, in southern Gaul. The heartland of this villa culture was in the south-​west, with fewer in the south-​east and with only a scatter in west-​central Gaul (Balmelle 2001). A common feature of the larger villas was the elaboration of their plan and architecture, making widespread use of the apsidal form for the principal room, and also on occasion the triconch plan that allows the identification of the main dining suite. In addition, elaborate bath-​suites were a regular feature of such complexes, with ambitious three-​dimensional structures. The major ‘public’ areas of these villas were the setting for the decor of display with the use of marble (available from the Pyrenees) for architectural detailing and the use of mosaic for flooring. The example of the villa at Chiragan south-​west of Toulouse makes clear the importance of sculpture as a status display medium; evidence from other villas in Gaul and Spain show that sculpture was more widespread and significant than has perhaps been appreciated. The subject matter of the mosaics and the sculpture was firmly in the Mediterranean traditions of divinities, heroes, and famous men and women, along with the mythology of which they formed part. This was not the only area of the west that exhibited such a concentration; others lay in Spain, on the Catalonian littoral, on the northern meseta, and in southern Lusitania (Chavarría 2007). Another small but important group lay in north-​eastern Gaul and the Rhineland, in the hinterlands of Cologne and especially Trier (Van Ossel 1992). What we are seeing is a major, alternative form of aristocratic self-​representation, contrasting with the militarized one dominant in northern Gaul. This aristocratic vocabulary was

142

142   Simon Esmonde Cleary one related to two hugely important social groupings in the late Roman world, the imperial court and the senatorial order; to an extent the elaborate villas of Gaul and Spain can be read as scaled-​down reflections of imperial grandeur and of senatorial opulence. It is also the aristocracy of education and learning, of paideia, whereby regional aristocracies were subsumed into a common framework of education and thus cultural formation and expectation, within which the display of the traditional mythologies and personalities operated. It is clear that in certain important respects southern Britain better matches southern than northern Gaul as regards the evidence of the cities and villas along with aspects of their material culture. The cities of late Roman Britain parallel their south-​Gaulish cousins in retaining their gridded layouts, many of their inherited public buildings and monuments, and in remaining centres of elite residence and economic activity. The retention, indeed curation, of earlier buildings and monuments varies by city and by class of monument, but as with their southern equivalents there is no evidence for a concerted plan of suppression of these structures, though again with variation by city and by monument class. A major difference at the cities of Britain is the presence of walls. This was a historical peculiarity of the island and had resulted in the majority of the cities of Britain being enclosed from the later second century by defences enclosing the then inhabited area—​with even those cities defended later on (for example, Canterbury, Chichester) enclosing larger areas than the ‘fortress cities’ of Gaul. Moreover, the British walls were less often provided with regular projecting towers (‘bastions’) than the north-​ Gaulish circuits. The similarities between elite residences and their appointments in the villa-​rich regions of southern Gaul and Spain, on the one hand, and southern Britain, on the other, is striking, in particular for villas, but also for urban residences, where our information for Britain is much better than for the continent. Comparison of the plans and layouts of the larger villas and town houses, types of principal rooms, types of decoration (such as mosaic and sculpture), along with their subject matter, with those of south-​western Gaul and Spain show the elites of Britain to have been well aware of fashions in these matters and to have participated in a wider common vocabulary of elite auto-​representation. The elaboration of the decor of these residences is well known, and the corpus of mosaics from Britain reveals the range of subjects of figured mosaics (Neal and Cosh 2002–​10). Like their counterparts in Gaul and Spain, they display their commissioners’ knowledge of Mediterranean culture and their internalization of the linguistic, literary, and philosophical markers of paideia. So at one level southern Britain forms one of a series of regions in the late Roman west exhibiting this sort of civilian elite display, but, as has been noted, even within Britain such elite display is regionalized. One region was East Anglia, where villas were architecturally modest but which nevertheless contained the largest and richest concentration of silver plate, coin, gold jewellery, and semi-​precious stones in the empire—​deposited it seems in the first half of the fifth century but probably manufactured and in circulation in the course of the fourth (Hobbs 2006).

Britain at the End of Empire    143 One area where southern Britain seems to diverge markedly from the continent is in its expressions of religious activities. In the northern half of Gaul the established narrative has been the destruction of the temples of the traditional religions either by barbarians in the later third century or by militant Christians such as Martin of Tours in the second half of the fourth. Britain has long been known for the way in which its temples continued to be used through the fourth century, along with other manifestations of cult or ritual practice such as the wet-​place deposition (Henig 1984: ch. 10; Woodward 1992). There is increasing evidence that the picture for Gaul may be overdrawn and that sanctuaries and cults of the traditional religions may have lasted longer than has been thought (Esmonde Cleary 2013a: 19–​197). On the other hand, the evidence for Christianity in Britain is not as well developed as that for Gaul (Petts 2003; see Petts, this volume, for further discussion of Romano-​British Christianity). Clearly there are key Christian sites in Britain, such as the Lullingstone wall-​paintings or Water Newton Christian plate, but as yet there is no good evidence for the sort of urban church-​building that was taking place from the second half of the fourth century in some of the cities of Gaul, and there are only one convincing and one probable late Roman Christian tombstones from Britain (RIB 955 Carlisle; RIB 787 Brougham, both interestingly in the ‘military zone’) compared with the burgeoning of such commemoration in Gaul (Handley 2003). Nor are there any sarcophagi of continental Christian type. These two absences suggest that in Christianity, as in the traditional religions, Britain differed from continental practice. The demise of the ‘civilian’ forms of elite auto-​representation in southern Gaul and Spain has been written largely from the settlement rather than the cemetery evidence, in particular the villas. It was the fifth century that saw the ‘militarization’ of the cities of southern Gaul by the construction of wall circuits, many of them very limited in extent (Heijmans 2006b). The form and dating of the latest occupation at many villas in Gaul and Spain are difficult to reconstruct—​partly because many of the major sites were excavated some time ago, partly because of difficulties in the dating. But recent surveys of the evidence for south-​western Gaul (Balmelle 2001: 115–​123) and for Iberia (Chavarría 2007: 114–​156) concur in seeing the middle of the fifth century as a horizon after which the high-​quality residences were no longer maintained and progressive degradation of the structures and their decor ensues—​with subdivision of space by timber structures (‘squatters’), new timber structures, or sometimes the installation of cemeteries. The reasons for this abandonment of the villa culture are as yet obscure. It was probably in part economic, with a combination of the reduction of resources consequent upon politico-​legal fragmentation and a reduction of the specialists required to sustain such a culture. But as well as the change in landholdings and the organization and social relations of production, it clearly signals the progressive collapse of the ‘civilian’ elite culture of paideia and its material correlates. The other component of the classic late Roman city was, of course, the church, which became a more general feature of the Gaulish urban landscape in the fifth century and preserved many of the features of ‘civilian’ prestige architecture and a Christianized paideia. This summary of developments in southern Gaul should be familiar in outline to those studying the progressive dereliction of

144

144   Simon Esmonde Cleary Romano-​British cities and villas, save that the dating for this in Britain is the first rather than the second half of the fifth century and for Britain there is not the displacement to prestige church buildings and decoration on the continental model. However, the chronology for the remaining villas of northern Gaul fits rather better with that for Britain than with their cousins south of the Loire (Van Ossel 1992: 8–​82). The demise of Britain’s cities and the villas may, in part, have been due to the politico-​military withdrawal of the imperial presence early in the fifth century. It may also have been, in part, a consequence of the economic dislocations that ensued (Esmonde Cleary 1989: ch. 4), but what it also represents is the collapse of the ‘civilian’ modes of elite self-​representation, probably, in large part, because they no longer had validity without the socio-​political system that gave them a frame of meaning. What may have replaced them we shall now consider.

The Archaeology of the Fifth Century: Settlements, Cemeteries, and ‘Militarization’ If the fifth century witnessed the demise of the ‘civilian’ forms of status display, then what replaced them? In the first section of this chapter it was suggested that in Gaul north of the Loire the later Roman period was characterized by the adoption of military-​derived vocabularies of power and representation, which contrast with the vocabularies we have just been looking at south of the Loire and southern Britain. It will be the contention of this section that in the course of the fifth century the ‘civilian’ elite vocabularies were progressively replaced by the ‘military’ ones. This is a trend that has traditionally been subsumed within another narrative—​that of the spread of the historically attested ‘Germanic’ peoples across Gaul, Spain, and Britain. This has been based in particular upon burial evidence and the supposed ‘Germanic’ style both of the burial rite, especially ‘weapon graves’, and of the grave-​goods, above all the metalwork and, to a lesser extent, the pottery. Here a certain separation of the two narratives will be attempted. The demise of the tradition of construction in stone (churches apart) and the (re-​) emergence of traditions of building in timber is one of the widespread archaeological signals of the transition from the Roman to the early medieval. It has often been seen as an index of ‘Germanization’, especially where particular building forms such as the ‘hall’ or the sunken-​featured building (SFB) are present. More recent evidence and analysis suggests proceeding with more caution. The resurgence in timber building was already under way in northern Gaul in the fourth century, including at villa sites (Van Ossel 2003, 2010), and continued into the fifth, with both ‘halls’ and SFBs in association with former villas as well as on new sites, but always on a more restricted scale than their villa predecessors (Peytremann 2003: 318–​319). The presence of structures of the ‘long house’ or Wohnstallhaus as early as the fourth century in the far north of Gaul

Britain at the End of Empire    145 (Toxandria) as well as occasionally further south has been linked to the narratives of the settlement of the Salian Franks in the former and of ‘Germanic’ troops and others in the latter. Although these building types do have strong geographical links with areas either side of the lower Rhine and beyond, their origins, especially of the SFBs, are not always clear (Hamerow 2002: 31–​35). The presence in them from the fifth century of metalwork and ceramics of ‘Germanic’ types has been read likewise as evidence of the spread of Germanic, more specifically Frankish or Anglo-​Saxon, peoples. A problem here is that this material is often decontextualized from the larger quantities of Roman-​provincial material culture at these sites, risking giving a misleading impression by focusing on the ‘Germanic’ material. A major recent consideration of early medieval settlements in northern Gaul (Peytremann 2003) carefully eschewed any ‘ethnic’ interpretation and instead discussed such sites in terms of agrarian and artisan production and social structures; this allows us an alternative explanation of changes of society and productivity at and after the end of the Roman period without any particular need for ‘ethnic’ change. One result of such a reading is the move to an archaeology of settlements that shows far less social gradient than that of the later Roman period, something that could also be argued for the archaeology of the dead. A similar problem of decontextualization affects the study of cemeteries, particularly rural ones, in the northern Gaul of the fifth century. Metalwork, ceramics, and other features such as burial rite or cranial deformation have been read in an ‘ethnic’ way. In addition, the burial of some males with weapons has also been seen as ‘Germanic’. In fact, the tradition of burial with weapons can be detected from what appear to be otherwise Gallo-​Roman burial grounds in northern Gaul from the later fourth century on (cf. Theuws 2009), so we may be seeing further change in the indigenous status-​markers. From the fifth century, more in northern than southern Gaul, there was a quantity of material whose stylistic origins lay outside what had been imperial territory, but it is now recognized that seeing objects as direct ethnic indicators is an approach fraught with imponderables. It is important to recall that such burials and material usually form a minority, sometimes a small minority, of burials in a cemetery—​for instance, in the well-​known site of Frénouville in Normandy (Pilet 1980). In such cemeteries the predominant rite is the unfurnished inhumation, characteristic of the late Roman provincial population, and such material culture as is present is likewise of provincial-​Roman type. The risk is clearly that, by decontextualizing certain types of burial and classes of material, a misleading impression is given of the composition of the burying population. Recent appreciation of the very debatable relationship between object and ‘ethnic’ or other identity compounds the problems. Instead it is possible to read the ‘weapon burials’ not in ethnic terms but in terms of the increased importance of the vocabulary of military power as markers of social status for ‘Romans’ and ‘Germans’ equally. One other point worth making about some of these cemeteries, such as Frénouville, is that they demonstrate considerable stability and longevity, mirroring some of the settlement evidence for continuity from the fourth century into the fifth and sixth, again raising questions as to what extent we are looking at ‘incomers’.

146

146   Simon Esmonde Cleary Fifth-​century settlements and cemeteries in southern Britain have been analysed almost exclusively within an ‘ethnic’ framework, above all as evidence for the arrival of the Anglo-​Saxons. Do the ideas outlined above relating to changes in building technology, settlement types, and the militarization of the elites have anything to contribute this side of the Channel? The shift to timber building has long been acknowledged as a marker of the end of the Roman order as much as, or more than, of the arrival of new peoples. Even the types of structure found in eastern and southern Britain at ‘Anglo-​ Saxon’ sites sometimes have an equivocal relationship with ‘Germanic’ building types, and the Wohnstallhaus is barely represented this side of the North Sea (Hamerow 2002). As with the discussion of comparable settlements in northern Gaul, consideration is now focusing more on questions of the social formation they represent and the agrarian and artisan practices they embody rather than simply ethnic identity. Likewise, the presence of weapons and militaria in graves of this period has been discussed almost exclusively in terms of Anglo-​Saxon social structures (e.g. Böhme 1986). But with the increasing appreciation of the part indigenous British people may have played in populating ‘Anglo-​Saxon’ cemeteries, the presence of such material might also be looked at in terms of the ‘militarization’ of male, particularly elite male, identity more generally. Linked with this is the question of the significance of burials containing ‘Quoit Brooch’ material (Suzuki 2000), with its evidently Roman-​style decoration and significance and therefore its presence in supposedly Anglo-​Saxon cemeteries. This is emphasized by the amount of such material from the continental coasts of the Channel (Soulat 2009), suggesting a trans-​Manche region of prestige metalwork that does not fit well into standard narratives about ‘Anglo-​Saxon’ and ‘Frank’.

Epilogue As stated in the preamble, given this chapter’s place in the volume, its purpose has been briefly to try to set developments in the archaeology of Britain in the context of developments on the continent and to identify those aspects of fourth-​and fifth-​century Britain where trends in the island derived from wider trends within the late empire and thus should be interpreted in that context. Equally, it has tried to identify those ways in which practices in the regions of Britain deviated from those across the Channel, so requiring explanations to be sought closer to home. What I hope this has demonstrated is that late Roman Britain was more integrated with other regions of the west than has often been appreciated. This is not to say that Britain was identical with the continent; the archaeology both of the end of imperial power in the island and of the relationship between the indigenous population and incomers in the fifth century strongly suggests insular particularities, but at least we should, in due course, be better able to separate the insular from the wider processes. One might close by citing the growing evidence, particularly from stable isotope analysis (see the chapters in Eckardt 2010; see also Eckardt, this volume), for the cosmopolitan nature of the population of later Roman Britain, at least in

Britain at the End of Empire    147 its urban centres, with substantial percentages of incomers from the continent in the fourth century, let alone the question of incomers in the fifth. We may also be beginning to see the corresponding development of this, with some evidence identifying fourth-​ century Britons on the continent (Swift 2010; see also chapters by Eckardt and Ivleva, this volume, for further discussion of population movements). Along with the more conventional archaeological evidence considered above, this shows how dangerous it would be to continue to try to consider late Roman Britain in isolation from its neighbours in the west.

Abbreviation RIB  R  . G. Collingwood and R. P. Wright, The Roman Inscriptions of Britain. Volume I: Inscriptions on Stone. Oxford: Clarendon Press, 1965–​to date.

References Adrian, Y.-​M. (2006). ‘Céramiques et verreries des IVe et Ve s. dans la basse vallée de la Seine; Les Exemples de Rouen, Lillebonne, Caudebec-​lès-​Elbeuf, Tourville-​la-​Rivière (Seine-​ Maritime)’, in P. Ouzoulias and P. Van Ossel (eds) Les Céramiques de l’Antiquité tardive en Île-​de-​France et dans le Bassin parisien Volume I. Ensembles régionaux. Diœcesis Galliarum Document de travail 7. Paris: Diœcesis Galliarum, 331–​389. Balmelle, C. (2001). Les Demeures aristocratiques d’Aquitaine: Société et culture de l’Antiquité tardive dans le Sud-​Ouest de la Gaule. Aquitania Supplément 10. Bordeaux:  Fédération Aquitania. Böhme, H. W. (1974). Germanische Grabfunde des 4. bis 5. Jahrhunderts zwischen unterer Elbe und Loire: Studien zur Chronologie und Bevölkerungsgeschichte. Munich: C. H. Beck’sche. Böhme, H. W. (1986). ‘Das Ende der Römerherrschaft in Britannien und die angelsächsische Besiedlung Englands im 5.  Jahrhundert’, Jahrbuch des Römisch-​ Germanischen Zentralmuseums Mainz, 33/​2: 469–​574. Brulet, R., Vilvorder, F., and Delage, R. (2010) (eds). La Céramique romaine en Gaule du nord: La Vaisselle à large diffusion. Turnhout: Brepols. Chavarría Arnau, A. (2007). El Final de las Villae en Hispania, (Siglos IV–​VII d.C.), Bibliothèque le l’Antiquité Tardive. Turnhout: Brepols. Collins, R. (2008). ‘The Latest Roman Coin from Hadrian’s Wall: A Small Fifth-​Century Purse Group’, Britannia, 39: 256–​261. Collins, R. (2012). Hadrian’s Wall and the End of Empire: The Roman Frontier in the 4th and 5th Centuries. London: Routledge. Cunliffe, B. (1976). Excavations at Portchester Castle: Volume II Saxon. Reports of the Research Committee of the Society of Antiquaries of London 23. London: Society of Antiquaries of London. Drack, W. and Fellmann, R. (1988). Die Römer in der Schweiz. Stuttgart: Theiss. Eckardt, H. (2010) (ed.). Roman Diasporas: Archaeological Approaches to Mobility and Diversity in the Roman Empire. Journal of Roman Archaeology Supplementary Series 78. Portsmouth, RI: Journal of Roman Archaeology.

148

148   Simon Esmonde Cleary Esmonde Cleary, S. (1989). The Ending of Roman Britain. London: Batsford. Esmonde Cleary, S. (2013a). The Roman West, ad 200–​ 500:  An Archaeological Study. Cambridge: Cambridge University Press. Esmonde Cleary, S. (2013b). ‘Southern Britain in the Fifth Century:  A  Collapsed State?’ in F. Hunter and K. Painter (eds), Late Roman Silver:  The Traprain Treasure in Context. Edinburgh: Society of Antiquaries of Scotland, 49–​53. Evans, J. (2000). ‘The End of Roman Pottery in the North’, in A. Wilmott and P. Wilson (eds), The Late Roman Transition in the North: Papers from the Roman Archaeology Conference Durham 1999. BAR British Series 299. Oxford: Archaeopress, 39–​46. Foy, D. (1995) (ed.). Le Verre de l’Antiquité tardive et du haut Moyen Âge: Typologie, chronologie, diffusion. Guiry-​en-​Vexin: Association Française pour l’Archéologie du verre. Fulford, M., and Hodder, I. (1974). ‘A Regression Analysis of Some Late Romano-​British Pottery: A Case Study’, Oxoniensia, 39: 26–​33. Guest, P. (2005). The Late Roman Gold and Silver Coins from the Hoxne Treasure. London: British Museum. Hamerow, H. (2002). Early Medieval Settlements: The Archaeology of Rural Communities in North-​West Europe 400–​900. Oxford: Oxford University Press. Handley, M. (2003). Death, Society and Culture: Inscriptions and Epitaphs in Gaul and Spain, AD 300–​750. BAR International Series 1135. Oxford: Archaeopress. Heijmans, M. (2006a). ‘La Place des monuments publics du Haut-​Empire dans les villes de la Gaule méridionale durant l’Antiquité tardive (IVe–​VIe s)’, Gallia, 63: 25–​41. Heijmans, M. (2006b). ‘La Mise en défense de la Gaule méridionale au IVe–​VIe s’, Gallia, 63: 59–​74. Henig, M. (1984). Religion in Roman Britain. London: Batsford. Hobbs, R. (2006). Late Roman Precious Metal Deposits c. ad 200–​700: Changes over Time and Space. BAR International Series 1504. Oxford: Archaeopress. Kent, J. (1994). Roman Imperial Coinage Volume X:  The Divided Empire and the Fall of the Western Parts. London: Spink. Neal, D., and Cosh, S. (200–​10). Roman Mosaics of Britain. 4 vols. London: Society of Antiquaries of London. Oldenstein, J. (1986). ‘Neue Forschungen im spätrömischen Kastell von Alzey: Vorbericht über die Ausgrabungen 1981–​1985’, Bericht der Römisch-​Germanischen Kommission 67: 290–​356. Pearson, A. (2002). The Roman Shore Forts:  Coastal Defences of Southern Britain. Stroud: Tempus. Petts, D. (2003). Christianity in Roman Britain. Stroud: Tempus. Peytremann, E. (2003). Archéologie de l’habitat rural dans le nord de la France du IVe au XIIe siècle, Mémoires de l’Association Française de l’Archéologie Mérovingienne XIII. 2  vols. Condé-​sur-​Noireau: Association Française de l’Archéologie Mérovingienne. Pilet, C. (1980). La Nécropole de Frénouville: Étude d’une population de la fin du IIIe à la fin du VIIe siècle. BAR International Series 83. Oxford: British Archaeological Reports. Pirling, R. (1986). Römer und Franken am Niederrhein. Mainz: Philipp von Zabern. Pirling, R., Siepen, M., Noeske-​Winter, B., and Tegtmeier, U. (2000). Das römisch-​fränkische Gräberfeld von Krefeld-​Gellep 1989–​2000. Stuttgart: Steiner. Rigoir, J. (1968). ‘Les Sigillées paléochrétiennes grises et oranges’, Gallia, 26: 177–​243. Soulat, J. (2009). Le Matériel archéologique de type saxon et Anglo-​Saxon en Gaule Mérovingienne. Mémoires publiés par l’Association Française de l’Archéologie Mérovingienne XX. St-​ Germain-​en-​Laye: Association Française de l’Archéologie Mérovingienne.

Britain at the End of Empire    149 Suzuki, S. (2000). The Quoit Brooch Style and Anglo-​Saxon Settlement. Woodbridge: Boydell. Swift, E. (2000). Regionality in Dress Accessories in the Late Roman West. Monographies Instrumentum 11. Montagnac: Éditions Monique Mergoil. Swift, E. (2010). ‘Identifying Migrant Communities:  A  Contextual Analysis of Grave Assemblages from Continental Late Roman Cemeteries’, Britannia, 41: 237–​282. Theuws, F. (2009). ‘Grave Goods, Ethnicity and the Rhetoric of Burial Rites in Late Antique Northern Gaul’, in T. Derks and N. Roymans (eds), Ethnic Constructs in Antiquity: The Role of Power and Tradition. Amsterdam Archaeological Studies 13. Amsterdam:  Amsterdam University Press, 293–​319. Van Ossel, P. (1992). Établissements ruraux de l’Antiquité tardive dans le nord de la Gaule, (51e Supplément à Gallia). Paris: Centre National de la Recherche Scientifique. Van Ossel, P. (2003). ‘De la villa au village: Les Prémices d’une mutation’, in J.-​M. Yante and A.-​M. Bultot-​Verleysen (eds), Autour du ‘village’: Établissements humains, finages et communautés rurales entre Seine et Rhin (IVe–​XIIIe siècles) Actes du colloque international de Louvain-​la-​Neuve, 16–​17 mai 2003. Turnhout: Brepols, 1–​19. Van Ossel, P. (2010). ‘Transformations et continuités aux contacts de trois cultures:  Les Campagnes de la Gaule septentrionale au Ve siècle’, in P. Delogu and S. Gasparri (eds), Le trasformazioni del V secolo:  L’Italia, I  barbari e ‘Occidente romano, Seminari del Centro Iteruniversitario per la storia e l’archeologia dell’alto medioevo 2.  Turnhout:  Brepols, 579–​600. Van Ossel, P. (2011). ‘Les Sigillées du groupe Argonne dans le Bassin parisien au Bas-​ Empire: Caractérisation, production et diffusion’, in P. Van Ossel (ed.) Les Céramiques de l’Antiquité tardive en Île-​de-​France et dans le Bassin parisien. Volume II. Synthèses. Diœcesis Galliarum Document de travail no. 9. Paris: Diœcesis Galliarum, 231–​254. Whittaker, C. (1994). Frontiers of the Roman Empire: A Social and Economic Survey. Baltimore and London: Johns Hopkins University Press. Wickham C. (2005). Framing the Early Middle Ages: Europe and the Mediterranean 400–​800. Oxford: Oxford University Press. Woodward, A. (1992). Shrines and Sacrifice. London: Batsford/​English Heritage.

150

Chapter 8

Britain be fore the Roma ns Timothy Champion

Introduction In 1849 the historian Thomas Macaulay (1849: i. 4) wrote that the inhabitants of ancient Britain were ‘little superior to the natives of the Sandwich Islands’ and ‘received only a faint tincture of Roman arts and letters’. Other authors in this volume will debate whether ‘faint tincture’ is an appropriate description of what happened after the conquest. Here I would like to consider the state of Britain immediately prior to that event. Macaulay undoubtedly intended to be uncomplimentary, but he may have made a serious misjudgement. The Sandwich Islands was the name given by Captain Cook and still in use in the 1840s for the islands now known collectively as Hawaii, which have become a classic case study of the development of a complex chiefdom society in archaeological and anthropological textbooks (Kirch 2010, 2012). At the time of Cook’s first visit in 1778 the islands were characterized by dense population, efficient irrigation and rain-​fed agriculture, elite control of surplus production, divine kingship, and wealth finance based on prestige goods. In the 120 years between Cook’s first visit and the annexation of the islands by the empire of the United States in 1898, they experienced growing contact with the European world through military and naval expeditions, politicians, traders, arms dealers, and missionaries; under such external military, economic, and religious influence from the USA, Britain, and Russia, they saw the emergence of a unified political structure, largely through external support for one particular chief, new economic structures, especially for exports, as well as new forms of architecture, religion, language, and literacy (Kirch and Sahlins 1992). The parallel with the situation in Britain in the century between Caesar and Claudius is not exact, but it serves as a clear and well-​documented example of how a society can be transformed by contact with very different cultural groups even before formal conquest and annexation.

Britain before the Romans    151 Macaulay had little sympathy for pre-​Roman Britain and certainly no notion of significant changes in the centuries before the conquest. Archaeology developed rapidly in the following decades, and by the early twentieth century an orthodoxy had emerged. Sir John Evans’ Coins of the Ancient Britons (1864), a work that laid the foundations for all future numismatic study of the Iron Age (de Jersey 2008), demonstrated clearly that coins had been minted from at least the middle of the second century bc, and that some of them showed evidence of a knowledge of Latin and the existence of the institution of kingship. When Evans and his son Arthur visited a sand quarry at Aylesford in Kent in 1888, they found evidence for cremation burials of the Late Iron Age (LIA), one of which contained an imported Italian bronze jug and pan. Arthur Evans’ eventual publication (1890: 386) of the finds traced the continental origins of LIA culture in Britain, and attributed these innovations to an ‘invasion’ by ‘an intrusive Gaulish tribe’. The cremation tradition was confirmed by the excavation of a small cemetery at Swarling in Kent (Bushe-​Fox 1925), while connections with the classical world were again demonstrated by the discovery of two rich graves at Welwyn, Hertfordshire, which contained not only Roman bronze vessels and amphorae, but also a pair of Augustan silver cups (Smith 1912). Excavations in the 1930s at major sites such as St Albans (Wheeler and Wheeler 1936) and Colchester (Hawkes and Hull 1947) clearly demonstrated the Iron Age precursors of some important Roman towns and the extent of contact with the Roman world, seen especially in the presence of imported pre-​conquest pottery. By that time the cross-​ Channel connections had been explored in detail by Hawkes and Dunning (1930), and Hawkes (1931) had published the original version of his influential tripartite scheme for the British Iron Age, later to be revised and embellished (1959); in this scheme the final stage, Iron Age C, was represented by two waves of immigration from the continent, the first into Kent and Essex and the second further west, into Hampshire, Berkshire, and Wiltshire. Though the so-​called second Belgic invasion in the west never won much support, the idea of an Iron Age C culture centred on Kent, Essex, and Hertfordshire, brought by immigrants from Belgium or northern France, remained the prevailing theory for the explanation of LIA innovations in burial practice, pottery, coinage, settlement, and social organization. Behind these upheavals in north-​western Europe was the expansion of the Roman Empire, first into southern France in the late second century bc, followed in the 50s by Caesar’s conquest of Gaul. The impact of Rome in the post-​Caesarian period was now clear: ‘Romanization had struck deep roots before the Claudian conquest’ (Kendrick and Hawkes 1932: 207). Much of this orthodoxy is now questioned, if not yet actually discarded: the vision of Britain as affected only by externally generated factors, the role of immigration, the geographical extent of major social changes in the later Iron Age, the chronology and pace of social change, the nature of social organization immediately prior to the conquest, and Romanization whether before or after Claudius. Some of this revision has been due to the explosion of archaeological evidence since the 1990s, some to changing modes of interpretation. The changing modes of interpreting the evidence for the LIA will be reviewed in more detail at the end of this chapter. There is now a growing literature of critical assessment

152

152   Timothy Champion (e.g. Gardner 2013) and historiographical review (e.g. Hingley 2000, 2008; Hoselitz 2007) of the treatment of developments both before and after the conquest, but these studies have been conducted mainly by specialists in Roman archaeology and have focused on the dominant Roman side of the clash. By contrast, little (e.g. Webster 1996) has been written by scholars specializing in the Iron Age or from the point of view of the subaltern British. One notable exception to this is the topic of the Druids, who have been the focus for several recent studies (Piggott 1985; Webster 1999; Hutton 2007, 2009; Aldhouse-​Green 2010). Smiles’ scholarly study (1994) of the ancient Britons ranges far more widely than the ‘images’ of the ‘romantic imagination’ that its title suggests, but its chronological range stops before the rise of archaeology. We still lack a thorough historiographical review of the study of the Iron Age.

The Evidence With Caesar’s narrative of his cross-​Channel adventures in 55 and 54 bc, Britain enters the written record of the classical world. His account (De Bello Gallico IV. 2–​36; V. 8–​23) is fairly brief and not without problems, but it is fuller and more detailed than the patchy record that survives for the following century before the Claudian invasion. This evidence, primarily short passages in Strabo and Diodorus Siculus and even briefer references in Suetonius and Dio Cassius, as well as some enigmatic references in the Roman poets and one incomplete paragraph in Augustus’ Res Gestae, has been reviewed by several authors (e.g. Braund 1996). There have been no additions to this corpus of sources, though new editions, especially of the Res Gestae (Cooley 2009), have clarified some issues. More important has been Woolf ’s exploration (2011) of the cultural and political context in which knowledge of the west was created by the expanding Roman Empire. By contrast, the archaeological evidence for Iron Age Britain has increased dramatically since the last decades of the twentieth century (Haselgrove and Moore 2007a; Haselgrove and Pope 2007). Some parts of this evidence will be discussed in detail later in this chapter, but particularly noteworthy has been work on some of the major centres of the LIA, such as Silchester (Fulford and Timby 2000) and Canterbury (Blockley et al. 1995), though the organization of the landscape and the pattern of ordinary rural settlement remain poorly understood. Knowledge of the burial tradition in south-​eastern England, mainly cremations, has been advanced by excavation of important cemeteries (Parfitt 1995; Fitzpatrick 1997), warrior burials (Stevenson 2013), and elite cremations (Niblett 1999; Crummy 2007). The Portable Antiquities Scheme has led to the recording of a vast number of new finds of metalwork; the significance of these finds has not been fully explored, though it is clear that there is great regional variation in the types of material deposited in the Iron Age (Worrell 2007) and that detailed analysis of the finds from a specific region can produce important evidence for the circulation and deposition of coins and other metal objects (Hutcheson 2004; Farley 2012). There has been particular progress in the study of coinage: as well as more general reviews (Haselgrove

Britain before the Romans    153 1987; Creighton 2000) and catalogues (van Arsdell 1989; Hobbs 1996; Rudd et al. 2010), there have been important studies of some of the regional series, including the northeastern (Farley 2012), the East Anglian (Chadburn 2006), the south-​eastern or Kentish (Holman 2000, 2005), the South Thames (Bean 2000), and the western (van Arsdell 1994). The following sections will review how the new evidence and changing modes of interpretation have altered perceptions of the later Iron Age, with particular attention to southern and eastern England, where the first effects of contact with Rome were felt most strongly. There were important developments in the preceding centuries, which drew parts of Britain into a network of relationships with their continental neighbours, so that, when the Roman Empire eventually extended into northern France, it was more a case of transforming existing power structures than establishing completely new ones.

Changing Perceptions of the Iron Age Since Arthur Evans’ publication (1890) of the rich LIA graves from Aylesford in Kent, it has been recognized that there were significant cultural changes in the century or two immediately preceding the Roman conquest. A distinctive late Iron Age as the culmination of the insular Iron Age sequence has been identified predominantly in the south-​east of England, though the idea has also found wider application in southern and eastern England. More recently, however, archaeologists working in regions other than the south-​east have tended to prefer a simpler division into earlier and later Iron Age (Haselgrove et al. 2001; Moore 2006: 214–​224; Haselgrove and Moore 2007b: 1–​5), with the dividing point somewhere around 40–​300 bc, suggesting that the changes seen in the south-​east were not reflected elsewhere, but also that internal processes of change were important over a much longer period. This vision of Iron Age developments has not gone completely unchallenged, and the reality of widespread changes in the first century bc has been reasserted; instead of steady social evolution, a pattern of punctuated equilibrium, with long periods of comparatively static social organization interrupted by brief episodes of change, has also been suggested (Barrett et al. 2011). The question of the nature of Iron Age social organization has been the topic of recent debate, though much of the discussion has revolved around the specific issue of the extent to which society was hierarchically organized, especially in the earlier parts of the Iron Age. In much early writing about the Iron Age there is little, if any, explicit discussion of the nature of social organization or social relations, though there is an implicit concept of a hierarchically organized society, presumably derived from the authority of the classical references to chiefs and kings in the LIA, projected backwards into earlier periods. Pictorial representations of ancient Britons in the eighteenth and nineteenth centuries were also dominated by warriors and chiefs, as well as Druids (Smiles 1994). More explicit discussion of the nature of Iron Age society began to appear in the latter part of the twentieth century. One influential model was derived from Cunliffe’s

154

154   Timothy Champion discussion of his excavations at the hillfort of Danebury. Early attempts at explaining its social role (Cunliffe 1984: 549–​562) included a comparison of Iron Age society with the better-​documented historical societies of early medieval Ireland, underpinned by an implicit assumption of widespread and long-​term homogeneity in the nature of ‘Celtic’ society; a concept of the hillfort as the possible residence of kings or nobles; a vision of the role of the hillfort as an emerging centre of regional power and of redistribution, drawing on neo-​evolutionary ideas of the role of a chiefdom; and the progression from hillfort to developed hillfort to oppidum as the material expression in settlement terms of the social evolutionary trajectory from tribe to chiefdom to state (Cunliffe 1976). These interpretations have been the subject of sustained critique (e.g. Hill 1995, 1996), as much on the grounds of the archaeological evidence as from a theoretical consideration of the appropriateness of the analogies. Various attempts have been made to draw on other models of pre-​industrial social organization: Crumley’s idea (1995) of ‘heterarchy’, in which power is derived from a variety of different structures within society, has attracted some writers, such as Armit (2007) and Cripps (2007); others, looking to structural Marxist anthropology, have explored the concept of the Germanic Mode of Production (e.g. Hingley 1984; Hill 1995). In an attempt to move beyond the concepts of neo-​evolutionary anthropology and structural Marxism, Sharples (2010: 292–​294) has adopted Mary Douglas’ grid and group matrix for social analysis. Hill (2011) has explored non-​hierarchical structures for Iron Age society, citing the concept of segmentary societies (Fortes and Evans-​Pritchard 1940). He also analysed some of the possibilities for exploring elements of social organization such as households, communities, and tribes, and, as the focus of attention has moved on from the debate about hierarchical or non-​hierarchical social organization in the earlier periods of the Iron Age, so these themes have been receiving more attention. Much of this recent critique has been aimed at the concept of a hierarchical structure in earlier Iron Age societies, and the attempt to find ways of discussing other forms of society in which there may have been leaders rather than rulers. There has been surprisingly little reconsideration of social organization in the final centuries of the Iron Age, especially in the south-​east, when the classical authors refer to chiefs, military leaders, and kings, supported by the evidence of some of the inscribed coinage, considered further below. Hill (2007: 30) refers to kingship as ‘a novel experiment’ in the LIA of south-​ eastern England, but there has been little detailed consideration of the extent or the basis of such ‘royal’ power. Caesar regularly describes the political groupings of southern England in the mid-​first century bc by using the word civitates, usually translated as ‘tribes’, a term that Moore (2011) has argued to be so ideologically laden as to be unhelpful and even misleading. It has also been shown that the distributions of the regional series of coins in southern and eastern England, formerly attributed to such tribes and thought to mark their distinct territories (Williams 2003, 2005a), are much more complex, revealing patterns at a variety of scales and making such an attribution very problematic (Leins 2008). Like the nature of political authority, the nature of political groupings in the LIA has become much more difficult to discern, and almost certainly much more varied; at the very least, projecting post-conquest structures back into prehistory is unwise.

Britain before the Romans    155

Settlement and Economy There is clear evidence for close contact between Britain and the continent throughout the Iron Age (Webley 2015), though the precise nature of the social relationships represented by the material is often unclear. Much writing about the Iron Age has been based on assumptions that such contacts declined after the end of the Bronze Age, that interaction across the North Sea and the English Channel was intermittent until the LIA, that these waters represented some form of cultural boundary, and that cultural innovations began on the continent and crossed to Britain. All of these propositions should now be regarded as doubtful. Britain shared a number of significant cultural practices with the near continent at a variety of scales, such as an architectural tradition of predominantly round rather than rectangular structures, as well as the production and use of utilitarian artefacts such as triangular clay loomweights (Champion 1975; Wilhelmi 1977, 1987), bone weaving combs (Tuohy 1999), or, at a more localized scale, pottery (Leman-​Delerive 1984; Hurtrelle et al. 1990; Blancquaert and Bostyn 1998). In such cases, the items form a unified cultural zone in which the Channel and North Sea did not constitute any form of cultural boundary; it is more appropriate to talk of regions of shared cultural practice than of points of innovation and areas of expansion or diffusion. In the case of other items, such as weaponry or art, discussed further later in the chapter, it may make more sense to think of a point or area of origin, and diffusion from there. The direction of such diffusion, however, was not necessarily always in a north-​westerly direction from the continent to Britain: some of the earliest evidence for rotary querns currently available in Europe comes from southern England (Peacock and Cutler 2011; Wefers 2011). A clear distinction between indigenous processes of change and externally generated innovation is therefore problematic. Indications of significant change in Iron Age economy and society can be seen from perhaps 400 or 300 bc. The patterns of change vary considerably in chronology and are highly variable regionally, being especially marked in southern and eastern England. One obvious area of innovation is in the production and usage of material artefacts. The sheer volume of such material items increased greatly, and there were important developments in the scale and specialization of production. Pottery manufacture was, at least in some areas, increasingly in the hands of more specialized producers, especially for finer wares, and their products were more standardized (Morris 1994, 1996). Quern production also became more centralized, probably through the work of specialist producers working from major quarry sites such as Lodsworth (Peacock 1987). Production of iron also increased, with the emergence of some major centres of smelting that distributed their iron for smithing in characteristically shaped bars of standardized size and weight, the so-​called currency bars (Allen 1967; Hingley 1990, 2005; Crew 1994). Heavy stone weights are known from a number of sites in southern England—​for example, at Danebury (Cunliffe and Poole 1991: 383); though their function is uncertain, it is likely

156

156   Timothy Champion that they were used for weighing out heavy or bulky items, suggesting the exchange of agricultural produce. More specialized production would have been accompanied by more complex relationships for the distribution and acquisition of products, and control over these processes of production and distribution would have provided new opportunities for the accumulation of wealth and status. There are also developments in the nature of settlement sites, though their relationship to other changes is not always clear, and there is great regional variation. In the regions of Wessex with a high concentration of hillforts, many sites were abandoned, while a few were enlarged or elaborated to form the ‘developed hillforts’ in Cunliffe’s terminology (2005: 388–​396). The term indicates a suggested role as emerging centres of regional power and economic organization, but they may alternatively be the result of a process of nucleation of a scattered population (Davis 2013). There are different histories of hillfort construction and usage in other areas: in Sussex, hillforts were built in a variety of locations and for a variety of reasons from the late Bronze Age onwards, while in Surrey and Kent the first forts were being built around the third century (Hamilton and Manley 2000). In several regions, recent research has revealed a switch from open to enclosed settlement forms (Knight 2007; Moore 2007). As well as changes in settlement types, there is growing evidence for fluctuating densities of population; areas previously densely occupied show much less evidence of activity, or less intensive usage, as activity moved into new zones. New types of sites also began to appear, especially from the second century onwards. There is considerable variation in location, size, form, and function, as well as chronology of foundation. Some of these sites have been grouped together and classed as oppida; the term, originally applied by Caesar to some of the major defended sites in France, is applied in continental, especially German, archaeology to some very large nucleated sites of the LIA, but its usage in Britain is ‘more confused’ (Collis 1984: 6). Cunliffe (1976) has suggested a typology of oppida: his scheme includes enclosed, undefended, and territorial oppida, seen as types of nucleated centres that are different from the earlier hillforts. Nevertheless, the range of sites included in this category makes it highly problematic: some are indistinguishable from hillforts, while others that are in many ways similar to oppida are excluded. It is doubtful whether the term is of much use for understanding the actual processes of settlement development: Woolf ’s critique (1993a) of the term in continental archaeology could apply equally to its use in Britain (Bryant 2007; Pitts 2010: 34–​35). The enclosed oppida play an important role in Cunliffe’s model of social evolution (2005: 402–​406), replacing the developed hillforts as the settlement evidence for an enhanced level of social organization and political centralization, and representing settlements that were urban or of an urban character, while the territorial oppida are seen as the centres of LIA power and authority. The fact that some of these sites later became the location for Roman towns may indicate that they were places of importance in the pre-​conquest period, but cannot be used to indicate their function at an earlier date; the Roman Empire had very different structures of administrative organization. One important development was the appearance of major sites on the coast, with a significant role in cross-​Channel exchange. By the second half of the second century,

Britain before the Romans    157 links between south-​western England and north-​western France, though perhaps never totally absent, become more visible (Cunliffe and de Jersey 1997), articulated through ports such as Hengistbury Head in Dorset (Cunliffe 1987) and Mount Batten near Plymouth in Devon (Cunliffe 1988); though imported items such as coins, fine pottery, and Roman amphorae, discussed further later in this chapter, did percolate into the hinterlands, concentrations of such items at the ports suggest that they may have been places with a distinct culture. Somewhat later, exchange between the continent and south-​eastern England became more prominent; sites on the east coast of Kent, known mostly from surface coin collections and small-​scale excavation (Holman 2005), may have played a major role in this connection, a process beginning before the end of the second century. Elsewhere, there are signs of significant nucleation of settlement, though the chronology is often uncertain. In eastern England, occupation at large sites such as Dragonby (May 1996) and Sleaford (Elsdon 1997) had begun before the first century bc. At Winchester a large rectangular enclosure had been occupied in the second and first centuries (Qualmann and Whinney 2004). At Baldock, Hertfordshire, the main occupation had begun ‘at least by the middle of the first century bc’ (Stead and Rigby 1986: 84), and at Braughing, Hertfordshire (Partridge 1981; Potter and Trow 1988) and Heybridge, Essex (Atkinson and Preston 1998) probably soon after that; at other sites, however, such as Dyke Hills, Dorchester (Hingley and Miles 1984: 65 and fig. 4.9) and Abingdon (Lambrick 2009: 362), both in Oxfordshire, the chronology of occupation is obscure, though probably in the second or first centuries bc. These processes had clearly begun in at least some regions by the second century, but were intensified in the first, especially in the years after 50 bc.

Wealth, Power, and Ritual As well as the similarities in architecture, pottery, and some artefact types noted above, Britain also shared in more widespread fashions for prestigious or symbolic items (Joy 2015). There are very few objects that could be actual imports, but indigenous British imitations and developments of continental styles suggest widespread familiarity with continental practice, whether in weaponry such as daggers (Jope 1961) or swords (Stead 2006), in art (Jope 2000), or in bodily adornment and clothes fastening, with the brooch replacing the pin (Adams 2014). From the second century onwards, however, there are significant changes in the evidence. Perhaps the most important of these is the reappearance of gold in the archaeological record, especially in the form of coinage. Gold had been comparatively plentiful in the late Bronze Age, but after about 800 bc it disappears from the archaeological record in much of Britain, though recent metal-​detector finds suggest it may not have been totally absent. In many parts of the central and western regions of continental Europe gold became rarer, but continued to be deposited primarily in the richly furnished

158

158   Timothy Champion graves of the elite in the form of personal ornaments or occasional vessels (Eluère 1987). In the fourth and third centuries large quantities of gold flowed northwards from the Mediterranean world as a result of barbarian raiding and the role of Celtic warriors as the favoured mercenaries in the Hellenistic world. This influx of wealth played a major role in the restructuring of social relationships in the barbarian world and the creation of a new set of usages for wealth; gold was preferred in the west, though silver was more common in some areas further east. From the late fourth century gold appears less and less frequently in graves and more regularly as hoards or votive deposits, most often in the form of coinage, though neck-​rings, or torques, of gold are also found, sometimes in association with coins (Furger-​Gunti 1982; Fitzpatrick 2005). Celtic coins were being minted from the early third century, copying Macedonian and other Mediterranean prototypes. Gold was now used not for lavish display in the form of ornaments and vessels, though torques may have been worn as a symbol of male prowess, but more commonly in the form of coins. The earliest coins in western Europe were of precious metal, mostly gold, and were of high-​value denominations. They were not used for everyday commercial exchange, but could have been used as a medium for accumulating, assessing, and storing wealth, as well as being ideally suited to use in the transfer of wealth. They are the material representation of transactions involved in a limited set of social relationships (Allen 1976; Nash 1987); these might have included payment for services rendered or tribute due to a superior, dowry or bride price, gift exchange, and other acts of generosity, including the deposition of votive offerings, which is the predominant archaeological context for their discovery. Polybius, writing in the second century, describes Celtic society thus: Their possessions consisted of cattle and gold, since these were the only objects which they could easily take with them whatever their circumstances and transport wherever they chose. It was of the greatest importance to them to have a following, and the man who was believed to have the greatest number of dependents and companions about him was the most feared and most powerful member of the tribe. (Hist. II. 17, trans. I. Scott-​Kilvert)

At a later date, Caesar provides good evidence for the existence of such patron–​client relationships in Gaul, especially in his account of the internal politics of the Helvetii (De Bello Gallico I. 4): Orgetorix escaped trial when he turned up in court with his 10,000 followers. Some powerful men maintained their own permanent retinue of armed and mounted men, such as Dumnorix of the Aedui (De Bello Gallico I. 18) or Commius of the Atrebates (De Bello Gallico VIII. 23). Internal politics was dominated by factional rivalry within the polities, often between brothers or other close male relatives in a ruling lineage, such as the brothers Dumnorix and Diviciacus of the Aedui (De Bello Gallico I. 18), and by alliances between factions within different polities, such as that between Orgetorix, Dumnorix, and Casticus of the Sequani (De Bello Gallico I. 3). Such patron–​client relationships and inter-​tribal alliances would have been maintained by a variety of transactions, possibly including dynastic marriage alliances and the exchange

Britain before the Romans    159 of gifts; in extreme cases, where alliances were called on to provide support in warfare, large quantities of gold would have been needed. In these circumstances, coinage would have provided a useful means for the new uses to which wealth could be put, especially for its transfer. Parts of eastern and southern Britain were clearly drawn into this network of social relationships and wealth transfer, and into shared practices in the circulation and deposition of wealth. Continental coins began to appear in Britain perhaps before 200 bc, and more frequently in the first half of the following century. Some of the earliest coin types, found on both sides of the English Channel, are usually thought to have been struck in France, but it is not impossible that some were already being produced in England by the middle of the second century. Local British coin series with imagery derived from the Gaulish prototypes were being produced in southern and eastern England from the second half of the second century, but some of the earliest coins certainly produced locally were of potin, a high-​tin alloy of copper, which were cast rather than being struck; the alloy and the production technology as well as the designs were copied from coin issues of the Marseilles region (Haselgrove 1988, 1995, 2006). Though not made of precious metal, they were hoarded and deposited in the same way as gold, and seem to have been used and valued equivalently. Gold also circulated and was deposited in the form of torques, especially in parts of East Anglia, as at Snettisham (Stead 1991; Hutcheson 2003, 2004), though individual items have been found much more widely throughout the country. As well as the precious metals, other items began to appear in southern England. Among the most significant are objects ultimately from the Roman world, transmitted through late La Tène Gaul. Wine from Italy came in amphorae of Dressel 1 type: the earlier form, Dressel 1A, arriving from around 125 bc, is found in southern central England, while the later form, Dressel 1B, dating from around 8–​15 bc, is more clustered in the south-​east (Peacock 1971, 1984; Fitzpatrick 1985; Carver 2001). Bronze vessels from Campania (Boube 1991; Feugère and de Marinis 1991), though difficult to date precisely but quite possibly pre-​Caesarian, were also found widely in late La Tène Europe (Werner 1954), and reached Britain, as shown by the jugs from Welwyn (Smith 1912) and the jug and pan from Aylesford (Evans 1890). Britain also adopted other practices from the late La Tène world, including new styles of clothes fastening (Haselgrove 1997: 56–​57); from c. 120 bc brooches became much more common, and new continental forms replaced indigenous ones. An important technological innovation was the first production of wheel-​thrown pottery in some parts of the south-​east of England, though much pottery was still hand-​made (Hill 2002). The new technology was accompanied by new forms and a wider range of more specialized vessel shapes, with close similarities to those in use in northern Gaul (Hawkes and Dunning 1930; Tyers 1980). The new ceramic repertoire suggests new modes of preparing, serving, and consuming food and drink, as well as perhaps a new social setting and significance for these activities. One context where wheel-​thrown pottery is found was in cremation burials, a new rite also adopted from the continent (Collis 1977). The rite, often called the Aylesford

160

160   Timothy Champion or Aylesford–​Swarling tradition after two of the earliest sites excavated (Evans 1890; Bushe-​Fox 1925), is found most frequently in Kent, Essex, and Hertfordshire, though it extends beyond this region. It involved the burial of a portion of the cremated remains of the deceased in a pot (Fitzpatrick 2000), though occasional inhumation burials are known, as well as regional variation in the details of deposition (Fitzpatrick 1997; Hill et al. 1999). The absence of other artefacts in most of the graves makes precise dating difficult, but a small number of burials contain brooches, and the earliest such associations would now be dated to the first half of the first century bc (Stead 1976; Fitzpatrick 1997: 95–​96) and the practice may have begun even before 100 bc. Among these burials are a small number that are clearly much more richly furnished than the rest. As originally identified (Stead 1967), this group was characterized by the presence of a large rectangular burial chamber, and the inclusion of pottery, one or more imported Roman amphorae and usually some glass or metal vessels, and was distinct from another series of burials that contained metal-​bound wooden buckets (Stead 1971). Subsequently discovered burials (e.g. Dorton:  Farley 1983; Baldock:  Stead and Rigby 1986) blurred the distinction between the groups and widened the range of grave goods that occurred. Attempts to group the burials into categories defined on the basis of wealth (Haselgrove 1982, 1984) have not been wholly persuasive. Though there is considerable variation in the actual selection of goods for deposition in the graves, there are predominant themes of feasting and drinking, with a mixture of indigenous items such as buckets, cauldrons, and firedogs with exotics such as a bear-​skin rug or cloak, and imports from the Roman world, including silver cups, bronze jugs, and pans, as well as amphorae and other pottery vessels. Weaponry is notably absent, as are items decorated with so-​called Celtic art; such decoration was most common in the second and the first half of the first centuries bc, and there are few dates after about 50 bc, before a revival in the middle of the first century ad (Garrow et al. 2009). A very small number of graves are distinguished by their extreme wealth and a different burial rite, best documented at Folly Lane, St Albans (Niblett 1999), but also seen at Lexden (Foster 1986) and Stanway (Crummy 2007), both at Colchester: the deceased person was laid out in a sunken burial chamber, with a rich array of grave goods; the grave goods were then smashed, and some placed on the pyre together with the body and some animal parts; the chamber was demolished and the shaft backfilled, while a portion of the cremated remains was placed in a separate burial pit. The ritual can be paralleled in some of the richest graves of northern France (Niblett 1999: 394–​404) and must represent the appropriate funerary treatment for some of the members of the most important lineages in LIA Britain. The selection of grave goods is also different from the feasting-​related emphasis of the other rich graves: Folly Lane contained local and Roman horse harness, a tunic of iron chain mail, bronze and ivory furniture fittings, and many other bronze and silver items melted beyond recognition, while Lexden also included chain mail, furniture including a folding stool, and a denarius of Augustus. The adoption of new burial rites closely modelled on those prevailing in northern France would have served to underline the social and political relationships that extended across the Channel, and to express social differences in a very visible way;

Britain before the Romans    161 they may also have helped to establish new relationships if the settlement evidence is rightly interpreted as indicating significant shifts in population density and new forms of landscape organization. Burial, however, was only one way in which social pre-​eminence could be expressed, and the dominant theme of feasting is echoed in non-​funerary contexts, suggesting its significance in LIA social relations. Feasting may have been an important feature of earlier periods in the first millennium, as at late Bronze Age ringworks (Champion 2014: 291–​292) or early Iron Age hillforts (Jones 2007), based on evidence for the storage and preparation of food. Joy (2014) has also argued that cauldrons were used for the preparation of food or drink for feasts throughout the period. Much has been made of the significance of imported wine in LIA Britain (e.g. Carver 2001), but the cauldrons, firedogs, and buckets suggest an indigenous element to feasting rituals. In a survey of possible evidence for feasting in LIA East Anglia, Ralph (2007: 108) commented on the lack of evidence for drinking; but the absence of amphorae on such sites does not mean an absence of drink, only an absence of Roman drink. Pitts (2005) has argued that, in the fifty years before the Roman conquest, the regular presence of large drinking vessels, such as butt beakers, in ceramic assemblages implies communal drinking of beer rather than wine. Whatever was consumed, feasts played a major role in many societies, and the role of feasting may have changed greatly over time (Hayden 2014). Dietler (1998, 2001) has suggested several different categories of feasting that might be appropriate for LIA Britain; Fitzpatrick (2009: 397) has argued that the rich burials of the period represent the burials of the organizers of feasts of his ‘patron-​ role’ type, acts of conspicuous entertainment designed to attract and retain a following of clients, though Ralph (2007) has suggested a wider range of social contexts. The Greek grammarian Athenaeus, writing about ad 200 (Deipnosophistae IV. 37), quotes Posidinius’ much earlier description of the Arvernian nobleman Louernios, who rode in a chariot distributing gold and silver, and prepared a feast with a huge quantity of food and drink, to win popular support. This anecdote encapsulates some of the major themes of LIA society, if social organization was broadly similar in central France and in Britain: social difference, patron–​client relationships, circulation of wealth in fulfilment of these relationships, feasting. One other practice involving many of these elements was the ritual deposition of wealth. Coins, whether found singly or in hoards, as well as other wealthy items such as the gold jewellery in the Winchester hoard (Hill et al. 2004), are very rarely discovered in settlement sites. More often they are from isolated findspots where little is known about their context, though those places may have been significant in the Iron Age landscape; nor, of course, is there any evidence of the people who carried out the deposition or the rites associated with the act. Occasionally, it is possible to see that there were sacred places or even shrines, as at Snettisham in Norfolk (Stead 1991). At Hallaton, Leicestershire (Score 2011), a hill-​top shrine of the conquest period was the site of multiple deposits including over 5,000 coins and a Roman silver-​covered cavalry helmet; large quantities of pig bones suggest feasting on the site. At a few of these sacred sites a structure interpreted as a temple was built (Haselgrove 2005). The best documented is at Hayling Island, Hampshire, where an Iron Age predecessor underlay a

162

162   Timothy Champion later Romano-​Celtic temple and was the focus for deposits of coins, brooches, and other ornaments, horse gear and chariot parts, as well as spearheads and chain mail (King and Soffe 1994, 2008). One other frequent context for the ritual deposition of artefacts was water, continuing a tradition with a long history in Britain (Bradley 1998): in eastern England from the Thames northwards the predominant context for Iron Age swords is rivers such as the Thames and the Trent (Stead 2006), while the Thames was also the context for items such as the Battersea shield (Stead 1985). Springs and ponds may also have been the location for deposits. Ritual activity, including deposition of coins, had begun at the source of the Ebbsfleet river in Kent before the conquest, foreshadowing the later religious complex at Springhead (Andrews et al. 2011). By the middle of the first century bc, therefore, large parts of Britain had been incorporated into the west European network of wealth circulation and deposition, and a smaller part of south-​eastern England had started to produce its own wealth items, with some regional variation in preference for coins or torques, gold or bronze. An even smaller region had begun to bury its dead according to continental practices, with some marked evidence of social differences. Quite how this had happened is still in need of detailed explanation, but it should be borne in mind, as noted above, that close relationships across the North Sea and English Channel had existed earlier in the Iron Age, so that it was more a case of transforming existing links than creating new ones. Though the historical record is minimal, Caesar’s account of Britain suggests that, at least in south-​eastern England, similar political structures may have existed to those he had encountered in Gaul and that political alliances or dependencies may have linked Britain and Gaul. An enigmatic reference to Diviciacus of the Suessiones (De Bello Gallico II. 4), who is said to have exerted some form of political authority across the water into England, suggests the possibility of political alliances, whether between partners of nominally equal status and authority or between those in a relationship of dominance and dependence. It is also clear that such alliances could be called on to provide military support; again Caesar provides the evidence, with references to support from Britain for his opponents in almost all his battles in Gaul (De Bello Gallico IV. 20). The leaders of the Bellovaci fled to Britain, suggesting some form of real or potential alliance (De Bello Gallico IV. 20). The numerous coins of Gallo-​Belgic E type appearing in Britain in the Caesarian period are generally interpreted as payment for British involvement in the opposition to Rome. This represents not so much an unusual event as an extreme example of the operation of social and political relationships through which people and goods regularly crossed the Channel.

From Caesar to Claudius: Britain and Rome in the Final Iron Age When Caesar first set foot in Britain in the late summer of 55 bc, it was Rome’s first direct encounter with Britain, but perhaps not with Britons. Rome had entered into a

Britain before the Romans    163 world where there was an existing network of frequently shifting political relationships, whether alliances or antagonisms, and proceeded to play its part as a powerful new force, with a mixture of diplomatic and military strategies. By the time of Caesar’s second invasion in 54 bc, Rome was seen as a potential ally in the political rivalries of south-​eastern England; Cassivellaunus had killed the king of the Trinovantes and Mandubracius, the son of the dead king, had crossed the Channel to seek Caesar’s support (De Bello Gallico V. 20). Caesar’s expeditions across the Channel undoubtedly boosted his reputation in Rome, even if they did not result in formal conquest and incorporation of any part of Britain. Despite some doubts about the effectiveness or permanence of Roman control, it does look as though the arrangements that Caesar made in 54 bc, including a demand for annual payments and the taking of hostages, did establish the principle of Roman authority in Britain. The historical and archaeological evidence gives some insight into how this may have worked. Rome continued its policy of recognizing client kings outside the empire, though the details of any formal alliance are sparse. Caesar had installed Mandubracius as ruler of the Trinovantes after his father’s death (De Bello Gallico V. 22), and he may also have set up Commius as a ruler in central southern England. The historical record is problematic (Bean 2000: 115–​117; Williams 2005a: 74–​75): Caesar made Commius king of the Atrebates in northern France in 57 bc, and he served Caesar well in his expeditions to Britain, but subsequently joined in the Gallic revolt of 52 bc, before moving back to Britain; the name Commius appears on coins of southern England, and Creighton (2000: 59–​64) has argued that Caesar established him as a king in southern England, despite his rebellion. These Caesarian arrangements may have laid the foundations for the emergence of two major polities, termed the Eastern Kingdom and the Southern Kingdom in southern East Anglia and southern central England respectively. At a later date, Strabo (Geog. 4.5.3) refers to British rulers sending embassies to Augustus and making dedications on the Capitol in Rome; there are no details of the date or the circumstances, but Braund (1996:  85)  has argued that one likely context could have been the formal recognition of client kings. Whether or not they were formally recognized as client kings, some British rulers, like Mandubracius before them, regarded the Roman emperor as a powerful ally, especially at times of factional competition: Augustus (Res Gestae 32) records that two British rulers, Dumnobellaun[us] and Tincom[arus], fled to him, while Adminius, son of Cunobelin, was banished by his father and fled to Rome (Suetonius, Caligula 44.2) and Verica did the same shortly before the Claudian invasion, ‘having been driven out by an uprising’ (Dio Cassius, Hist. lx. 19). Despite Caesar’s demand for annual payments, the archaeological evidence suggests that wealth was flowing from the Roman world to Britain. In the decades after his expeditions, there are significant changes in the coinage of south-​eastern England. The gold coins of the pre-​Caesarian period are rarely found in association with later issues, and may have been deliberately withdrawn from circulation. Instead, new series began to be produced from a different stock of metal, made from a different gold alloy and visible in a change of colour from a yellow gold to a red gold: the source was no longer the

164

164   Timothy Champion wealth of the LIA barbarian world, but that of the Roman world (Creighton 2000: 68–​ 74). Large quantities of silver also began to appear, again deriving from Roman bullion and coinage, and this was the source of new silver coin issues in many parts of south-​ eastern England (Farley 2012: 34–​130); some silver circulated in other forms, such as the Augustan silver cups found in the Welwyn and Welwyn Garden City burials (Smith 1912; Stead 1967). From c. 20 bc there were two other significant changes in the coinage of the Southern and Eastern Kingdoms. The imagery on the coins adopted a new set of motifs, closely following the coinage of Augustus (Creighton 2000: 80–​125). Some issues also began to carry inscriptions: the meaning of all the inscriptions is far from clear, but some are the names of rulers known from other sources. A small number of issues bear the name with the Latin title rex. Others have the Latin formula for claimed ancestral descent, F (for filius) and a personal name in the genitive. Thus, in the Eastern Kingdom, some coins bear the inscriptions cunobelinus and tasciovani f:  Tasciovanus is known only from such coins, but Cunobelinus is the person known to us from Suetonius as rex Britannorum (Caligula 44.2) and from Dio Cassius (Hist. lx. 20). In the Southern Kingdom several rulers are described as sons of Commius. Whether such relationships were actual biological ones or more like claims to ideologically important ancestry is not clear, but these inscriptions show a familiarity with Latin and with Roman usage. These developments in the imagery and inscriptions of the coinage of the south-​east demonstrate a close connection to Rome, and Creighton (2000: 117–​124) has suggested that this was due to the recognition of the rulers of the region as client kings and the practice of members of those ruling dynasties spending time in Rome as guests or hostages. The historical sources, together with the archaeological evidence of the coins and the burials, clearly demonstrate the existence of powerful ruling lineages in south-​eastern England, but the precise nature of that power is unclear. The classical sources naturally concentrate on their relationships with Rome, and so we know that they were able to raise and lead armies, to negotiate peace and give and take hostages; power could be inherited within the lineage and was the subject of factional competition, so Rome could be a powerful ally for those under internal pressure. What is not clear is how these lineages had achieved such positions, or the real extent of their powers: the use of the title rex and the occasional reference to them as kings in the historical sources says more about the classical perception of such rulers than about the reality of their rule. The concept of the Eastern and Southern Kingdoms is rooted in the distribution of coins, but the relationship between coins and rulers, or between coins and polities, is uncertain. Some of the coins also bear inscriptions that can be interpreted as the names of places, whether of mints or of other significant locations. In the Eastern Kingdom, Camulodunum (Colchester) and Verulamium (St Albans) are named, and in the Southern Kingdom, Calleva (Silchester). These were unlike the nucleated centres of population such as Braughing. They were important new sites, apparently founded in the second half of the first century bc, apparently in areas with little previous occupation. At Colchester (Gascoyne and Radford 2013), large parts of the enclosed area were

Britain before the Romans    165 unoccupied; there was a probable area of industrial production at Sheepen (Hawkes and Hull 1947), and rich burials at Lexden (Foster 1986) and Stanway (Crummy 2007), but the function of the Gosbecks complex in the Iron Age is unknown, despite its significance in the Roman period (Creighton 2006: 61–​64). Verulamium was very different (Haselgrove and Millett 1997): activity around a marshy river valley included the minting of coins and the deposition of burials, especially the large cemetery at King Harry Lane (Stead and Rigby 1989) and the rich Folly Lane cremation of the conquest period (Niblett 1999). In the Southern Kingdom, an area around Chichester was enclosed by a series of embankments that may have been begun in the middle Iron Age (Bradley 1971), but despite an assemblage of fine pre-​conquest ceramic imports there is no structural evidence for occupation (Davenport 2003); a suggested focus of activity at Selsey is more likely to be the result of multiple ritual deposits than domestic occupation (Bean 2000: 269–​271). The case is very different at Silchester, where excavation has produced unique evidence for the adoption of new forms of settlement and living (Fulford and Timby 2000). Below the later forum was part of a rectilinear street grid laid out in the late first century bc, flanked by rectangular structures. With imported amphorae that had brought wine and olive oil from Italy and Spain, as well as food remains including oysters and a high percentage of pig, it shows a remarkable contrast with contemporary sites in southern England and much closer similarities to Gallo-​Roman sites in northern France (Fulford and Timby 2000: 545–​564). Other sites were founded at this period and went on to be important later, though their status in the LIA is still obscure. At Canterbury structural evidence is fragmentary but includes a triple-​ditched enclosure (Blockley et al. 1995), but, on the basis of coins (Haselgrove 1987: 139–​145) and amphorae (Arthur 1986), the site may have been founded in the middle of the first century bc. The coin evidence, however, now suggests that it was one of several important sites in East Kent in the pre-​conquest period (Holman 2005). Leicester too was an important centre before the conquest, with imported pottery and evidence of coin production (Clay and Mellor 1985). Apart from the gold and silver, other items from the classical world also reached Britain. Italian wine had been imported from the late second century, as noted earlier, but the volume of imported ceramics grew considerably towards the end of the first century bc. Wine from Italy and Spain, and olive oil from Spain, were reaching south-​ eastern England (Carver 2001; Fitzpatrick and Timby 2002). The silver cups from the Welwyn and Welwyn Garden City burials may have been used for drinking such wine. Other imports included flagons from central Gaul and Italian-​type terra sigillata, but the largest volume of imports came from the Gallo-​Belgic industries founded in northern Gaul after about 15 bc. These products represented a fusion of Roman and late La Tène types, such as platters and beakers; they were startlingly different from the products of the indigenous tradition and were widely copied in Britain. The distribution of Gallo-​Belgic imports shows marked concentrations in central southern and in south-​ eastern England, roughly coincident with the Southern and Eastern Kingdoms defined by the coinage, but with other clusters in south-​western and eastern England, where the

166

166   Timothy Champion imports may have arrived independently or have been distributed onwards from the south-​east (Fitzpatrick and Timby 2002: figure 14.4). Though most attention has been paid to the amphorae and other ceramics, other innovations suggest a more complex pattern of interactions across the Channel. The inscribed coinage shows a familiarity not only with Latin (Williams 2007), but also with literacy more generally. Other signs of literacy in pre-​conquest contexts exist, though they are not frequent, including styli and graffiti on pottery, and a probably just post-​ conquest grave at Stanway, Colchester, which contained an inkwell (Crummy 2007); Williams (2001) has argued that literacy and its uses must have been familiar to Britain through contact with the literate societies of Roman Gaul, but that it was adopted only on a limited basis. Another innovation in Britain was the adoption of Roman-​style toilet equipment; this suggests a new concern with the adornment of the body, though the precise form that that took is uncertain (Hill 1997).

Changing Models of Interaction The evidence cited demonstrates clearly that there was close contact between Britain and the continent during the two centuries or more before the Roman invasion of ad 43. The nature of that contact, and the resulting processes of cultural change in Britain, have been conceptualized in changing ways since the later nineteenth century. The idea that prevailed then, and well into the twentieth century, of migration or invasion from the continent, has now been largely rejected. Too much attention has been seduced by Caesar’s reference to a migration from Belgium (De Bello Gallico V. 12: ‘qui … ex Belgio transierant’) as a historical event in need of archaeological substantiation or as an explanation for cultural change; there is much to be said for Williams’s suggestion (2005a: 75) that it is an echo of an indigenous tradition, perhaps an origin myth. Invasion hypotheses in general have gone out of favour, but there are more detailed objections in the case of the LIA. The long chronology of cultural change that is now evident is not compatible with a single causal event, or even with a small number of such events; in any case, the idea of a migration does not of itself explain the adoption in Britain of new cultural practices, and the rejection of others. A much longer history of interaction is required, and a scenario that envisages regular cross-​Channel communication, with people crossing the water for very different reasons. Some people may have crossed and returned, perhaps on many occasions, but others may have crossed and settled. It is the nature of these crossings and their cultural consequences that is at the heart of the debate. Subsequent explanations have, to some extent, followed contemporary archaeological fashions. The heyday of processual archaeology can be seen in studies of LIA Britain, emphasizing trade as the mechanism of change (Haselgrove 1982, 1984). These were embedded in a spatial model of core and periphery, where the core (Rome) was the active force in cultural change in an expanding periphery (the provinces), driven

Britain before the Romans    167 by trade in prestige goods that allowed the indigenous elites to accumulate wealth and power. The problems with this vision were that it placed too little weight on the political connections that were implied by the historical evidence and the coinage, and gave an exaggerated role to trade, both in its social significance (Woolf 1993b) and in its volume (Willis 1994). The early trade was mostly confined to amphorae of wine, and did not include other items such as Campanian ware, and only comparatively modest quantities reached Britain, to be used and deposited in a variety of ways appropriate to indigenous society. Nor were there clear signs of direct trade with Rome; without a historical record or an agreed archaeological methodology for recognizing Roman traders in Britain, claims for their presence, as at Braughing (Partridge 1981: 351), are not totally convincing. Nor did the concept of Romanization before the conquest (Kendrick and Hawkes 1932: 207; Haselgrove 1984) stand up to scrutiny. It became increasingly clear that many of the innovations in LIA Britain were derived from contacts with the late La Tène world of continental Europe, whether coinage, burials, or pottery. As already argued, even the objects of Italian manufacture such as amphorae and bronze vessels were common in the late La Tène world and could have been transmitted from there (Millett 1990: 38–​39). When, after 15 bc, larger quantities of ceramic finewares were imported, they were from the Gallo-​Belgic industries of northern Gaul, very different from the wares that were supplied to Roman sites after the conquest (Pitts 2005). As Gosden (2004: 109) wrote, ‘South-​eastern Britain was as much Gallicised as Romanised in this period, with the potentates of northern Gaul vital intermediaries for Mediterranean influences.’ More recent discussion has moved away from large-​scale models of interaction to consideration of the processes of cultural change involved, with particular attention to the active decisions of the indigenous populations as conscious agents, and to the meaning of the material culture involved. It is clear that there was regular traffic across the Channel and that people in south-​eastern England would have been well aware of social practices and material culture on the continent. The nature of the connections are inevitably somewhat speculative, but political alliances, marriage ties, and military support are likely, first with the polities of northern Gaul and then with Rome. There are some hints of the possibilities: a late La Tène helmet from a cremation burial near Canterbury (Farley et al. 2014) and the Roman military horsegear in the Folly Lane burial. The magnificent Roman cavalry helmet from Hallaton, Leicestershire, may have been captured after the invasion, but could equally well have been the result of pre-​ conquest service in the Roman army or a diplomatic gift (Score 2011). Many people may have crossed and returned, others may have settled permanently: settlers from Gaul have been suggested for the foundation of Silchester (Fulford and Timby 2000: 563–​564). These contacts would have been accompanied by the exchange of material items, most importantly the Roman gold and silver that was transformed into insular coins; they would also have given the opportunity of knowledge of continental practices, some of which were adopted or imitated in Britain. Such adoption, however, was very variable. It varied geographically: coins, for example, were widely distributed in Britain, but only

168

168   Timothy Champion regularly used in southern and eastern England; cremation burial was adopted in an even more restricted region, and perhaps by only part of the population. Other things, such as Latin and literacy, were known, but used only in minimal ways in restricted regions. Even where there are definite imports, their cultural significance and the social context of their use may have been shaped by indigenous customs. Wine was certainly imported, but its acceptability may have been because it fitted local practices of feasting, alongside other drinks such as beer (Williams 2005b); its inclusion in burials in Britain was most un-​Roman. To try to explain this clash of cultures, a number of scholars have found an instructive parallel for the interaction of Rome and the non-​Roman societies of north-​western Europe in the confrontation of European and non-​European societies in the early modern period. They have looked not to Macaulay’s Sandwich Islands, but to the east coast of North America, and in particular to the work of Richard White (1991) and his concept of the ‘middle ground’ to describe the relatively stable world that was formed for a century or more in the Great Lakes region as the British and French penetrated deeper into the territories of the native Americans. It was a world in which the old cultural norms had been fatally disrupted, but neither side had established a new permanent social or political order; in which conflicting languages, religions, technologies, modes of exchange, and patterns of social relationships were accommodated. The exchange of material items such as furs and iron axes was an important part of this interaction. In a brief discussion, Gosden (2004: 110) envisaged the interaction of the Roman and non-​ Roman worlds as a series of middle grounds beyond the contemporary frontier, ‘bringing new sets of cultural resources … which could be used, refused or subverted’. Woolf (2011) also drew on the concept of the middle ground to explore how the classical world tried to come to terms intellectually with its barbarian neighbours in Spain, Gaul, and Britain. Within a relatively short period, new stories were created, drawing on classical and native sources, to create a new hybrid body of historical and ethnographic knowledge. We may never know how the pre-​conquest inhabitants of Britain conceptualized their new Roman neighbours, but the classical literary tradition illuminates the reverse process. Williams (2005b: 37–​38) has also looked to North America as a comparative case, in particular to the successful integration of imported alcohol into indigenous social rituals. In the case of LIA Britain, he emphasized that, if we are looking for the meaning of innovations, we must look ‘within the changing circumstances of Britain itself ’. The most extended discussion of North American parallels has been by Farley (2012: 131–​ 183), who has contrasted different phases of European and native American interaction: an earlier phase where attempts to understand each other led to confusion and frequent misunderstandings, followed by a phase in which a hybrid world of meaning had been negotiated. The problem of understanding this clash of cultures in LIA Britain can be exemplified by the imported Italian bronze vessels, such as the jug and pan first identified by Evans at Aylesford that began this debate. In the classical world such vessels are interpreted as parts of a set for the washing of hands (Nuber 1972; Bolla 1991; Feugère and de

Britain before the Romans    169 Marinis 1991), but their association with evidence of feasting in elite burials in England has suggested to some (e.g. Carver 2001; Ralph 2007) that they were used for the service of wine. What did they mean in LIA Britain? Were they simply strange exotics, valued for being exotic? Or did they signify the adoption of Roman dining rituals among the British elites? Or were they incorporated into local feasting practices, and perhaps used in very different ways from their original function?

References Primary Sources Athenaeus, Deipnosophistae. Augustus, Res gestae. Caesar, De Bello Gallico. Dio Cassius, History of Rome. Polybius, Histories. Strabo, Geography. Suetonius, Caligula.

Secondary Sources Adams, S. A. (2014). ‘The First Brooches in Britain: From Manufacture to Deposition in the Early and Middle Iron Age’. Ph.D. thesis, University of Leicester. Aldhouse-​Green, M. J. (2010). Caesar’s Druids:  An Ancient Priesthood. New Haven and London: Yale University Press. Allen, D. F. (1967). ‘Iron Currency Bars in Britain’, Proceedings of the Prehistoric Society, 33: 307–​335. Allen, D. F. (1976). ‘Wealth, Coinage and Money in a Celtic Society’, in J. V. S. Megaw (ed.), To Illustrate the Monuments: Essays on Archaeology Presented to Stuart Piggott. London: Thames and Hudson, 199–​208. Andrews, P., Biddulph, E., Hardy, A., and Brown, R. (2011). Settling the Ebbsfleet Valley: High Speed 1 Excavations at Springhead and Northfleet, Kent: The Late Iron Age, Roman, Saxon, and Medieval Landscape, Volume 1:  The Sites. Oxford and Salisbury:  Oxford Wessex Archaeology. Armit, I. (2007). ‘Social Landscapes and Identities in the Irish Iron Age’, in C. Haselgrove and T. Moore (eds), The Later Iron Age in Britain and Beyond. Oxford: Oxbow, 130–​139. Arthur, P. (1986). ‘Roman Amphorae from Canterbury’, Britannia, 17: 239–​258. Atkinson, M., and Preston, S. J. (1998). ‘The Late Iron Age and Roman Settlement at Elms Farm, Heybridge, Essex, Excavations 1993–​1995: Interim Report’, Britannia, 29: 85–​110. Barrett, J., Bowden, M., and McOmish, D. (2011). ‘The Problem of Continuity: Reassessing the Shape of the British Iron Age Sequence’, in T. Moore and X.-​L. Armada (eds), Atlantic Europe in the First Millennium bc: Crossing the Divide. Oxford: Oxford University Press, 439–​448. Bean, S. C. (2000). The Coinage of the Atrebates and the Regni. Oxford:  Oxford University School of Archaeology. Blancquaert, G., and Bostyn, F. (1998). ‘L’Âge du fer à Coquelles et Fréthun (Pas-​de-​Calais) (Fouilles du Transmanche 1986–​1988)’, Revue du Nord, 80: 109–​137.

170

170   Timothy Champion Blockley, K., Blockley, M., Blockley, P., Frere, S., and Stow, S. (1995). Excavations in the Marlowe Car Park and Surrounding Areas. The Archaeology of Canterbury 5. Canterbury: Canterbury Archaeological Trust. Bolla, M. (1991). ‘Considerazioni sulla funzione dei vasi in bronzo tardorepubblicani in Italia settentrionale’, in M. Feugère and C. Rolley (eds), La Vaisselle tardo-​républicaine en bronze: Actes de la table-​ronde CNRS organisée à Lattes du 26 au 28 avril 1990 par l’UPR 290 (Lattes) et le GDR 125 (Dijon). Centre de recherches sur les techniques gréco-​romaines, no.  13. Dijon. Université de Bourgogne, Centre de recherches sur les techniques gréco-​ romaines, 143–​153. Boube, C. (1991). ‘Les Cruches’, in M. Feugère and C. Rolley (eds), La Vaisselle tardo-​ républicaine en bronze: Actes de la table-​ronde CNRS organisée à Lattes du 26 au 28 avril 1990 par l’UPR 290 (Lattes) et le GDR 125 (Dijon). Centre de recherches sur les techniques gréco-​ romaines, no. 13. Dijon: Université de Bourgogne, Centre de recherches sur les techniques gréco-​romaines, 23–​45. Bradley, R. (1971). ‘A Field Survey of the Chichester Entrenchments’, in B. Cunliffe (ed.), Excavations at Fishbourne. Research Report 26–​27. London:  Society of Antiquaries of London, 17–​36. Bradley, R. (1998). The Passage Of Arms: An Archaeological Analysis of Prehistoric Hoards and Votive Deposits. 2nd edn. Oxford: Oxbow. Braund, D. (1996). Ruling Roman Britain: Kings, Queens, Governors and Emperors from Julius Caesar to Agricola. Abingdon: Routledge. Bryant, S. (2007). ‘Central Places or Special Places? The Origins and Development of “Oppida” in Hertfordshire’, in C. Haselgrove and T. Moore (eds), The Later Iron Age in Britain and Beyond. Oxford: Oxbow, 62–​80. Bushe-​Fox, J. P. (1925). Excavations of the Late-​Celtic Urnfield at Swarling, Kent. Society of Antiquaries of London Research Report. London: Society of Antiquaries. Carver, E. (2001). The Visibility of Imported Wine and its Associated Accoutrements in Later Iron Age Britain. British Archaeological Reports, British Series 325. Oxford: Archaeopress. Chadburn, A. D. B. (2006). ‘Aspects of the Iron Age Coinage of Northern East Anglia with Especial Reference to Hoards’. Ph.D. thesis, University of Nottingham. Champion, T. C. (1975). ‘Britain and the European Iron Age’, Archaeologia Atlantica, 1: 127–​145. Champion, T. C. (2014). ‘Food, Technology and Culture in the Late Bronze Age of Southern Britain: Perforated Clay Plates of the Lower Thames Valley’, Proceedings of the Prehistoric Society, 80: 279–​298. Clay, P., and Mellor, J. (1985). Excavations in Bath Lane, Leicester. Archaeological Report 10. Leicester: Leicestershire Museums, Art Galleries and Record Service. Collis, J. R. (1977). ‘Pre-​Roman Burial Rites in North-​Western Europe’, in R. Reece (ed.), Burial in the Roman World. Research Report 22. London: Council for British Archaeology, 1–​13. Collis, J. R. (1984). Oppida: Earliest Towns North of the Alps. Sheffield: Department of Prehistory and Archaeology, University of Sheffield. Cooley, A. (2009). Res gestae divi Augusti:  Text, translation, and commentary. Cambridge: Cambridge University Press. Creighton, J. (2000). Coins and Power in Late Iron Age Britain. Cambridge:  Cambridge University Press. Creighton, J. (2006). Britannia: The Creation of a Roman Province. London: Routledge. Crew, P. (1994). ‘Currency Bars in Great Britain:  Typology and Function’, in M. Mangin (ed.), La Sidérurgie ancienne de l’Est de la France dans son contexte européen: Archéologie

Britain before the Romans    171 et archéométrie: Actes du colloque de Besançon, 10–​13 novembre 1993. Annales littéraires de l’Université de Besançon 536. Paris: Belles Lettres, 340–​350. Cripps, L. J. (2007). ‘Re-​Situating the Later Iron Age in Cornwall and Devon: New Perspectives from the Settlement Record’, in C. Haselgrove and T. Moore (eds), The Later Iron Age in Britain and Beyond. Oxford: Oxbow, 140–​155. Crumley, C. L. (1995). ‘Heterarchy and the Analysis of Complex Societies’, in R. M. Ehrenreich, C. L. Crumley, and J. E. Levy (eds), Heterarchy and the Analysis of Complex Societies. Ann Arbor: American Anthropological Association, 1–​5. Crummy, P. (2007). Stanway: An Élite Burial Site at Camulodunum. London: Society for the Promotion of Roman Studies. Cunliffe, B. (1976). ‘The Origins of Urbanisation in Britain’, in B. Cunliffe and T. Rowley (eds), Oppida: The Beginnings of Urbanisation in Barbarian Europe. Oxford: British Archaeological Reports, Supplementary Series 11, 135–​161. Cunliffe, B. (1984). Danebury: An Iron Age Hillfort in Hampshire. Volume 2: The Excavations 1969–​1978: The Finds. Research Report 52. London: Council for British Archaeology. Cunliffe, B. (1987). Hengistbury Head Dorset: Volume 1: The Prehistoric and Roman Settlement, 3500 bc–​ad 500. Monograph 13. Oxford:  Oxford University Committee for Archaeolgy, Institute of Archaeology. Cunliffe, B. (1988). Mount Batten Plymouth: A Prehistoric and Roman Port. Monograph 26. Oxford: Oxford University Committee for Archaeology, Institute of Archaeology. Cunliffe, B. (2005). Iron Age Communities in Britain:  An Account of England, Scotland and Wales from the Seventh Century bc to the Roman Conquest. 4th edn. London: Routledge. Cunliffe, B., and de Jersey, P. (1997). Armorica and Britain:  Cross-​Channel Relationships in the Late First Millennium BC. Monograph 45. Oxford: Oxford University Committee for Archaeology. Cunliffe, B., and Poole, C. (1991). Danebury:  An Iron Age hillfort in Hampshire. Volume 5: The Excavations 1979–​1988: The Finds. Research Report 73. London: Council for British Archaeology. Davenport, C. (2003). ‘The Late Pre-​Roman Iron Age of the West Sussex Coastal Plain: Continuity or Change?’, in D. Rudling (ed.), The Archaeology of Sussex to ad 2000. King’s Lynn: Heritage Marketing and Publications, 101–​110. Davis, O. (2013). ‘Re-​ Interpreting the Danebury Assemblage:  Houses, Households, and Community’, Proceedings of the Prehistoric Society, 79: 353–​375. de Jersey, P. (2008). ‘Evans and Ancient British Coins’, in A. MacGregor (ed.), Sir John Evans 1823–​ 1908:  Antiquity, Commerce and Natural Science in the Age of Darwin. Oxford: Ashmolean Museum, 152–​172. Dietler, M. (1998). ‘Feasts and Commensal Politics in the Political Economy: Food, Power and Status in Prehistoric Europe’, in P. Wiessner and W. Schiefenhövel (eds), Food and the Status Quest: An Interdisciplinary Perspective. Providence: Berghahn Books, 87–​125. Dietler, M. (2001). ‘Theorizing the Feast: Rituals of Consumption, Commensal Politics, and Power in African Contexts’, in M. Dietler and B. Hayden (eds), Feasts: Archaeological and Ethnographic Perspectives on Food, Politics and Power. Washington: Smithsonian Institution Press, 65–​114. Elsdon, S. (1997). Old Sleaford Revealed: A Lincolnshire Settlement in Iron Age, Roman, Saxon and Medieval Times: Excavations 1882–​1995. Monograph 78. Oxford: Oxbow. Eluère, C. (1987). L’Or des Celtes. Paris: Bibliothèque des Arts. Evans, A. J. (1890). ‘On a Late-​Celtic Urn-​Field at Aylesford, Kent’, Archaeologia, 52: 178–​192.

172

172   Timothy Champion Evans, J. (1864). The Coins of the Ancient Britons. London: J. Russell Smith. Farley, J. M.-​A. (2012). ‘At the Edge Of Empire: Iron Age and Early Roman Metalwork in the East Midlands’. Ph.D. thesis, University of Leicester. Farley, J. M.-​A., Parfitt, S. A., and Richardson, A. (2014). ‘An Iron Age Helmet Burial from Bridge near Canterbury, Kent’, Proceedings of the Prehistoric Society, 80: 379–​388. Farley, M. (1983). ‘A Mirror Burial at Dorton, Buckinghamshire’, Proceedings of the Prehistoric Society, 49: 269–​302. Feugère, M., and de Marinis, R. (1991). ‘Les poêlons’, In M. Feugère and C. Rolley (eds), La Vaisselle tardo-​républicaine en bronze: actes de la table-​ronde CNRS organisée à Lattes du 26 au 28 avril 1990 par l’UPR 290 (Lattes) et le GDR 125 (Dijon). Centre de recherches sur les techniques gréco-​romaines; no 13. Dijon: Université de Bourgogne, Centre de recherches sur les techniques gréco-​romaines, 97–​112. Fitzpatrick, A. P. (1985). ‘The Distribution of Dressel 1 Amphorae in North-​West Europe’, Oxford Journal of Archaeology, 4: 305–​340. Fitzpatrick, A. P. (1997). Archaeological Excavations on the Route of the A27 Westhampnett Bypass, West Sussex, 1992. Volume 2: the Late Iron Age, Romano-​British and Anglo-​Saxon cemeteries. Salisbury: Trust for Wessex Archaeology. Fitzpatrick, A. P. (2000). ‘Ritual, Sequence and Structure in Late Iron Age Mortuary Practices in North-​Western Europe’, in J. Pearce, M. Millett, and M. Struck (eds), Burial, Society and Context in the Roman World. Oxford: Oxbow, 15–​29. Fitzpatrick, A. P. (2005). ‘Gifts for the Golden Gods: Iron Age Hoards of Torques and Coins’, in C. Haselgrove and D. Wigg-​Wolf (eds), Iron Age Coinage and Ritual Practices. Studien zu Fundmünzen der Antike. Mainz: von Zabern, 157–​182. Fitzpatrick, A. P. (2009). ‘The Champion’s Portion:  Feasting in the Celtic Pre-​ Roman Iron Age’, in G. Cooney, K. Becker, J. Coles, M. Ryan, and S. Sievers (eds), Relics of Old Decency: Archaeological Studies in Later Prehistory. Dublin: Wordwell, 389–​403. Fitzpatrick, A. P., and Timby, J. (2002). ‘Roman Pottery in Iron Age Britain’, in A. Woodward and J. D. Hill (eds), Prehistoric Britain: The Ceramic Basis. Oxford: Oxbow, 161–​172. Fortes, M., and Evans-​Pritchard, E. E. (1940) (eds). African Political Systems. London: Oxford University Press. Foster, J. (1986). The Lexden Tumulus: A Re-​Appraisal of an Iron Age Burial from Colchester, Essex. British Archaeological Reports, British Series 56. Oxford. Fulford, M. G., and Timby, J. R. (2000). Late Iron Age and Roman Silchester: Excavations on the Site of the Forum-​Basilica 1977, 1980–​86. Britannia Monograph Series 15. London: Society for the Promotion of Roman Studies. Furger-​Gunti, A. (1982). ‘Der “Goldfund von Saint-​Louis” bei Basel und ähnliche keltische Schatzfunde’, Zeitschrift für Schweizerische Archaeologie und Kunstgeschichte, 39: 1–​47. Gardner, A. (2013). ‘Thinking about Roman Imperialism: Postcolonialism, Globalisation and Beyond?’, Britannia, 44: 1–​25. Garrow, D., Gosden, C., Hill, J. D., and Bronk Ramsey, C. (2009). ‘Dating Celtic Art: A Major Radiocarbon Dating Programme of Iron Age and Early Roman Metalwork in Britain’, Archaeological Journal, 166: 79–​123. Gascoyne, A., and Radford, D. (2013). Colchester: Fortress of the War God: An Archaeological Assessment. Oxford: Oxbow Books. Gosden, C. (2004). Archaeology and Colonialism: Cultural Contact from 5000 bc to the Present. Cambridge: Cambridge University Press.

Britain before the Romans    173 Hamilton, S., and Manley, J. (2000). ‘Hillforts, Monumentality and Place: A Chronological and Topographic Review of First Millennium bc Hillforts in South-​East England’, European Journal of Archaeology, 4: 7–​42. Haselgrove, C. (1982). ‘Wealth, Prestige and Power:  The Dynamics of Late Iron Age Political Centralisation in South-​East England’, in A. C. Renfrew and S. J. Shennan (eds), Ranking, Resource and Exchange:  Aspects of the Archaeology of Early European Society. Cambridge: Cambridge University Press, 79–​88. Haselgrove, C. (1984). ‘“Romanization” before the Conquest: Gaulish Precedents and British Consequences’, in T. F.  C. Blagg and A. C. King (eds), Military and Civilian in Roman Britain: Cultural Relationships on a Frontier Province. British Archaeological Reports, British Series 136. Oxford: British Archaeological Reports, 5–​63. Haselgrove, C. (1987). Iron Age Coinage in South-​East England: The Archaeological Context. British Archaeological Reports, British Series 174. Oxford: British Archaeological Reports. Haselgrove, C. (1988). ‘The Archaeology of British Potin Coinage’, Archaeological Journal, 145: 99–​122. Haselgrove, C. (1995). ‘Potin Coinage in Iron Age Britain, Archaeology and Chronology’, Gallia, 52: 117–​127. Haselgrove, C. (1997). ‘Iron Age Brooch Deposition and Chronology’, in A. Gwilt and C. Haselgrove (eds), Reconstructing Iron Age Societies. Oxford: Oxbow, 51–​72. Haselgrove, C. (2005). ‘A Trio of Temples: A Reassessment of Iron Age Coin Deposition at Hayling Island, Harlow and Wanborough’, in C. Haselgrove and D. Wigg-​Wolf (eds), Iron Age Coinage and Ritual Practices. Studien zu Fundmünzen der Antike. Mainz: von Zabern, 381–​418. Haselgrove, C. (2006). ‘Early Potin Coinage in Britain: An Update’, in P. de Jersey (ed.), Celtic Coinage:  New Discoveries, New Discussion. British Archaeological Reports, International Series 1532. Oxford: Archaeopress, 17–​28. Haselgrove, C., and Millett, M. (1997). ‘Verlamion Reconsidered’, in A. Gwilt and C. Haselgrove (eds), Reconstructing Iron Age Societies. Oxford: Oxbow, 282–​296. Haselgrove, C., and Moore, T. (2007a) (eds). The Later Iron Age in Britain and Beyond. Oxford: Oxbow. Haselgrove, C., and Moore, T. (2007b). ‘New Narratives of the Later Iron Age’, in C. Haselgrove and T. Moore (eds), The Later Iron Age in Britain and Beyond. Oxford: Oxbow, 1–​15. Haselgrove, C., and Pope, R. E. (2007) (eds). The Earlier Iron Age in Britain and the Near Continent. Oxford: Oxbow. Haselgrove, C., Armit, I., Champion, T. C., Creighton, J., Gwilt, A., Hill, J. D., Hunter, F., and Woodward, A. (2001). Understanding the British Iron Age:  An Agenda for Action. Salisbury: Trust for Wessex Archaeology. Hawkes, C. F. C. (1931). ‘Hillforts’, Antiquity, 5: 60–​97. Hawkes, C. F. C. (1959). ‘The ABC of the British Iron Age’, Antiquity, 33: 170–​182. Hawkes, C. F. C., and Dunning, G. C. (1930). ‘The Belgae of Gaul and Britain’, Archaeological Journal, 87: 150–​335. Hawkes, C. F. C., and Hull, M. R. (1947). Camulodunum: First Report on the Excavations at Colchester 1930–​1939. Research Report XIV. Oxford: Society of Antiquaries of London. Hayden, B. (2014). The Power of Feasts: From Prehistory to the Present. Cambridge: Cambridge University Press. Hill, J. D. (1995). ‘How Should we Understand Iron Age Societies and Hillforts? A Contextual Study from Southern Britain’, in J. D. Hill and C. G. Cumberpatch (eds), Different Iron

174

174   Timothy Champion Ages:  Studies on the Iron Age in Temperate Europe. British Archaeological Reports, International Series 602. Oxford: Tempus Reparatum, 45–​66. Hill, J. D. (1996). ‘Hillforts and the Iron Age of Wessex’, in T. C. Champion and J. R. Collis (eds), The Iron Age in Britain and Ireland: Recent Trends. Sheffield: J. R. Collis Publications, 95–​116. Hill, J. D. (1997). ‘“The End of One Kind of Body and the Beginning of Another Kind of Body”? Toilet Instruments and “Romanization”’, in A. Gwilt and C. Haselgrove (eds), Reconstructing Iron Age Societies. Oxford: Oxbow, 96–​107. Hill, J. D. (2002). ‘Just about the Potter’s Wheel? Using, Making and Depositing Middle and Later Iron Age Pots in East Anglia’, in A. Woodward and J. D. Hill (eds), Prehistoric Britain: The Ceramic Basis. Oxford: Oxbow, 143–​160. Hill, J. D. (2007). ‘The Dynamics of Social Change in Later Iron Age Eastern and South-​Eastern England’, in C. Haselgrove and T. Moore (eds), The Later Iron Age in Britain and Beyond. Oxford: Oxbow, 16–​40. Hill, J. D. (2011). ‘How did British Middle and Late Pre-​Roman Iron Age Societies Work (if they did)?’, in T. Moore and X.-​L. Armada (eds), Atlantic Europe in the First Millennium bc: Crossing the Divide. Oxford: Oxford University Press, 242–​263. Hill, J. D., Evans, C., and Alexander, M. (1999). ‘The Hinxton Rings: A Late Iron Age Cemetery at Hinxton, Cambridgeshire, with a Reconsideration of Northern Aylesford–​Swarling Distributions’, Proceedings of the Prehistoric Society, 65: 243–​273. Hill, J. D., Spence, A. J., LaNeice, S., and Worrell, S. (2004). ‘The Winchester Hoard: A Find of Unique Iron Age Gold Jewellery from Southern England’, Antiquaries Journal, 84: 1–​22. Hingley, R. (1984). ‘Towards Social Analysis in Archaeology: Celtic Society in the Iron Age of the Upper Thames Valley’, in B. Cunliffe and D. Miles (eds), Aspects of the Iron Age in Central Southern England. Oxford: Oxford University Committee for Archaeology, 72–​88. Hingley, R. (1990). ‘Iron Age “Currency Bars”:  The Archaeological and Social Context’, Archaeological Journal, 147: 91–​117. Hingley, R. (2000). Roman Officers and English Gentlemen: The Imperial Origins of Roman Archaeology. London: Routledge. Hingley, R. (2005). ‘Iron Age “Currency Bars” in Britain:  Items of Exchange in Liminal Contexts?’, in C. Haselgrove and D. Wigg-​Wolf (eds), Iron Age Coinage and Ritual Practices. Studien zu Fundmünzen der Antike. Mainz: von Zabern, 183–​205. Hingley, R. (2008). The Recovery of Roman Britain:  1586–​ 1906:  A  Colony so Fertile. Oxford: Oxford University Press. Hingley, R., and Miles, D. (1984). ‘Aspects of Iron Age Settlement in the Upper Thames Valley’, in B. Cunliffe and D. Miles (eds), Aspects of the Iron Age in Central Southern Britain. Oxford: University of Oxford Commitee for Archaeology, 52–​7 1. Hobbs, R. (1996). British Iron Age Coins in the British Museum. London: British Museum Press. Holman, D. (2000). ‘Iron Age Coinage in Kent’, Archaeologia Cantiana, 120: 205–​234. Holman, D. (2005). ‘Iron Age Coinage and Settlement in East Kent’, Britannia, 36: 1–​54. Hoselitz, V. (2007). Imagining Roman Britain:  Victorian Responses to a Roman Past. Woodbridge: Boydell. Hurtrelle, J., Monchy, E., Roger, F., Rossignol, P., and Villes, A. (1990). Les Débuts du second âge du fer dans le Nord de la France. Les Dossiers de Gauheria. 1. Givenchy-​en-​Goelle: Gauheria. Hutcheson, N. (2003). ‘Material Culture in the Landscape: A New Approach to the Snettisham Hoards’, in J. Humphrey (ed.), Re-​ Searching the Iron Age. Leicester:  University of Leicester, 87–​97.

Britain before the Romans    175 Hutcheson, N. (2004). Later Iron Age Norfolk:  Metalwork, Landscape and Society. British Archaeological Reports. British Series 361. Oxford: Archaeopress. Hutton, R. (2007). The Druids: A History. London: Hambledon Continuum. Hutton, R. (2009). Blood and Mistletoe: The History of the Druids in Britain. New Haven and London: Yale University Press. Jones, M. K. (2007). ‘A Feast of Beltain? Reflections on the Rich Danebury Harvests’, in C. Gosden, H. Hamerow, P. de Jersey, and G. Lock (eds), Communities and Connections: Essays in Honour of Barry Cunliffe. Oxford: Oxford University Press, 143–​153. Jope, E. M. (1961). ‘Daggers of the Early Iron Age in Britain’, Proceedings of the Prehistoric Soceity, 27: 307–​343. Jope, E. M. (2000). Early Celtic Art in the British Isles. Oxford: Oxford University Press. Joy, J. (2014). ‘“Fire burn and cauldron bubble”: Iron Age and Early Roman Cauldrons of Britain and Ireland’, Proceedings of the Prehistoric Society, 80: 327–​362. Joy, J. (2015). ‘Connections and Separation? Narratives of Iron Age Art in Britian and its Relationship with the Continent’, in H. Anderson-​Whymark, D. Garrow, and F. Sturt (eds), Continental Connections: Exploring Cross-​Channel Relationships from the Mesolithic to the Iron Age. Oxford: Oxbow, 145–​165. Kendrick, T. D., and Hawkes, C. F.  C. (1932). Archaeology in England and Wales 1914–​1931. London: Methuen. King, A. C., and Soffe, G. (1994.) ‘The Iron Age and Roman Temple on Hayling Island’, in A. P. Fitzpatrick and E. L. Morris (eds), The Iron Age in Wessex: Recent Work. Salisbury: Trust for Wessex Archaeology, 114–​116. King, A. C., and Soffe, G. (2008). ‘Hayling Island: A Gallo-​Roman Temple in Britain’, in D. Rudling (ed.), Ritual Landscapes of Roman South-​ East Britain. Great Dunham and Oxford: Heritage Marketing & Publications and Oxbow, 139–​151. Kirch, P. V. (2010). How Chiefs Became Kings: Divine Kingship and the Rise of Archaic States in Ancient Hawai’i. Berkeley: University of California Press. Kirch, P. V. (2012). A Shark Going Inland Is My Chief: The Island Civilization of Ancient Hawai’i. Berkeley: University of California Press. Kirch, P. V., and Sahlins, M. D. (1992). Anahulu: The Anthropology of History in the Kingdom of Hawaii. Chicago: University of Chicago Press. Knight, D. (2007). ‘From Open to Enclosed: Iron Age Landscapes in the Trent Valley’, in C. Haselgrove and T. Moore (eds), The Later Iron Age in Britain and Beyond. Oxford: Oxbow, 190–​218. Lambrick, G. (2009). The Thames through Time:  The Archaeology of the Gravel Terraces of the Upper and Middle Thames:  The Thames Valley in Late Prehistory, 1500 bc–​ad 50. Oxford: Oxford University School of Archaeology for Oxford Archaeology. Leins, I. (2008). ‘What can be Inferred from the Regional Stylistic Diversity of Iron Age Coinage?’, in D. Garrow, C. Gosden, and J. D. Hill (eds), Rethinking Celtic Art. Oxford: Oxbow, 100–​112. Leman-​Delerive, G. (1984). ‘Céramique laténienne domestique de la région lilloise (Nord)’, Gallia, 42: 79–​95. Macaulay, T. B. (1849). The History of England from the Accession of James II. London: Longman, Green, Brown and Longmans. May, J. (1996). Dragonby: Report on Excavations at an Iron Age and Romano-​British Settlement in North Lincolshire. Oxford: Oxbow.

176

176   Timothy Champion Millett, M. (1990). The Romanization of Britain:  An Essay in Archaeological Interpretation. Cambridge: Cambridge University Press. Moore, T. (2006). Iron Age Societies in the Severn-​Cotswolds: Developing Narratives of Social and Landscape Change. British Archaeological Reports, British Series 421. Oxford: Archaeopress. Moore, T. (2007). ‘The Early to Late Iron Age Transition in the Severn–​Cotswolds: Enclosing the Household?’, in C. C. Haselgrove and R. E. Pope (eds), The Earlier Iron Age in Britain and the Near Continent. Oxford: Oxbow, 259–​278. Moore, T. (2011). ‘Detribalizing the Later Prehistoric Past: Concepts of Tribes in Iron Age and Roman Studies’, Journal of Social Archaeology, 11/​3: 334–​360. Morris, E. L. (1994.) ‘Production and Distribution of Pottery and Salt in Iron Age Britain: A Review’, Proceedings of the Prehistoric Society, 60: 371–​393. Morris, E. L. (1996). ‘Iron Age Artefact Production and Exchange’, in T. C. Champion and J. Collis (eds), The Iron Age in Britain and Ireland:  Recent Trends. Sheffield:  J R Collis Publications, 41–​65. Nash, D. (1987). Coinage in the Celtic World. London: Seaby. Niblett, R. (1999). The Excavation of a Ceremonial Site at Folly Lane, Verulamium. Britannia Monograph Series 14. London: Society for the Promotion of Roman Studies. Nuber, H. U. (1972). ‘Kanne und Griffschale: Ihr Gebrauch im täglichen Leben und die Beigabe in Gräbern der römischen Kaiserzeit’, Bericht der Römisch-​Germanischen Kommission, 53: 1–​232. Parfitt, K. (1995). Iron Age Burials from Mill Hill, Deal. London: British Museum Press. Partridge, C. (1981). Skeleton Green: A Late Iron Age and Romano-​British Site. Britannia Monograph Series 2. London: Society for the Promotion of Roman Studies. Peacock, D. P. S. (1971). ‘Roman Amphorae in Pre-​Roman Britain’, in D. Hill and M. Jesson (eds), The Iron Age and its Hill-​Forts. Southampton: Southampton University Archaeology Society, 161–​188. Peacock, D. P. S. (1984). ‘Amphorae in Iron Age Britain: A Reassessment’, in S. Macready and F. H. Thompson (eds), Cross-​Channel Trade between Gaul and Britain in the Pre-​Roman Iron Age. Occasional Paper (ns) 4. London: Society of Antiquaries, 37–​42. Peacock, D. P. S. (1987). ‘Iron Age and Roman Quern Production at Lodsworth, West Sussex’, Antiquaries Journal, 67: 61–​85. Peacock, D. P. S., and Cutler, L. (2011). ‘The Earliest Rotary Querns in Southern England’, in D. Williams and D. Peacock (eds), Bread for the People: The Archaeology of Mills and Milling. Oxford: Archaeopress, 70–​80. Piggott, S. (1985). The Druids. London: Thames & Hudson. Pitts, M. (2005). ‘Pots and Pits: Drinking and Deposition in Late Iron Age South-​East Britain’, Oxford Journal of Archaeology, 24/​2: 143–​161. Pitts, M. (2010). ‘Re-​Thinking the Southern British Oppida: Networks, Kingdoms and Material Culture’, European Journal of Archaeology, 13/​1: 32–​63. Potter, T. W., and Trow, S. D. (1988). ‘Puckeridge–​Braughing, Herts.:  The Ermine Street Excavations, 1971–​1972’, Hertfordshire Archaeology, 10: 1–​191. Qualmann, K. E., and Whinney, R. J.  B. (2004). Oram’s Arbour:  The Iron Age Enclosure at Winchester. Winchester: Winchester Museums Service. Ralph, S. (2007). Feasting and Social Complexity in Later Iron Age East Anglia. British Archaeological Reports British Series 451. Oxford: Archaeopress. Rudd, C., Cottam, E., de Jersey, P., and Sills, J. (2010). Ancient British Coins. Aylsham: Chris Rudd.

Britain before the Romans    177 Score, V. (2011). Hoards, Hounds and Helmets:  A  Conquest-​Period Ritual Site at Hallaton, Leicestershire. Leicester Archaeology Monograph 21. Leicester:  University of Leicester Archaeological Services. Sharples, N. (2010). Social Relations in Later Prehistory: Wessex in the First Millennium bc. Oxford: Oxford University Press. Smiles, S. (1994). The Image of Antiquity:  Ancient Britain and the Romantic Imagination. London: Yale University Press. Smith, R. A. (1912). ‘On Late-​Celtic Antiquities Discovered at Welwyn, Herts’, Archaeologia, 63: 1–​30. Stead, I. M. (1967). ‘A La Tène III Burial at Welwyn Garden City’, Archaeologia, 101: 1–​62. Stead, I. M. (1971). ‘The Reconstruction of Iron Age Buckets from Aylesford and Swarling’, in G.  d.  G. Sieveking (ed.), Prehistoric and Roman Studies. London:  British Museum, 250–​282. Stead, I. M. (1976). ‘The Earliest Burials of the Aylesford Culture’, in G.  d.  G. Sieveking, I. H. Longworth, and K. E. Wilson (eds), Problems in Economic and Social Archaeology. London: Duckworth, 401–​416. Stead, I. M. (1985). The Battersea Shield. London: British Museum Publications. Stead, I. M. (1991). ‘The Snettisham Treasure: Excavations in 1990’, Antiquity, 65: 447–​465. Stead, I. M. (2006). British Iron Age Swords and Scabbards. London: British Museum Press. Stead, I. M., and Rigby, V. (1986). Baldock: The Excavation of a Roman and Pre-​Roman Settlement, 1968–​72. Britannia Monograph Series 7. London: Society for the Promotion of Roman Studies. Stead, I. M., and Rigby, V. (1989). Verulamium:  The King Harry Lane Site. Archaeological Report 12. London: English Heritage. Stevenson, J. (2013). Living by The Sword:  The Archaeology of Brisley Farm, Ashford, Kent. Portslade: SpoilHeap Publications. Tuohy, C. (1999). Prehistoric Combs of Antler and Bone. British Archaeological Report British Series 285. Oxford: British Archaeological Reports. Tyers, P. A. (1980). ‘Correspondances entre la céramique commune La Tène III du sud-​est de l’Angleterre et du nord de la France’, Septentrion, 10: 61–​70. van Arsdell, R. D. (1989). Celtic Coinage of Britain. London: Spink. van Arsdell, R. D. (1994). The Coinage of the Dobunni: Money Supply and Coinage Circulation in Dobunnic Territory. Oxford: Oxford University Committee for Archaeology. Webley, L. (2015). ‘Rethinking Iron Age Connections across the Channel and North Sea’, in H. Anderson-​Whymark, D. Garrow, and F. Sturt (eds), Continental Connections: Exploring Cross-​Channel Relationships from the Mesolithic to the Iron Age. Oxford: Oxbow, 122–​144. Webster, J. (1996). ‘Ethnographic Barbarity: Colonial Discourse and “Celtic Warrior Societies”’, in J. Webster and N. Cooper (eds), Roman Imperialism:  Post-​ Colonial Perspectives. Leicester: School of Archaeological Studies, University of Leicester, 111–​123. Webster, J. (1999). ‘At the End of the World: Druidic and Other Revitalization Movements in Post-​Conquest Gaul and Britain’, Britannia, 30: 1–​20. Wefers, S. (2011). ‘Still Using your Saddle Quern? A Compilation of the Oldest Known Rotary Querns in Western Europe’, in D. Williams and D. Peacock (eds), Bread for the People: The Archaeology of Mills and Milling. Oxford: Archaeopress, 67–​76. Werner, J. (1954). ‘Die Bronzekanne von Kelheim’, Bayerische Vorgeschichtsblätter, 20: 43–​73. Wheeler, R. E. M., and Wheeler, T. V. (1936). Verulamium: A Belgic and Two Roman Cities. Research Report XI. London: Society of Antiquaries of London.

178

178   Timothy Champion White, R. (1991). The Middle Ground: Indians, Empires, and Republics in the Great Lakes Region, 1650–​1815. Cambridge: Cambridge University Press. Wilhelmi, K. (1977). ‘Zur Funktion und Verbreitung dreieckige Tongewichte der Eisenzeit’, Germania, 55: 180–​184. Wilhelmi, K. (1987). ‘Zur Besiedlungsgenese Englands und des nordwestlichen Kontinent von 1500 vor bis Christi Geburt’, Acta Praehistorica Archaeologica, 19: 71–​84. Williams, J. H.  C. (2001). ‘Coin Inscriptions and the Origins of Writing in Pre-​Conquest Britain’, British Numismatic Journal, 71: 1–​17. Williams, J. H. C. (2003). ‘Iron-​Age and Roman Coins’, British Numismatic Journal, 73: 44–​57. Williams, J. H. C. (2005a). ‘Coinage and Identity in Pre-​Conquest Britain: 50 bc–​ad 50’, in C. Howgego, V. Heuchert, and A. Burnett (eds), Coinage and Identity in the Roman Provinces. Oxford: Oxford University Press, 69–​78. Williams, J. H. C. (2005b). ‘ “The newer rite is here”: Vinous Symbolism on British Iron Age Coins’, in C. Haselgrove and D. Wigg-​Wolf (eds), Iron Age Coinage and Ritual Practices. Studien zu Fundmünzen der Antike. Mainz: von Zabern, 25–​41. Williams, J. H. C. (2007). ‘New Light on Latin in Pre-​Conquest Britain’, Britannia, 38: 1–​11. Willis, S. (1994). ‘Roman Imports into Late Iron Age British Societies:  Towards a Critique of Existing Models’, in S. Cottam, D. Dungworth, S. Scott, and J. Taylor (eds), TRAC 94: Proceedings of the Fourth Annual Theoretical Roman Archaeology Conference, Durham, 1994. Oxford: Oxbow, 141–​150. Woolf, G. D. (1993a). ‘Rethinking the Oppida’, Oxford Journal of Archaeology, 12: 223–​232. Woolf, G. D. (1993b). ‘The Social Significance of Trade in Late Iron Age Europe’, in C. Scarre and F. Healey (eds), Trade and Exchange in Prehistoric Europe. Oxford: Oxbow, 211–​218. Woolf, G. D. (2011). Tales of the Barbarians:  Ethnography and Empire in the Roman West. Chichester: John Wiley & Sons. Worrell, S. (2007). ‘Detecting the Later Iron Age:  A  View from the Portable Antiquities Scheme’, in C. Haselgrove and T. Moore (eds), The Later Iron Age in Britain and beyond. Oxford: Oxbow, 371–​388.

Chapter 9

Beyond Hadria n’ s  Wa l l Fraser Hunter

Introduction The edges of empires are fascinating places. Britannia’s northern frontier was a varied one, with considerable fluidity in its location until the early third century. The vagaries in frontier location and historical sources for conflict and turmoil indicate that the relationship with Rome’s northern neighbours was far from easy. Yet archaeological evidence can present other sides of this relationship, with Roman material culture finding varied uses within these Iron Age societies, while the long and often difficult relationship had a series of unexpected consequences on both sides. This chapter will start with a pen portrait of Iron Age societies before the Roman invasion, consider the military history and the wider picture of life on the frontier from the Roman perspective, before turning once more to the indigenous population and their relationship with Rome over four centuries.

The Iron Age before Rome Recent Developments The study of the Scottish Iron Age has been revolutionized over the past generation by a series of factors. Increasing numbers of sites have reliable radiocarbon sequences, especially with the advent of Bayesian statistical analysis, which has started to free the chronology from a dependence on imports and stylistic parallels; examples include the Traprain Law Environs Project (Haselgrove 2009), the debate over the dating of brochs (Dockrill et  al. 2006) and patterns of hillfort development (the so-​called Hownam sequence (Armit 1999)). The boom in developer-​funded archaeology, as elsewhere, has

180

180   Fraser Hunter vastly increased the available evidence (e.g. Bradley 2007), much of it in areas such as north-​east Scotland that had previously seen limited work, although there is still little synthesis of the results. More generally, like the rest of the British Iron Age, it has benefited from the application of an increasing range of theoretical views, not all necessarily of lasting value but providing sets of alternative perspectives to challenge and develop ways of thinking. There has been increasing concern over how exactly Iron Age societies worked and whether typical models of chieftain-​based tribal societies are appropriate (e.g. Hill 2006). This has been encouraged by the wider ‘Celts debate’, seeking to allow archaeological evidence to speak for itself rather than incorporating it into interpretations derived from an assemblage of classical and insular historical sources about groups termed ‘Celts’ by ancient or modern writers (e.g. Fitzpatrick 1996; Megaw and Megaw 1996, 1998; Sims-​Williams 1998; James 1999; Collis 2003). These debates continue and what is sketched below will be an interim picture. A more specifically Scottish debate has been over chronology and terminology. Scholars working in the north and west of the country have pushed for a ‘long Iron Age’ from c. 800 bc–​ad 800 on Scandinavian models, rather than following the English Iron Age chronology, which stops with Rome (Harding 2004: 3–​5). This has much to commend it, not least in linking periods that are often studied separately. The concept of a ‘Roman Iron Age’—​common among scholars working beyond the frontier on the continent—​is a powerful and useful one, reflecting the relationships the area had with the Roman world without treating this as a disciplinary boundary. For broader works on the Scottish Iron Age, see Hingley (1992), Harding (2004), Armit (2016), and Hunter and Carruthers (2012a).

Conceptualizing Iron Age Societies By the first centuries bc/​ad, in lowland Scotland at least the picture is one of a densely settled farming landscape, with woodland extensively cleared and remaining trees carefully husbanded (Tipping 1997; Tipping and Tisdall 2005). The settlement pattern and social framework show marked regional variation. In the south of the country, while hillforts may still dominate the landscape today, by the late pre-​Roman Iron Age the age of these classic type-​fossils of the Iron Age was past—​there is very little evidence for their construction at this time, with unenclosed clusters of houses overlying their ramparts. In the north and west, it seems that the main period of broch construction was also over. These dramatic drystone round-houses, often tower-​like in form, remained as focal points in the landscape, however, and many show long settlement sequences over a thousand years or more (for example, Howe, Orkney, and Old Scatness, Shetland (Ballin Smith 1994; Dockrill et al. 2010, 2015)). The Geography of Ptolemy has been unduly influential in shaping our understanding of Iron Age society. There have long been efforts to map his tribes onto the landscape and the archaeology, efforts that have been contentious and largely unsuccessful (see, e.g., the differing views of Richmond 1958: 131–1​46; Hind 1983; Mann and Breeze 1987;

Beyond Hadrian’s Wall   181 Barrow 1989; Fraser 2005: 34–3​7). This should not surprise us: Ptolemy is a poor guide. The sparsity of names for some areas must arise from an incomplete record, but there are more general issues that pose wider concerns over its reliability. The information is, at best, a snapshot of a complicated situation, moderated by unknown agents who may have had little or no knowledge of local politics, trying to impose a Roman idea of appropriate levels of social organization onto a situation that, from anthropological parallels elsewhere, was probably put in flux as a result of contact with empire. Such large-​ scale ‘peoples’ or ‘tribes’ are a poor basis for understanding the Scottish Iron Age; only in extreme circumstances, such as the Roman invasion, is there evidence of large-​scale collected action from the disparate units (Tacitus, Agricola 29). The emerging picture is of small-​scale societies, with the extended household the key unit in most areas: single houses or small clusters dominate our evidence. There is little evidence of larger-​scale units or centres. The exception may be in south-​eastern Scotland, where larger communities may be represented in the clusters of houses over old hillforts. This is not to see the late pre-​Roman Iron Age as a community of equals. Large parts of the population may well be archaeologically invisible—​historical parallels from the early Medieval period point to a critical difference between the free and the unfree, the latter being almost impossible to detect, while among the free there would be variations according to ownership of and access to land (Fraser 2009: 34–​35). An increasing concern with marking individual identities and differences is seen in the growing use of personal ornaments, a phenomenon noted widely across Britain (Hill 1995: 85). Emergent local and regional identities (for instance, the development of broch architecture in the Atlantic; the oblong forts of north-​east Scotland, or, later, the distinctive late Iron Age metalwork of this area) suggest some broader-​scale patterns of community from the later first millennium bc (Harding 2004: 85–​90, 108–​132; Hunter 2006a). But the model does not seem to be the classic ‘Celtic chief and his tribe’: individual households or small communities seem broadly similar, with differences perhaps arising from personal prowess or gained prestige and not necessarily persisting from one generation to the next. In southern and north-​eastern Scotland, a notable feature of the period is the increasing use of decorated metalwork, normally termed Celtic or La Tène art. This reflects the increasing concern with personal identity and status noted above, but may also be seen as a direct response to the increasing power of Rome. Across much of Britain, the late pre-​Roman and early Roman Iron Age saw an explosion of Celtic art (Garrow 2008: 30–​ 36). These societies were coming under stress, with the presence or proximity of Rome causing tensions and changes: Celtic art has been seen as a very visual, deliberately non-​ Roman response to this, perhaps emerging from the conflict situation itself, perhaps from wider social tensions and shufflings caused by Rome’s proximity (Hunter 2007a: 289; Davis and Gwilt 2008). The chronology of many changes in relation to the Roman conquest remains frustratingly vague, as there are still too few secure radiocarbon-​dated sequences at this critical time, but the move to an increasing concern with the individual starts in the second or first century bc, while at least some of these styles of northern Celtic art

182

182   Fraser Hunter pre-​date the Roman conquest, if only by a few decades (e.g. Fitzpatrick 1989). How much impact did the Roman world have before conquest? There is none of the large-​ scale use of Roman imports or influence from Roman habits seen in south-​eastern Britain in the period c. 50 bc–​ad 50 (e.g. Haselgrove 1984; Creighton 2000), nor yet the equivalent of the privileged contacts to client kings as at Stanwick (North Yorkshire (Fitts 1998)). But sparse evidence of early Mediterranean travellers to the area indicates some pre-​military contacts (Breeze 2002: 11–​12), while evidence of links to other areas of Iron Age Britain (e.g. Stevenson 1966, although within a diffusionist paradigm) meant that at least some people in the north would have heard rumours of change.

A Military History The peoples of what is now Scotland had a very direct experience of the Roman military. Motives behind Rome’s expansion have seen much discussion. Each of the various moves into Scotland was imperially directed, and the desire to conquer the whole country was certainly on the agenda in the first century and, apparently, in the early third: the second-​century invasion may have been more limited in its aims. Total conquest would have made the occupation easier in the long run, with no ‘outside’ to continue resistance. The reasons why this was not accomplished have long been debated and range from the supposed ferocity of the local people and landscape, to the lack of Roman interest in an area that did not fit strategic priorities, as well as the interplay with imperial politics. The truth is likely to lie between these. Northern Britain was always expendable: on several occasions, troops were withdrawn to deal with more pressing problems elsewhere (such as Domitian’s Dacian wars or Antoninus Pius’ Moorish wars (Hanson 1987: 152; Woolliscroft 2000; Breeze 2006a: 100–​101)). But the frontier was a restless one—​indeed, the very presence of a frontier probably created tensions that led to conflict, and military interventions were increasingly unsuccessful in dealing with the problem. This will be considered later in the chapter. Reconstruction of the history of Roman Scotland comes from a combination of literary sources and detailed archaeological work. At times the latter has been rather subservient to the former, but there is now more belief in (and argument over) the interpretation of the archaeological remains in their own right, which is adding to and questioning earlier models; some examples are mentioned later. It is worth noting, however, that the overall picture of the occupations remains partial; for instance, the distribution of Roman military sites is highly biased to eastern, arable-​producing areas that provide good cropmark evidence (Figure 9.1). The distribution of sites in the wetter, predominantly pastoral west of the country is clearly highly incomplete (e.g. Keppie 1990; Hunter and Carruthers 2012b: 10–​27). For general works on Roman Scotland see, inter alia, Breeze (1982, 2006b), Maxwell (1989, 1998), Keppie (2004), and Hunter and Carruthers (2012b).

(a)

(b)

Figure  9.1 Distribution of Roman military sites in Scotland. (a)  Flavian (c. ad 78–​86); (b) Antonine (c. ad 139–​165). Source: Reproduced by courtesy of Professor D. J. Breeze.

184

184   Fraser Hunter

The First Invasion: The Flavian Period The governor Agricola is usually credited with leading the army into Scotland, thanks to the testament of his son-​in-​law Tacitus, but recent years have seen considerable discussion over this. Long-​known evidence for earlier contacts (Birley 1946) was brought back to centre-​stage when dendrochronological dates revealed that the fort at Carlisle was built in ad 72, a good six years before Agricola headed further north. Did the army really stop here or did it move forward to conquer new land or soften the area up in the intervening years? This has seen extensive debate, focusing on pre-​or early Flavian finds and prolonged building sequences. While some of the marching camps in southern Scotland could well relate to pre-​Agricolan activity, at the moment there is no clear evidence to overturn Agricola as the prime mover (see papers in Breeze et al. 2009, with further references; Fraser 2009: 17–​18). This first invasion seems to have conquered southern Scotland quickly. The campaign paused around the Tay, perhaps because the death of Vespasian, and then Titus, required Agricola to await fresh orders (Hanson 1987: 107, 115), but Domitian ordered the legions forward once more, with the north-​east of Scotland invaded and subdued following a crushing victory at the battle of Mons Graupius. The quest for its site has intrigued scholars for centuries, but, without firm evidence of a victory monument or the discarded debris of battle, we are, in truth, no nearer to finding its location (see Maxwell 1990; Fraser 2005). The temporary camps built by troops on the move have long been a major piece of evidence for those wishing to track the movements of Agricola and subsequent generals. Recent synthesis has put this on a much firmer footing (Jones 2011: 99–​107). Likely campaign lines can be identified north of the Forth—​although not all are closely dated—​ but, to the south, the camps resolutely refuse to fall into neat series. This reflects the messier situation of large numbers of troops moving around southern Scotland over two centuries in the business of subduing and controlling it (Jones 2011: 121–​123). A striking development has been the extensive excavation of the Flavian camp at Kintore (Aberdeenshire). Such camps were long dismissed as unworthy of large-​scale excavation, but stripping of a substantial area uncovered latrine and rubbish pits, field ovens, and a wide range of finds that breathe fresh light into our picture of the army on campaign (Cook and Dunbar 2008). A notable finding was the evidence for multiple firings of some ovens, implying this was far more than just an overnight stopping point, while excavation of other sites has provided evidence of the army returning to an earlier camp with little or no discernible reworking of its defences (Dunwell and Keppie 1995). A network of control in the shape of forts and fortlets had already been placed over southern Scotland (Tacitus, Agricola 20), and was extended north in the years immediately before and after Mons Graupius. The linchpin was the legionary fortress at Inchtuthil on the river Tay, well placed to control the newly conquered area and act as a springboard for further campaigns (Pitts and St Joseph 1985; Woolliscroft and Hoffmann 2006: 62–​72). It was supported by a range of auxiliary forts and by the Gask Ridge—​a heavily supervised road variously seen as a frontier line, an element in a wider

Beyond Hadrian’s Wall   185 frontier system, or a carefully controlled supply route (Hanson 1987: 153–​ 157; Woolliscroft and Hoffmann 2006: 225–​34; Dobat 2009). Evidence for frequent structural changes on excavated sites indicates this was a rapidly evolving situation, with garrisons shifted and installations modified as needs changed. Before it could settle down—​indeed, before the Inchtuthil fortress was completed—​the decision was taken to withdraw, as up to a quarter of the British garrison (Legio II Adiutrix and probably some auxiliary units) was transferred to the continent as a result of the tense situation on the lower Danube (Hanson 1987: 152). ‘Britain was conquered and immediately abandoned’ was Tacitus’ acid comment (Histories 1.2; trans. Ireland 2008: 92). Initially, parts of southern Scotland were retained but, around ad 100, these forts were also abandoned as the frontier fell back to the Tyne–​Solway line.

A Tale of Two Walls: The Efforts of Hadrian and Antoninus Pius The decision to withdraw from the north may have been related to wider politics, but there were, clearly, local problems. Various sources indicate conflict during Hadrian’s reign—​ indeed, the building of Hadrian’s Wall may have exacerbated this, as it probably dissected pre-​existing communities and restricted contact. This running sore of conflict probably influenced Antoninus Pius’ decision to reconquer southern Scotland on his accession and build another wall, along with wider motives such as his desire to show he was a worthy commander in chief (Breeze 2006a: 12–​14). Both wall systems have seen enormous amounts of detailed scholarship and, in recent years, innovative approaches and exciting new results and debate over their planning, perceived role, and actual function (for Hadrian’s Wall, see Hill 2004; Breeze 2006c; Hodgson 2009; Poulter 2009; Symonds and Mason 2009; Graafstal 2012; for the Antonine Wall, Swan 1999; Robertson and Keppie 2001; Breeze 2006a; Poulter 2009; Graafstaal et al. 2015). In the case of the Antonine Wall, not only did its plan evolve but the building process was a protracted one (Gillam 1976a; Swan 1999; Breeze 2006a: 97–​102). Indeed, it may barely have been finished when the decision was taken to abandon it. Evidence increasingly suggests that the Wall was under-​garrisoned (Keppie 2009a), implying that the reconquest of Scotland had overstretched the army. The issue of resistance remains under discussion. The old view of two separate Antonine occupations, split by a revolt among the Brigantes, has been convincingly dismissed (Hodgson 1995), but there are signs of localized resistance, notably in Dumfriesshire (Hodgson 2009), and this may have been another factor in the decision to return to Hadrian’s Wall.

A Restless Frontier The sparse historical sources point to continuing troubles north of Hadrian’s Wall, leading eventually to the British expedition of Septimius Severus and Caracalla in ad 208–​11 (Breeze 1982: 128–​136; Birley 1988: 170–​187). These sources stress a military solution, but

186

186   Fraser Hunter archaeological evidence can cast a different light on things. There is a strong cluster of silver coin hoards at this date in and around the troublesome areas, and work at the indigenous site of Birnie (Moray) has provided an excavated context: here, two such hoards came to a local power centre in the 190s ad as part of a long-​running relationship to the Roman world. This has been seen as a system of gifts, subsidies, or bribes, depending on one’s perspective (Hunter 2007b, 2007c, 2015a; Holmes 2014). The potential effects of this will be considered in a later section. There is no further mention of conflict in the third century, but the fourth century produced a litany of troubles and violence caused by a group now termed ‘Picts’, with various military responses and expeditions to try to deal with them (conveniently summarized in Maxwell 1987: 43). This issue too is considered later, but in terms of Roman military archaeology it has left little or no recognizable trace, apart, arguably, from a scatter of late Roman finds at earlier fort sites (Hunter and Carruthers 2012b: 25).

Frontier Lives Until recently, research in the military zone could readily be caricatured as obsessed with troop units, campaigns, and military strategy. A rare exception was the remarkable excavation at Newstead, near Melrose (Borders) by James Curle. His report drew together structures, finds, and environmental evidence to present a picture of ‘a Roman frontier post and its people’, with the people a key part of the story (Curle 1911; Hunter and Keppie 2012). Only recently has academic focus returned to such a broad understanding of life on the Roman frontier. Newstead stressed the need to look beyond the fort walls in order to understand the Roman period. Unusually large numbers of British fort sites are equipped with annexes: there is an ongoing debate as to whether these should be seen as essentially military, for instance, as supply depots and industrial zones, or as military vici for the camp followers (Sommer 1984: 18–​22; 2006; Bailey 1994; Hanson 2007: 667–66​8). Only limited excavations have yet been carried out in them. At Newstead, geophysical survey followed up by targeted excavation produced a picture of a landscape rich in activity all round the fort, with indications of different functions in different areas (e.g. Clarke 2000). Inveresk is the only other Scottish fort to have seen substantial work beyond its walls (Figure 9.2). It seems to have an unenclosed civil settlement rather than a defended annexe: excavations have revealed a street grid, industrial activities, and indications of burials and religious sites (e.g. Bishop 2002, 2004). Much of this work has been conducted through developer-​funded archaeology; there has been little attempt yet to compare lifestyles and activities in different excavated areas, but the potential is considerable. Cemeteries and ritual sites, by contrast, are all but unknown. Only burial areas at Camelon and Inveresk have seen some limited work (Breeze and Rich-​Gray 1980; Chapman et al. 2012: 287–​288); otherwise the picture is dominated by a few stray finds and sculpture (Collard and Hunter 2000). The distribution of altars beyond forts may

Beyond Hadrian’s Wall   187

Figure 9.2  The Antonine military complex at Inveresk (East Lothian). Source: Drawing by Alan Braby.

offer clues to the sacred landscape, but this was rather obscured by earlier scholarship’s habit of gathering them under the fort’s umbrella (Collingwood and Wright 1965). The enigmatic and long-​destroyed ‘Arthur’s O’on’ was probably a temple (Steer 1958), and further examples must lurk around forts (as perhaps the now-​destroyed platform at Easter Langlee, near Newstead (Steer 1966); a Mithraeum turned up unexpectedly in development work near Inveresk in 2010 (Hunter et al. forthcoming)). Some evidence of ritual deposits has been discussed. Evidence for on-​site deposition, for instance in

188

188   Fraser Hunter the Newstead wells, remains hotly contested (Clarke and Jones 1994; cf. Manning 2006). There were also deposits of valued material (such as bronze vessels) in wet locations, some long hallowed by votive use, but whether these should be linked to beliefs of soldiers or habits of the locals remains for debate (Hunter 1997: 117–​118). Evidence of craft activities taking place outside the fort and the field systems that surround a number of them are reminders of the key role of supply and logistics. This has traditionally been approached through the proxy of pottery, with evidence for a changing balance of imports and local production (e.g. Gillam 1976b; Hartley 1976; Breeze 1986; Swan 1988). An integrated, synthetic view of patterns of production and supply would be highly desirable. In terms of the economic resources that the newly conquered area could have provided, our evidence remains weak. ‘Invisible exports’ such as fur, slaves, wild animals, and agricultural produce are the usual (and plausible) candidates but hard to demonstrate, although stable isotope work on animal bones might ultimately provide evidence of droving or similar extended chains of meat procurement practices (Stallibrass 2009). Another key question is the role of metal resources. Within the conquered area a number of sources of copper, lead, silver, and gold are known, but with little direct evidence for their exploitation. Analysis of a lead ingot from Strageath (Perth and Kinross) matched a Southern Uplands source (Hunter 2006b: 85), providing a tantalizing hint of Roman working of the Leadhills/​Wanlockhead area. Extensive modern exploitation may have removed or concealed direct evidence, but proxy records of smelting in peat cores (cf. Mighall 2003) offer a way forward. The potential of the material culture and its distribution to explore different lifeways on forts and their surrounding areas is largely untapped. Artefact distribution as a key to social use of space has seen useful application on other frontiers (e.g. Allison 2006) and could usefully be tackled with some of the Scottish data (see Giles 2012). Clarke (2000) has begun to use the data from different parts of the Newstead complex to point to different lifestyles reflected in differing access and approaches to material culture, and there is enormous potential to explore this further, especially as developer-​funded work offers ready samples from different areas of sites. One way to investigate changing lifeways in this area is by studying the development of hybrid forms of material culture, which is seen across the northern frontier zone spanning Hadrian’s Wall. The development of Romano-​British brooches, often showing so-​called Celtic design elements, has long been recognized (e.g. Collingwood 1930), and the regional nature of such brooch distributions is increasingly clear thanks to data from the Portable Antiquities Scheme (e.g. McIntosh 2011). Other object categories similarly can offer insights into this, such as the wide range of copper-​alloy objects developing from or influenced by indigenous traditions (for example, beaded torcs or horse harness (Hunter 2008, 2010a, 2015b)), through the persistence of Iron Age types of worked bone objects (such as long-​handled combs), to newly emerging forms such as glass bangles, typical of the frontier zone in the Roman Iron Age. Standard works have focused on typology and distribution (Kilbride-​Jones 1938; Stevenson 1956a, 1976; Price 1988), and there is scope for much more consideration of how they were made and used across the different frontier communities (e.g. Hoffmann 2003).

Beyond Hadrian’s Wall   189 Such studies of material culture in context offer ways to study the different communities of the frontier zone, how they perceived themselves and others and how they shaped this—​a topic that has seen much interest in recent years on a broad level, but that leaves lots of scope for detailed analysis (e.g. Goldsworthy and Haynes 1999; James 2002; Mattingly 2006).

An Iron Age Perspective The Romans did not march into an empty landscape. What impacts did the invasion and proximity of the army have on the local populations? The answers are complex and varied, but ongoing research is providing a better-​rounded picture. The violence and resistance that feature in the historical sources are hard to trace archaeologically, as is often the case. The much-​debated complex at Burnswark (Dumfries and Galloway), where the Iron Age hillfort is threatened by two Roman camps, may provide one example of a violent climax to an act of resistance; opinion remains divided (Jobey 1978; Campbell 2003; Keppie 2009b; Breeze 2011), though recent fieldwork seems to support the siege theory (John Reid, pers. comm). Parallels elsewhere suggest we should expect a range of responses to Rome, from support through acceptance to opposition, both between communities in different areas and within different segments of the one community. This is hard to trace archaeologically. Our main evidence of interaction comes from the discovery of Roman objects on indigenous sites, a topic with a long history of study (e.g. Curle 1932; Robertson 1970; Macinnes 1984; Hunter 2001, 2007b; Bateson and Holmes 2006, with previous references; Wilson 2010; Campbell 2012). These are not without their problems. In contrast to areas beyond the German frontier, the general lack of burials means that we are presented only with fragments from settlement sites, and their value has often been overlooked because of this. There has been extended debate over the significance of such fragments: did they come to the sites whole, or as tokens or talismans, when did they arrive, and how long were they used for? (See Hunter 2007b: 11, 91 n. 3, for references.) While there is clear evidence that some objects had an extended life, especially samian sherds (which were prioritized for curation and reuse (Campbell 2012)), patterns visible in the data (explored in the next section) make sense when the objects are considered as vessels rather than sherds, suggesting this is how they arrived and were initially used. Improved understanding of the social roles of these objects will come from a broad, life-​cycle approach (cf. Kopytoff 1986), which considers the object’s use-​lives (as whole item and fragment) and deposition in the context of local material culture; the fruits of such work are beginning to appear (Wallace 2006; Campbell 2012, 2015; Hunter 2013). It is difficult to assess whether finds arrived predominantly when southern Scotland was directly occupied (as Erdrich et al. 2000 argue) or whether there was a more constant flow from the province lying to the south. This key question of whether and when the frontier was open or closed for contact is currently hard to answer. We are caught in a vicious circle, where the items that can be most closely dated (coins and samian)

190

190   Fraser Hunter are the ones suspected to have long use-​lives, and thus the production date is a poor reflection of the deposition date. As the data set grows and more independent dates are obtained, this should become easier to analyse. This issue depends also on the mechanisms and motives behind contacts, whether the results of violence (such as booty) or peaceful contact (for example, trade, diplomacy, military service). The former seems unlikely as a widespread process, especially when the material is contrasted to the much more random and metal-​focused debris increasingly found beyond the German frontier, where it is interpreted as loot for reuse (e.g. Becker 2006).

Selection Processes Research has sought ways to alleviate the taphonomic bias imposed by this settlement debris by looking at techniques such as presence/​absence data (rather than absolute numbers) or the range of finds as an indication of variations in access (Hunter 2001). This has revealed a strong selectivity in the material coming into indigenous hands, with a marked preference for jewellery (especially brooches) and for feasting and drinking gear—​in other words, material that could be used in local social displays. Clear signs of differential access to this material, both regionally and socially, provide clues to the social networks of the period (Figure 9.3). In southern Scotland at least, Roman material was not absolutely rare, but more is found on sites that are richer in other senses as well—​for instance, in access to restricted craft skills such as bronze-​working, or the presence of decorated metalwork or other imported goods. It suggests that Roman contact was mediated through higher-​status individuals and groups. This is most marked in south-​east and east-​central Scotland, suggesting the presence or development of more hierarchical or competitive societies in these areas. Elsewhere there is a much more equal pattern of access, matching the picture of relatively small-​scale societies already discussed. Roman material was valued and used in regions far beyond the area of direct occupation, reaching out to the Northern and Western Isles; indeed, small quantities may have had disproportionate impact in areas of more restricted access, but this has seen little detailed study.

Focal Sites One set of sites that are particularly rich in Roman finds are the ‘lowland brochs’ (Macinnes 1984). The broch, a drystone tower-​like round-house, was typical of the Atlantic north and west, but a small number are found in central and southern Scotland. Most southern excavated examples are of first–​second-​century ad date, and many were rich in finds, including a wealth of Roman material culture. The exotic architectural form was a clear mark of social difference, and this is confirmed in the finds, with preferential access to a wide range of Roman objects as well as plentiful evidence of restricted craft skills and rare objects. Most of these brochs were not new foundations, developing

(b)

(a) 0 0

100 km

0 0

100 miles

100 km 100 miles

Birnie

Traprain Law

(c) 0 0

100 km 100 miles

Figure 9.3  Distribution of Roman finds from non-​Roman sites north of Hadrian’s Wall. (a) overall distribution; (b) middle Roman Iron Age c. ad 160–​250 (the sites of Birnie and Traprain Law, mentioned in the text, are marked: *represents find-spots of denarius hoards); (c) late Roman Iron Age c. ad 250–​400; *marks the major centres of Traprain Law, Edinburgh Castle, Eildon Hill, and Dumbarton Rock. Source: © Fraser Hunter.

192

192   Fraser Hunter on existing settlement sites. However, modern excavations suggest the brochs were a boom-​and-​bust phenomenon with little sign of extended use, and a number show evidence of deliberate destruction or demolition. This suggests a rather unstable situation where the influence of competing power-​hungry groups could rise and fall rapidly; perhaps because of changing relationships to the fickle Roman world, perhaps from changing local power politics and varying views on whether Rome was a friend or foe. One site stands out for its long-​ term, sustained relationship with the Roman world: Traprain Law in East Lothian (Jobey 1976; Hunter 2009). The defences of this mighty hillfort were denuded by the Roman period, but the site was densely inhabited—​ indeed, current evidence suggests that it was a Roman-​period ‘boom town’ with little immediately pre-​Roman settlement, and its rise to prominence was strongly connected to the links its occupants built to the Roman world. These long-​lasting connections brought a wealth of Roman material culture to the site from the first to the fifth century ad. Along with the very light Roman military presence in the area, it suggests this area acted as a client kingdom over a long period. Traprain stands alone at the moment, but there may be other such sites, especially in southern Scotland, yet to be recognized.

The Impact of Frontiers The differing impact of the Roman occupation in different areas is ripe for analysis, both in comparison to the continental frontier and within Britain. For instance, it is striking that the zone in front of Hadrian’s Wall, in north Northumberland, lacks indigenous sites rich in Roman finds, in contrast to south-​east Scotland. It suggests that the occupation had a dire effect on local societies in this area, especially the traditional power structures, with the army sitting at the top of the local social tree (e.g. Hodgson et al. 2012: 211–220). In contrast, further from Hadrian’s Wall (or around the Antonine Wall, according to period), more Roman material was more widely spread, with clearer indications of social hierarchies of access to it. The clearest examples are the lowland brochs, suggesting groups growing powerful on the back of Rome. As noted, these may have been a short-​lived phenomenon, but other sites show continuing, rich, Roman links in places beyond Hadrian’s Wall, especially in south-​east and north-​east Scotland. A regional picture of this varying impact of empire remains to be developed.

Diplomacy and Interference Access to and use of Roman material was not a neutral phenomenon; it deeply affected the people and societies involved. The precise sequence of cause and effect is rarely clear, but a series of social phenomena of the Roman Iron Age can plausibly be traced to the agency of Rome. Some may have been incidental byproducts of contact, but increasingly it seems that the Roman world was deliberately creating and manipulating links with local societies for political purposes, as it did on other frontiers. For north-​west Germany,

Beyond Hadrian’s Wall   193 Erdrich (2000) has argued that the Romans sought to control and foment unrest within local society by building some groups up at the expense of others, or by giving and then withdrawing favours. There are signs of similar patterns in the Scottish record. The best evidence comes from the mid–​late Roman Iron Age (c. ad 160–​400+; figure 9 b–c). In the mid-​Roman Iron Age (c. ad 160–​250), the distribution of silver coin hoards concentrated strongly in central and north-​eastern Scotland has been argued to reflect deliberate targeting of troublesome tribes and their neighbours (Hunter 2007b: 23–2​7, 2015a; Holmes 2014). The phenomenon is represented historically in references to unrest, especially among the Caledonians and Maeatae (e.g. Epitome of Dio Cassius LXXV, 5, 4; trans. Ireland 2008: 112–​113). Subsidies of denarii may have been the main diplomatic vehicle, but a range of other rich finds was targeted at the same area. Precise correlation with historical events is hazardous, but, in general terms, these can be seen as attempts to solve frontier unrest through diplomatic efforts; the subsequent Severan invasion suggests that this was not entirely successful, although a few denarius hoards in central and southern Scotland post-​dating the invasions indicate that the policy continued, targeted now perhaps at a ‘buffer zone’ south of the unfriendly area (Holmes and Hunter 1997). There are striking changes in patterns of access to Roman material in the succeeding centuries. The north-​east, which had been highly favoured, was essentially cut off from Roman contact in the later third and fourth centuries. Other areas, especially south-​east Scotland, saw continuing access, but now much more focused on a small number of sites. This perhaps suggests that social control over this material was much tighter—​and, arguably, that a small number of groups were becoming pre-​eminent (Hunter 2007b: 32–3​6; 2010b; 2014). The effects of this are much debated. In the north-​east it is generally agreed that Roman interference was ultimately responsible for the emergence of the ‘Picts’, but opinion is divided on whether this represents a coalescence of existing groups in the face of Roman threat (Mann 1974; Fraser 2009: 44–​49, 54–​61), or a more fundamental social collapse and the rise of altogether more hostile groups within society (Hunter 2007b: 50–​53). Roman contacts ran beyond the formal ‘end’ of Roman Britain. Traprain Law’s links with Rome flourished in the late Roman period, and the spectacular hoard of fifth-​ century ‘Hacksilber’ shows the value of material to which its inhabitants had access (Curle 1923). Earlier views of this as loot are now questioned, and the phenomenon is better related either to diplomatic gifts or to the payment of mercenaries serving in the late Roman army (Painter 2013).

The Long Term What was the lasting effect of the Roman world on Scotland? Some have seen the period as little more than ‘brief interludes’ (e.g. Hanson 2003), while others have argued for much more fundamental effects (Fraser 2009: 116–​117). The latter seems closer to what can be observed at a number of levels.

194

194   Fraser Hunter There were practical benefits, in terms of raw materials: much of the non-​ferrous metalwork of the Roman Iron Age reused Roman metal, for instance, while Roman ‘Hacksilber’ is argued to be the ultimate source of the wealth of early Medieval Scotland (Stevenson 1956b; Youngs 2013). Something of the social effects have been highlighted above: the emergence of the ‘Picts’ and, arguably, the development of larger-​scale social units in southern Scotland, can both be connected with Roman interference and contacts. The precise mechanisms merit more attention, but it seems that the ‘social shape’ of both Roman Iron Age and early Medieval societies was fundamentally affected by the Roman world, whether by accident or by design. An intriguing issue, but one that has received little attention, is a long-​term perspective on forts and their settings. How far did they influence the landscape in later centuries? The temporary camp at Kintore seems to have disrupted landscape use, as the interior was apparently left empty until the seventh century (Cook and Dunbar 2008: 354–​356). Surprisingly few fort sites have produced clear evidence for reuse before the high Medieval period, although the frequent instances of Medieval churches positioned on forts may have early Medieval roots (Hunter and Carruthers 2012b: 42; see also Maldonado 2015). The topic merits more study. Rome also became a model for the ‘barbarian world’ in the post-​Roman period, in Scotland as in much of the rest of early Medieval Europe. This was tied up with the adoption of Christianity, but Latin names of local leaders, the use of Latin in inscriptions, and perhaps a continuing taste for imported Mediterranean goods in some areas point to a lasting impact (Mann 1974; Campbell 2007: 134–​135; Fraser 2013).

Conclusions The study of Roman Scotland has moved a long way from its earlier military focus. Many aspects of older agendas remain valid, such as a more detailed understanding of the disposition of military sites, but recent work has begun to look much more broadly at the period. Rome must be seen not just as a military force but in terms of supply needs, impact on the landscape, and the different segments that made up a military community. The locals are not just opponents or background, but a key element in studying the frontier, affected by and affecting the military presence. The episodic nature of direct Roman occupation of Scotland, and the restricted geographical area, mean that there is tremendous potential for exploring complex issues of frontier control, supply, impact, and interaction.

Acknowledgements Co-​chairing the development of a research framework into Roman Scotland, an initiative of the Society of Antiquaries of Scotland, was an opportunity that forced me to

Beyond Hadrian’s Wall   195 think more broadly about the topic, and I am grateful to the colleagues who served on the panel and others who commented on the document for a fascinating and thought-​ provoking process (see Hunter and Carruthers 2012b). It has influenced the contents of this chapter. Earlier versions benefited greatly from the comments of David Breeze and David Clarke.

References Primary Sources Tacitus, Agricola, trans. A. R. Birley. Oxford: Oxford University Press, 1999. Tacitus, Histories, trans. K. Wellesley. Harmondsworth: Penguin, 1964.

Secondary Sources Allison, P. (2006). ‘Mapping for Gender: Interpreting Artefact Distribution in Roman Military Forts in Germany’, Archaeological Dialogues 13/​1: 1–​48. Armit, I. (1999). ‘Life after Hownam: The Iron Age in South-​East Scotland’, in B. Bevan (ed.), Northern Exposure: Interpretative Devolution and the Iron Ages in Britain. Leicester: School of Archaeological Studies, University of Leicester, 65–​79. Armit, I. (2016). Celtic Scotland: Iron Age Scotland in its European Context. 3rd edn. Edinburgh: Birlinn. Bailey, G. B. (1994). ‘The Provision of Fort-​Annexes on the Antonine Wall’, Proceedings of the Society of Antiquaries of Scotland, 124: 299–​314. Ballin Smith, B. (1994). Howe:  Four Millennia of Orkney Prehistory. Edinburgh:  Society of Antiquaries of Scotland. Barrow, G. W. S. (1989). ‘The Tribes of North Britain Revisited’, Proceedings of the Society of Antiquaries of Scotland, 119: 161–​163. Bateson, J. D., and Holmes, N. M. McQ. (2006). ‘Roman and Medieval Coins Found in Scotland, 2001–​2005’, Proceedings of the Society of Antiquaries of Scotland, 136: 161–​198. Becker, M. (2006). ‘Handelsgut, Sold und Beute’, in J. Stadler, B. Hanemann, M. Kolb, and R. Petrovszky (eds), Geraubt und im Rhein versunken: Der Barbarenschatz. Speyer: Historisches Museum der Pfalz, 219–​223. Birley, A. (1988). The African Emperor: Septimius Severus. London: Batsford. Birley, E. (1946). ‘Britain under the Flavians: Agricola and his Predecessors’, Durham University Journal (June), 79–​84. Bishop, M. C. (2002) (ed.). Roman Inveresk: Past, Present and Future. Duns: Armatura Press. Bishop, M. C. (2004). Inveresk Gate: Excavations in the Roman Civil Settlement at Inveresk, East Lothian, 1996–​2000. Edinburgh: STAR. Bradley, R. (2007). The Prehistory of Britain and Ireland. Cambridge:  Cambridge University Press. Breeze, D. J. (1982). The Northern Frontiers of Roman Britain. London: Batsford. Breeze, D. J. (1986). ‘The Manufacture of Pottery in Roman Scotland’, Proceedings of the Society of Antiquaries of Scotland, 116: 185–​189. Breeze, D. J. (2002). ‘The Ancient Geography of Scotland’, in B. Ballin Smith and I. Banks (eds), In the Shadow of the Brochs. Stroud: Tempus, 11–​14.

196

196   Fraser Hunter Breeze, D. J. (2006a). The Antonine Wall. Edinburgh: John Donald. Breeze, D. J. (2006b). Roman Scotland: Frontier Country. 2nd edn. London: Batsford. Breeze, D. J. (2006c). J. Collingwood Bruce’s Handbook to the Roman Wall. 14th edn. Newcastle upon Tyne: Society of Antiquaries of Newcastle upon Tyne. Breeze, D. J. (2011). ‘Burnswark: Roman Siege or Army Training Ground?’, Archaeological Journal, 168: 166–​180. Breeze, D. J., and Rich-​Gray, D. (1980). ‘“Fire Pits” at Camelon, Stirlingshire’, Proceedings of the Society of Antiquaries of Scotland, 110 [1978–​80], 513–​517. Breeze, D. J., Thoms, L. M., and Hall, D. W. (2009) (eds). First Contact: Rome and Northern Britain. Perth: Tayside and Fife Archaeological Committee. Campbell, D. B. (2003). ‘The Roman Siege of Burnswark’, Britannia, 34: 19–​33. Campbell, E. (2007). Continental and Mediterranean Imports to Atlantic Britain and Ireland, c. ad 400–​800. York: Council of British Archaeology Research Report 157. Campbell, L. (2012). ‘Modifying Material: Social Biographies of Roman Material Culture’, in A. Kyle and B. Jervis (eds), Make Do and Mend: The Archaeologies of Compromise. Oxford: Archaeopress/​BAR International Series 2408, 11–​25. Campbell, L. (2015). ‘Interfacing with the Empire: Materiality Matters’, in D. B. Breeze, R. H. Jones, and I. A. Oltean (eds), Understanding Roman Frontiers: A Celebration for Professor Bill Hanson. Edinburgh: Birlinn, 182–195. Chapman, E. M., Hunter, F., Booth, P., Wilson, P., Worrell, S., and Tomlin, R. S. O. (2012). ‘Roman Britain in 2011’, Britannia, 43: 271–​421. Clarke, S. (2000). ‘In Search of a Different Roman Period: The Finds Assemblage at the Newstead Military Complex’, in G. Fincham, G. Harrison, R. Holland, and L. Revell (eds), TRAC 99: Proceedings of the 9th Theoretical Roman Archaeology Conference, Durham, April 1999. Oxford: Oxbow Books, 22–​29. Clarke, S., and Jones, R. F. J. (1994). ‘The Newstead Pits’, Journal of Roman Military Equipment Studies, 5: 109–​124. Collard, M., and Hunter, F. (2000). ‘A Roman Burial from Cramond (Edinburgh) Rediscovered’, Proceedings of the Society of the Antiquaries of Scotland, 130: 525–​535. Collingwood, R. G. (1930). ‘Romano-​Celtic Art in Northumbria’, Archaeologia, 80: 37–​58. Collingwood, R. G., and Wright, R. P. (1965). The Roman Inscriptions of Britain. I: Inscriptions on Stone. Oxford: Oxford University Press. Collis, J. (2003). The Celts: Origins, Myths and Inventions. Stroud: Sutton. Cook, M., and Dunbar, L. (2008). Rituals, Roundhouses and Romans: Excavations at Kintore, Aberdeenshire 2000–​2006. Volume 1: Forest Road. Edinburgh: STAR. Creighton, J. (2000). Coins and Power in Late Iron Age Britain. Cambridge:  Cambridge University Press. Curle, A. O. (1923). The Treasure of Traprain. Glasgow: Maclehose. Curle, J. (1911). A Roman Frontier Post and its People: The Fort of Newstead in the Parish of Melrose. Glasgow: Maclehose. Curle, J. (1932). ‘An Inventory of Objects of Roman and Provincial Roman Origin Found on Sites in Scotland not Definitely Associated with Roman Constructions’, Proceedings of the Society of Antiquaries of Scotland, 66 [1931–​2], 277–​397. Davis, M., and Gwilt, A. (2008). ‘Material, Style and Identity in First Century AD Metalwork, with Particular Reference to the Seven Sisters Hoard’, in D. Garrow, C. Gosden, and J. D. Hill (eds), Rethinking Celtic Art. Oxford: Oxbow Books, 146–​184.

Beyond Hadrian’s Wall   197 Dobat, E. (2009). ‘The Gask “system” in Perthshire: The First Artificial Frontier Line of the Roman Empire’, in D. J. Breeze, L. M. Thoms, and D. W. Hall (eds), First Contact: Rome and Northern Britain. Perth: Tayside and Fife Archaeological Committee, 39–​48. Dockrill, S. J., Outram, Z., and Batt, C. M. (2006). ‘Time and Place: A New Chronology for the Origin of the Broch Based on the Scientific Dating Programme at the Old Scatness Broch, Shetland’, Proceedings of the Society of Antiquaries of Scotland, 136: 89–​110. Dockrill, S. J., Bond, J. M., Turner, V. E., Brown, L. D., Bashford, D. J., Cussans, J. E., and Nicholson, R. A. (2010). Excavations at Old Scatness, Shetland. Volume 1: The Pictish Village and Viking Settlement. Lerwick: Shetland Heritage Publications. Dockrill, S. J., Bond, J. M., Turner, V. E., Brown, L. D., Bashford, D. J., Cussans, J. E., and Nicholson, R. A. (2015). Excavations at Old Scatness, Shetland. Volume 2: The Broch and Iron Age Village. Lerwick: Shetland Heritage Publications. Dunwell, A. J., and Keppie, L. J.  F. (1995). ‘The Roman Temporary Camp at Dunning, Perthshire: Evidence from two Recent Excavations’, Britannia, 26: 51–​62. Erdrich, M. (2000). Rom und die Barbaren: Das Verhältnis zwischen dem Imperium Romanum und den Germanischen Stämmen vor seiner Nordwestgrenze von der späten römischen Republik bis zum Galischen Sonderreich. Mainz: Römisch-​Germanische Kommission. Erdrich, M., Giannotta, K. M., and Hanson, W. S. (2000). ‘Traprain Law:  Native and Roman on the Northern Frontier’, Proceedings of the Society of Antiquaries of Scotland, 130: 441–​456. Fitts, L. (1998). ‘Inde opes et rerum secundarum luxus, Stanwick and Melsonby’, in J. Bird (ed.), Form and Fabric: Studies in Rome’s Material Past in Honour of B. R. Hartley. Oxford: Oxbow Books, 1–​7. Fitzpatrick, A. P. (1989). ‘The Submission of the Orkney Islands to Claudius: New Evidence?’, Scottish Archaeological Review, 6: 24–​33. Fitzpatrick, A. P. (1996). ‘“Celtic” Iron Age Europe: The Theoretical Basis’, in P. Graves-​Brown, S. Jones, and C. Gamble (eds), Cultural Identity and Archaeology: The Construction of European Communities. London: Routledge, 238–​255. Fraser, J. E. (2005). The Roman Conquest of Scotland: The Battle of Mons Graupius ad 84. Stroud: Tempus. Fraser, J. E. (2009). From Caledonia to Pictland:  Scotland to 795. Edinburgh:  Edinburgh University Press. Fraser, J. E. (2013). ‘St Patrick and Barbarian Northern Britain in the Fifth Century’, in F. Hunter and K. Painter (eds), Late Roman Silver:  The Traprain Treasure in Context. Edinburgh: Society of Antiquaries of Scotland, 15–​27. Garrow, D. (2008). ‘The Time and Space of Celtic Art:  Interrogating the “Technologies of Enchantment” Database’, in D. Garrow, C. Gosden, and J. D. Hill (eds), Rethinking Celtic Art. Oxford: Oxbow Books, 15–​39. Giles, R. D. (2012). Roman Soldiers and the Roman Army:  A  Study of Military Life from Archaeological Remains. Oxford: Archaeopress/​BAR British Series 562. Gillam, J. P. (1976a). ‘Possible Changes in Plan in the Course of Construction of the Antonine Wall’, Scottish Archaeological Forum, 7: 51–​56. Gillam, J. P. (1976b). ‘Coarse Fumed Ware in North Britain and Beyond’, Glasgow Archaeological Journal, 4: 57–​80. Goldsworthy, A., and Haynes, I. (1999) (eds). The Roman Army as a Community. Portsmouth, RI: Journal of Roman Archaeology Supplementary Series 34.

198

198   Fraser Hunter Graafstal, E. P. (2012). ‘Hadrian’s Haste: A Priority Programme for the Wall’, Archaeologia Aeliana, 5th ser. 41: 123–​184. Graafstaal, E., Breeze, D. J., Jones, R. H. and Symonds, M. F. A. (2015). ‘Sacred Cows in the Landscape: Rethinking the Planning of the Antonine Wall’, in D. J. Breeze, R. H. Jones, and I. A. Oltean (eds), Understanding Roman Frontiers: A Celebration for Professor Bill Hanson. Edinburgh: Birlinn, 54–69. Hanson, W. S. (1987). Agricola and the Conquest of the North. London: Batsford. Hanson, W. S. (2003). ‘The Roman Presence: Brief Interludes’, in K. J. Edwards and I. B. M. Ralston (eds), Scotland after the Ice Age: Environment, Archaeology and History 8000 bc–​ad 1000. Edinburgh: Edinburgh University Press, 195–​216. Hanson, W. S. (2007). Elginhaugh:  A  Flavian Fort and its Annexe. London:  Society for the Promotion of Roman Studies. Harding, D. W. (2004). The Iron Age in Northern Britain. London: Routledge. Hartley, K. F. (1976). ‘Were Mortaria Made in Roman Scotland?’, Glasgow Archaeological Journal, 4: 81–​89. Haselgrove, C. (1984). ‘“Romanization” before the Conquest: Gaulish Precedents and British Consequences’, in T. F. C. Blagg and A. C. King (eds), Military and Civilian in Roman Britain. Oxford: ​BAR British Series 136, 5–​63. Haselgrove, C. (2009). The Traprain Law Environs Project: Fieldwork and Excavations 2000–​ 2004. Edinburgh: Society of Antiquaries of Scotland. Hill, J. D. (1995). ‘The Pre-​Roman Iron Age in Britain and Ireland (ca. 800 bc to ad 100): An Overview’, Journal of World Prehistory, 9/​1: 47–​98. Hill, J. D. (2006). ‘Are we any Closer to Understanding how Iron Age Societies Worked (or did not Work)?’, in C. Haselgrove (ed.), Celtes et Gaulois, l’archéologie face à l’histoire 4: Les Mutations de la fin de l’âge du Fer. Glux-​en-​Glenne: Bibracte, 169–​179. Hill, P. R. (2004). The Construction of Hadrian’s Wall. Oxford: Archaeopress/​BAR British Series 375. Hind, J. G.  F. (1983). ‘Caledonia and its Occupation under the Flavians’, Proceedings of the Society of Antiquaries of Scotland, 113: 373–​378. Hingley, R. (1992). ‘Society in Scotland from 700 bc to ad 200’, Proceedings of the Society of Antiquaries of Scotland, 122: 7–​53. Hodgson, N. (1995). ‘Were there Two Antonine Occupations of Scotland?’, Britannia, 26: 29–​49. Hodgson, N. (2009). ‘The Abandonment of Antonine Scotland:  Its Date and Causes’, in W. S. Hanson (ed.), The Army and Frontiers of Rome. Portsmouth, RI: Journal of Roman Archaeology Supplementary Series 74, 185–​193. Hodgson, N. (2009) (ed.). Hadrian’s Wall 1999–​2009. Kendal: Cumberland and Westmorland Antiquarian and Archaeological Society/​Society of Antiquaries of Newcastle upon Tyne. Hodgson, N., McKelvey, J., and Muncaster, W. (2012). The Iron Age on the Northumbrian Coastal Plain: Excavations in Advance of Development 2002–2010. Newcastle: TWM Archaeology/Arbeia Society. Hoffmann, B. (2003). ‘Roman Glass from Newstead and Vindolanda’, Annales du 15e Congrès de l’Association Internationale pour l’Histoire du Verre, 2001New York 2001. ​Nottingham: Association Internationale pour l’Histoire de Verre, 41–​44. Holmes, N. M. MCQ. (2014). ‘The Synton and Kippilaw Denarius Hoards: Further Numismatic Evidence for Late Antonine and Severan Scotland,’ Proceedings of the Society of Antiquaries of Scotland, 144, 138–168.

Beyond Hadrian’s Wall   199 Holmes, N., and Hunter, F. (1997). ‘Edston, Peebles-​shire’, in R. Bland and J. Orna-​Ornstein (eds), Coin Hoards from Roman Britain X. London:  Trustees of the British Museum, 149–​168. Hunter, F. (1997). ‘Iron Age Hoarding in Scotland and Northern England’, in A. Gwilt and C. Haselgrove (eds), Reconstructing Iron Age Societies. Oxford: Oxbow Books, 108–​133. Hunter, F. (2001). ‘Roman and Native in Scotland:  New Approaches’, Journal of Roman Archaeology, 14: 289–​309. Hunter, F. (2006a). ‘New Light on Iron Age Massive Armlets’, Proceedings of the Society of Antiquaries of Scotland, 136: 135–​160. Hunter, F. (2006b). ‘Recent Finds from Strageath Roman Fort’, Tayside and Fife Archaeological Journal, 12: 81–​88. Hunter, F. (2007a). ‘Artefacts, Regions and Identities in the Northern British Iron Age’, in C. C. Haselgrove and T. Moore (eds), The Later Iron Age in Britain and Beyond. Oxford: Oxbow Books, 286–​296. Hunter, F. (2007b). Beyond the Edge of the Empire: C ​ aledonians, Picts and Romans. Rosemarkie: Groam House Museum. Hunter, F. (2007c). ‘Silver for the Barbarians: Interpreting Denarii Hoards in North Britain and Beyond’, in R. Hingley and S. Willis (eds), Roman Finds: Context and Theory. Oxford: Oxbow Books, 214–​224. Hunter, F. (2008). ‘Celtic Art in Roman Britain’, in D. Garrow, C. Gosden, and J. D. Hill (eds), Rethinking Celtic Art. Oxford: Oxbow Books, 129–​145. Hunter, F. (2009). ‘Traprain Law and the Roman World’, in W. S. Hanson (ed.), The Army and Frontiers of Rome. Portsmouth, RI: Journal of Roman Archaeology Supplementary Series 74, 225–​240. Hunter, F. (2010a). ‘Changing Objects in Changing Worlds:  Dragonesque Brooches and Beaded Torcs’, in S. Worrell, G. Egan, J. Naylor, K. Leahy, and M. Lewis (eds), A Decade of Discovery:  Proceedings of the Portable Antiquities Scheme Conference 2007. Oxford: Archaeopress/​BAR British Series 520, 91–​107. Hunter, F. (2010b). ‘Beyond the Frontier: Interpreting Late Roman Iron Age Indigenous and Imported Material Culture’, in R. Collins and L. Allason-​Jones (eds), Finds from the Frontier: Material Culture in the 4th-​5th Centuries. York: Council for British Archaeology, 96–​109. Hunter, F. (2013). ‘The Lives of Roman Objects beyond the Frontier’, in P. S. Wells (ed.), Rome Beyond its Frontiers: Imports, Attitudes and Practices. Portsmouth, RI: Journal of Roman Archaeology Supplementary Series 94, 15–​28. Hunter, F. (2014). ‘Looking over the Wall: The Late and Post-Roman Iron Age North of Hadrian’s Wall,’ in F. K. Haarer (ed.), AD 430: The History and Archaeology of Late and PostRoman Britain. London: Roman Society, 206–215. Hunter, F. (2015a). ‘The Lure of Silver: Denarius Hoards and Relations across the Frontier,’ in D. J. Breeze, R. H. Jones and I. A. Oltean (eds), Understanding Roman Frontiers: A Celebration for Professor Bill Hanson. Edinburgh: Birlinn, 251–269. Hunter, F. (2015b) ‘Interpreting Celtic Art on the Roman Frontier—The Development of a Frontier Culture in Britain?’, in L. Vagalinski and N. Sharankov (eds), Proceedings of the 22nd International Congress of Roman Frontier Studies, Ruse, Bulgaria, September 2012. Sofia: National Archaeological Institute and Museum, 721–727. Hunter, F., and Carruthers, M. (2012a) (eds). ScARF Iron Age Panel Report (accessed 21 January 2013).

200

200   Fraser Hunter Hunter, F., and Carruthers, M. (2012b) (eds). ScARF Roman Scotland Panel Report (accessed 21 January 2013). Hunter, F., and Keppie, L. (2012) (eds). A Roman Frontier Post and its People: Newstead 1911–​ 2011. Edinburgh: National Museums Scotland. Hunter, F., Henig, M., Sauer, E., and Goodner, J. (forthcoming). ‘Mithras in Scotland: A Mithraeum at Inveresk (East Lothian)’, Britannia 47. Ireland, S. (2008). Roman Britain: A Sourcebook. 3rd edn. London: Routledge. James, S. (1999). The Atlantic Celts:  Ancient People or Modern Invention? London:  British Museum. James, S. (2002). ‘Writing the Legions: The Past, Present and Future of Roman Military Studies in Britain’, Archaeological Journal, 159: 1–​58. Jobey, G. (1976). ‘Traprain Law:  A  Summary’, in D. W. Harding (ed.), Hillforts. London: Academic Press, 191–​204. Jobey, G. (1978). ‘Burnswark Hill, Dumfriesshire’, Transactions of the Dumfries and Galloway Natural History and Antiquarian Society, 53 [1977–​8], 57–​104. Jones, R. H. (2011). Roman Camps in Scotland. Edinburgh: Society of Antiquaries of Scotland. Keppie, L. (1990). ‘The Romans in Southern Scotland:  Future Discoveries’, Glasgow Archaeological Journal, 16 [1989–​90], 1–​27. Keppie, L. (2004). The Legacy of Rome: Scotland’s Roman Remains. Edinburgh: John Donald. Keppie, L. (2009a). ‘The Garrison of the Antonine Wall: Endangered Species or Disappearing Asset?’, in A. Morillo, N. Hanel, and E. Martín (eds), Limes XX: XXth International Congress of Roman Frontier Studies. Madrid:  Consejo Superior de Investigaciones Científicas, 1135–​1145. Keppie, L. (2009b). ‘Burnswark Hill:  Native Space and Roman Invaders’, in W. S. Hanson (ed.), The Army and Frontiers of Rome. Portsmouth, RI:  Journal of Roman Archaeology Supplementary Series 74, 241–​252. Kilbride-​Jones, H. E. (1938). ‘Glass Armlets in Britain’, Proceedings of the Society of Antiquaries of Scotland, 72 [1937–​8], 366–​395. Kopytoff, I. (1986). ‘The Cultural Biography of Things:  Commoditization as Process’, in A. Appadurai (ed.), The Social Life of Things:  Commodities in Cultural Perspective. Cambridge: Cambridge University Press, 64–​91. Macinnes, L. (1984). ‘Brochs and the Roman Occupation of Lowland Scotland’, Proceedings of the Society of Antiquaries of Scotland, 114: 235–​249. McIntosh, F. (2011). ‘Regional Brooch-​Types in Roman Britain: Evidence from Northern England’, Archaeologia Aeliana, 5th ser. 40: 155–​182. Maldonado, A. (2015). ‘The Early Medieval Antonine Wall’, Britannia 46: 225–245. Mann, J. (1974). ‘The Northern Frontier after ad 369’, Glasgow Archaeological Journal, 3: 34–​42. Mann, J. C., and Breeze, D. J. (1987). ‘Ptolemy, Tacitus and the Tribes of North Britain’, Proceedings of the Society of Antiquaries of Scotland, 117: 85–​91. Manning, W. H. (2006). ‘The Roman Ironwork Deposits from the Fort at Newstead’, Bayerisches Vorgeschichtsblätter, 71: 15–​32. Mattingly, D. (2006). An Imperial Possession: Britain in the Roman Empire 54 bc–​AD 409. London: Penguin Books. Maxwell, G. S. (1987). ‘Settlement in Southern Pictland: A New Overview’, in A. Small (ed.), The Picts: A New Look at Old Problems. Dundee: Graham Hunter Foundation, 31–​44.

Beyond Hadrian’s Wall   201 Maxwell, G. S. (1989). The Romans in Scotland. Edinburgh: Mercat Press. Maxwell, G. S. (1990). A Battle Lost:  Romans and Caledonians at Mons Graupius. Edinburgh: Edinburgh University Press. Maxwell, G. S. (1998). A Gathering of Eagles: Scenes from Roman Scotland. Edinburgh: Birlinn. Megaw, J. V. S., and Megaw, M. R. (1996). ‘Ancient Celts and Modern Ethnicity’, Antiquity, 70: 175–​181. Megaw, J. V.  S., and Megaw, M. R. (1998). ‘The Mechanism of (Celtic) Dreams? A  Partial Response to our Critics’, Antiquity, 72: 432–​435. Mighall, T. (2003). ‘Geochemical Monitoring of Heavy Metal Pollution and Prehistoric Mining: Evidence from Copa Hill, Cwmystwyth, and Mount Gabriel, County Cork’, in P. Craddock and J. Lang (eds), Mining and Metal Production through the Ages. London: British Museum, 43–​51. Painter, K. (2013). ‘Hacksilber:  A  Means of Exchange?’, in F. Hunter and K. Painter (eds), Late Roman Silver: The Traprain Treasure in Context. Edinburgh: Society of Antiquaries of Scotland, 215–​242. Pitts, L., and St Joseph, J. K. (1985). Inchtuthil: The Roman Legionary Fortress. London: Society for the Promotion of Roman Studies. Poulter, J. (2009). Surveying Roman Military Landscapes across Northern Britain. Oxford: Archaeopress/​BAR British Series 492. Price, J. (1988). ‘Romano-​British Glass Bangles from East Yorkshire’, in J. Price, P. R. Wilson, C. S. Briggs, and S. J. Hardman (eds), Recent Research in Roman Yorkshire. Oxford: Archaeopress/​ BAR British Series 193, 339–​366. Richmond, I. A. (1958) (ed.). Roman and Native in North Britain. Edinburgh: Nelson. Robertson, A. (1970). ‘Roman Finds from Non-​Roman Sites in Scotland’, Britannia, 1: 198–​226. Robertson, A. S., and Keppie, L. (2001). The Antonine Wall:  A  Handbook to the Surviving Remains. Glasgow: Glasgow Archaeological Society. Sims-​Williams, P. (1998). ‘Celtomania and Celtoscepticism’, Cambrian Medieval Celtic Studies, 36: 1–​35. Sommer, C. S. (1984). The Military Vici in Roman Britain. Oxford: BAR British Series 129. Sommer, C. S. (2006). ‘Military Vici in Roman Britain Revisited’, in R. J.  A. Wilson (ed.), Romanitas: Essays in Roman Archaeology in Honour of Sheppard Frere on the Occasion of his Ninetieth Birthday. Oxford: Oxbow Books, 95–​145. Stallibrass, S. (2009). ‘The Way to a Roman Soldier’s Heart: A Post-​Medieval Model for Cattle Droving to the Hadrian’s Wall Area’, in M. Driessen, S. Heeren, J. Hendriks, F. Kemmers, and R. Visser (eds), TRAC 2008:  Proceedings of the 18th Theoretical Roman Archaeology Conference, Amsterdam 2008. Oxford: Oxbow Books, 101–​112. Steer, K. A. (1958). ‘Arthur’s O’on:  A  Lost Shrine of Roman Britain’, Archaeological Journal, 115: 99–​110. Steer, K. A. (1966). ‘A Roman Building at Easter Langlee, Roxburghshire’, Proceedings of the Society of Antiquaries of Scotland, 98 [1964–​6], 320–​1. Stevenson, R. B. K. (1956a). ‘Native Bangles and Roman Glass’, Proceedings of the Society of Antiquaries of Scotland, 88 [1954–​6], 208–​221. Stevenson, R. B. K. (1956b). ‘Pictish Chain, Roman Silver and Bauxite Beads’, Proceedings of the Society of Antiquaries of Scotland, 88 [1954–​6], 228–​230. Stevenson, R. B. K. (1966). ‘Metal-​Work and Some Other Objects in Scotland and their Cultural Affinities’, in A. L. F. Rivet (ed.), The Iron Age in Northern Britain. Edinburgh: Edinburgh University Press, 17–​44.

202

202   Fraser Hunter Stevenson, R. B. K. (1976). ‘Romano-​British Glass Bangles’, Glasgow Archaeological Journal, 4: 45–​54. Swan, V. G. (1988). ‘Comments on Inveresk Ware’ [pp. 167–​171], in G. Thomas ‘Excavations at the Roman Civil Settlement at Inveresk, 1976–​77’, Proceedings of the Society of Antiquaries of Scotland, 118: 139–​76. Swan, V. G. (1999). ‘The Twentieth Legion and the History of the Antonine Wall Reconsidered’, Proceedings of the Society of Antiquaries of Scotland, 129: 399–​480. Symonds, M. F. A., and Mason, D. J. P. (2009). Frontiers of Knowledge: A Research Framework for Hadrian’s Wall. Durham: Durham County Council/​Durham University. Tipping, R. (1997). ‘Pollen Analysis and the Impact of Rome on Native Agriculture around Hadrian’s Wall’, in A. Gwilt and C. Haselgrove (eds), Reconstructing Iron Age Societies. Oxford: Oxbow Books, 239–​47. Tipping, R., and Tisdall, E. (2005). ‘The Landscape Context of the Antonine Wall: A Review of the Literature’, Proceedings of the Society of Antiquaries of Scotland, 135: 443–​469. Wallace, C. (2006). ‘Long-​lived Samian?’, Britannia, 37: 259–​272. Wilson, A. (2010). Roman and Native in the Central Scottish Borders. Oxford: Archaeopress/​ BAR British Series 519. Woolliscroft, D. J. (2000). ‘More Thoughts on why the Romans Failed to Conquer Scotland’, Scottish Archaeological Journal, 22/​2: 111–​122. Woolliscroft, D. J., and Hoffmann, B. (2006). Rome’s First Frontier. Stroud: Tempus. Youngs, S. (2013). ‘From Chains to Brooches:  The Uses and Hoarding of Silver in North Britain in the Early Historic period’, in F. Hunter and K. Painter (eds), Late Roman Silver: The Traprain Treasure in Context. Edinburgh: Society of Antiquaries of Scotland, 403–​425.

Chapter 10

Mobilit y, M i g rat i on, and Diasp oras i n Roman Bri ta i n Hella Eckardt and Gundula Müldner

Introduction Did the Roman period see unprecedented levels of mobility? In order to answer this question, two main issues need to be addressed. The first is concerned with the various forms of evidence for mobility, notably written epigraphy, material culture, and scientific techniques such as isotopic analysis, and their respective strengths and weaknesses. The second is about how we can model interactions and relationships between incomers and locals, and indeed whether such distinctions mattered much in a cultural context that appears to have been characterized by an emphasis on Romanitas rather than geographical origin. This chapter builds on the results of a multidisciplinary AHRC-​funded project on mobility in Roman Britain held at the University of Reading from 2007 to 2009 and, in particular, on a paper contextualizing its results published in 2010 (Eckardt 2010). There is little point in repeating here the detailed arguments made in 2010. Rather, key issues will be summarized and only the evidence that has emerged since that publication will be examined in detail. This is especially relevant in the fast-​moving field of isotope analysis, where new data continue to challenge and help develop our original interpretations. However, there is also new research on material culture and epigraphy, and it is worth re-​examining very briefly how thinking about diasporas and hybridization may move us beyond established critiques of the ‘Romanization’ paradigm. There is also an increased awareness of the political context in which research on emotive themes such as migration is carried out, and this will be addressed by reviewing some recently developed educational resources.

204

204    Hella Eckardt and Gundula Müldner

Archaeological Evidence for Mobility Conquest by Rome brought a wide range of people to Britain, through both voluntary and forced migration; the military, imperial administration, and trade are usually thought to be the most important drivers, but the role of slavery also has to be considered (Eckardt 2010: 102–10​3; Webster 2010). Archaeologists have long attempted to identify migrants and to estimate the proportion they make up within a given site or region. Traditionally, this has been attempted through the study of written sources, epigraphy, and material culture, but, more recently, isotope analysis has offered new opportunities for research.

Epigraphy Epigraphy identifies foreigners through individual names, stated places of origin, and reference to regionally specific cults or calendars (e.g. Birley 1979; Noy 2000, 2010). These data produce personal, individual histories and allow for the exploration of wider patterns in terms of age, gender, and status. However, the evidence is unevenly distributed, with relatively few inscriptions known from Britain in comparison with other provinces, and inherently biased, for example, towards urban and military areas (Rowland 1976; Jones and Mattingly 2002: map 5.10; see also Pearce 2010). Another obvious proviso is that it is only those individuals stating their origins who are visible in the source material. Depending on the region and time, inscriptions may reflect only a few per cent of the total population, making estimates of the percentage of foreigners within epigraphic populations fraught with difficulty (Eckardt 2010, 2012). Nevertheless, epigraphy provides important insights and detailed information on those few individuals, and it is therefore worth adding a number of important studies that were missed, and new work that has been published since 2010. Thus Kakoschke (2002) provides an extremely useful parallel to the Gaulish data by analysing the evidence for incomers to the two Germanies (cf. Wierschowski 1995, 2001). Another assessment of immigration comes from Roman Dacia, which apparently saw intensive colonization after the Trajanic wars, with natives only very rarely attested (Mihailescu-​ Bîrliba 2011). Carroll (2006: 209–​232) provides an interesting comparison of migration into military (Mainz and Chester) and civilian (Lyon and Cologne) sites, highlighting the impact of the army, change through time, and again the diversity of people moving across the empire. Clay (2007) explores the epigraphic and linguistic evidence for Germanic people on Hadrian’s Wall, discussing in particular how religious practices may serve both to maintain relations with the homeland and form new ones with the inhabitants of the local area. Another recent case study highlights the self-​representation of large numbers of Milesians in Hellenistic and Roman Athens, and places them

Mobility, Migration, and Diasporas in Roman Britain    205 into the context of social relations and especially intermarriage between citizens and non-​citizens (Gray 2011). It is worth exploring one other new study of epigraphy and mobility in more detail, as it provides a fascinating chronological contrast by focusing on the late antique period (fourth–​eighth centuries ad). Handley recorded 623 examples of travellers or foreigners on a total of 567 inscriptions, revealing ‘an intricate web of travel in all directions: Spaniards in Italy, Pannonians in Dalmatia, Gauls in Africa’ (Handley 2011: 17). The study uses only individuals who state their origin or use regionally specific calendars, and demonstrates the unreliability of using names alone (Handley 2011: 21–​28). The advantage of the epigraphic material is that it allows detailed analysis of travellers in terms of gender, age, status, and profession, although numbers are low and the overall dataset is biased (Handley 2011: 37–​51). For example, male migrants dominate with about 80 per cent of the sample (while only making up 63 per cent of the whole epigraphic population), but some women also travel long distances, and factors like intermarriage might decrease the likelihood of their foreign origin being recorded. The age recorded in an inscription is, of course, not necessarily the age at which an individual first travelled, but the evidence from the late antique west shows more older male migrants leaving inscriptions, while in the city of Rome younger males are more strongly represented in the epigraphic data (Figure 10.1). Recorded professions range from soldiers and merchants to government office-​holders and bishops; there is also a lawyer and a stonemason. The reasons for travel are diverse, including war/​soldiery, pilgrimage, ecclesiastical business, exile, trade, and the emigration of individuals and families; there is also some evidence that the dead were moved, sometimes over considerable time periods and distances (Handley 2011: 53–​62). The study clearly illustrates that travel and mobility continued into the late antique period, and that travel had considerable economic implications. Ages of male travellers

25

Percentage

20 15 10 5 0

0–9

10–19

20–29

30–39

40–49

50–59

60–69

70–79

80–89

90–99

Males to Rome (Noy): N = 47 Males late antique west: N = 107

Figure 10.1  Epigraphically recorded age of travellers to Rome 300–​600 ad as recorded by Noy (2000) and for the late antique west by Handley (2011). Source: after Handley (2011: graphs 1 and 2). © Department of Archaeology, University of Reading.

206

206    Hella Eckardt and Gundula Müldner Handley (2011: 114) uses ancient sources to highlight the numbers of people recorded on ships and the potential financial value of such travellers. Thus an early sixth-​century letter of Cassiodorus records that 75 per cent of the ship’s profit came from the passenger fees, and only 25 per cent from the cargo, reversing the usual assumption about travel ‘piggy-​backing’ on the shipment of bulk goods (Handley 2011: 11). In summary, the epigraphic evidence still has much to offer, providing rich detail and a direct connection to people’s life stories; there is also much to learn about the relative proportions of migrants of different origin, gender, age, and status. However, it is still impossible truly to estimate numbers of foreigners from this limited and biased evidence.

Material Culture and Identity In practice, particular objects associated with personal adornment and food ways have long been used to identify ‘foreigners’. Thus certain beads, belt fittings, or brooches, when found outside their ‘normal’ distribution, may indicate the presence of migrants who carried small portable items with them, although the effects of trade and changing fashions always have to be considered. Pots, even when made in local fabrics, may indicate the movement of their makers and consumers, and this may be especially true for vessels associated with particular ways of preparing and consuming food (Swan 1992, 1999, 2009; Cool 2006a: 172–​183; 2006b). There are significant provisos and problems: for example, when attempting to establish the typological closeness of the forms found in Britain to those found in the supposed regions of origin (Fulford 2010). Burial evidence is of particular interest, as the rite may reveal ‘intrusive’ practices, although the picture is much more complicated than early archaeologists thought (cf. Millett 1999: 194–19​6; Cool 2010; Pearce 2010). More fundamentally, the very link between material culture and ethnicity, and indeed the meaning of ethnicity as a concept, has seen a significant review in theoretical archaeology. Rather than ethnicity being viewed as fixed and given, it is now usually seen as fluid, and there is an increased recognition that objects do not express just one aspect of identity, but that their use and interpretation are context-​dependent and open to multiple interpretations (e.g. Jones 1997; Gardner 2007: 197–​203). Thus an object may relate to age, gender, marital status, or religious beliefs, as well as act as an indicator of ethnicity (e.g. Meskell 2001), and, in some cases, an object type may appear intrusive or unusual to a modern artefact specialist but other aspects of the wearer’s dress may have been much more important signals of identity in the past. This is not the place to discuss these issues in detail, but merely to note that, in contrast to the post-​Roman period (e.g. Brather 2000, 2004; Curta 2007; Härke 2007), the question of ethnicity has not, until relatively recently, been a major point of discussion in Roman archaeology. This seems paradoxical, given the evidence for high levels of mobility in the Roman period and the fine-​grained and well-​published nature of the material culture, but this is now

Mobility, Migration, and Diasporas in Roman Britain    207 beginning to be addressed (e.g. Pohl 1998). Work on the archaeology of the Rhine delta in particular explored the nuanced ways in which Rome as a colonial power impacted on the creation of the Batavians as an ethnic and political entity (Roymans 2004; Derks and Roymans 2009). In other recent work, Swift (2010) examines a series of bracelets that have restricted distributions in late Roman Britain, suggesting production in regional workshops, but that also occur in limited numbers on continental sites. She argues that such finds are indicative of migrants, who brought relatively low-​value items of personal adornment with them and that this is the case in particular where the type-​site (or social) distribution of an object type is restricted to military and large urban sites. Using a combination of artefact types and burial context in the well-​published, large cemetery at Krefeld-​Gellep in north-​west Germany, Swift suggests the presence of migrants from Roman Britain and the Upper Danube provinces. Ivleva (2010) compares 31 epigraphic records (including military diplomas) and 241 British-​made brooches to trace the presence of British emigrants (and people of diverse origins who may have lived in Britain for some time) on the continent. The argued connection between British auxiliary soldiers, other British recruits, and returning veterans and brooch distributions is not always convincing, and it is impossible to know whether these brooches, thought to have been made in Britain, were worn as an expression of ‘British’ identity or simply because they were available and convenient personal possessions. Nevertheless, the paper does illustrate an interesting artefactual distribution and, at the very least, highlights close connections between Britain and the Rhineland in particular (Ivleva 2010: figure 2). What all these case studies share is an attempt to recognize ‘intrusive’ material culture in the archaeological record but to interpret these objects not simply as overt signals of ethnic identity but in their historical context and in relation to other aspects of identity such as gender or, for example, military identities.

The Application of Archaeological Science to the Question of Mobility Isotopes and Mobility: The Background The 2007–​9 Diasporas in Roman Britain project was the first to employ combined strontium (87Sr/​86Sr) and oxygen (δ18O) isotope analysis on a large scale in order to distinguish between locals and foreigners in a Roman province. ‘Isotopic signatures’ are incorporated into the body tissues of humans and animals through the food and drink they ingested at the time of tissue formation (see Eckardt 2010: 112–​113 for a summary). The strontium and oxygen isotope composition of tooth enamel, which is fixed in the early years of life, allows us to characterize broadly an individual’s place of childhood residence in terms of geology and climate. The ‘isotopic signature’ thus obtained can

208

208    Hella Eckardt and Gundula Müldner then be compared to available background data for different geographical regions, first in order to assess whether an individual’s isotope values are consistent with a local upbringing (that is, at the place of burial). If they are, the most parsimonious explanation is that this individual was indeed local, although it must always be borne in mind that many other places may share a similar geology and climate. For example, typical oxygen isotope values for Britain are shared by parts of western and southern Europe and even the Mediterranean coast, while much of the presently available strontium isotope data from Britain and mainland Europe appear to reflect the isotopic composition of a relatively small number of soils that are particularly suitable for agricultural production and are, therefore, sought out by humans everywhere (Montgomery 2010; Chenery et al. 2011; Evans et al. 2012; Whittle and Bickle 2013). Where the isotope values exclude a local origin, the information contained in the isotopic signal can be used to narrow down potential places of childhood residence. It should be emphasized, however, that it is usually impossible to assign an individual’s origins to a particular locale based on isotope evidence alone, simply because too many areas share the same strontium and oxygen isotope characteristics. Isotope analysis is most valuable as an ‘exclusive technique’ (Montgomery 2010: 336)—​that is, it can be used to eliminate local origin and other suggested homelands because they are inconsistent with the data, but, in order to make a case for one of the (usually) many possible places of childhood residence that are compatible with the isotope data, the archaeologist has to rely on other contextual evidence (see, e.g., Fitzpatrick (2011) for a carefully constructed argument for an origin of the ‘Amesbury Archer’ in the Alpine region, based on the combined analysis of isotopic and archaeological data). Isotope analysis is a destructive technique, requiring a small sample often taken from the second or third permanent molar, which represents an enamel development period between c. 3 and 13 years of age, and this can only partly be compensated for by photographs and casts of the teeth made prior to sampling. For oxygen and—​especially—​ strontium isotope analysis, tooth enamel is the preferred analyte, not only because it fixes geographical information from a well-​definable period in childhood, but also because, unlike bone or tooth dentine, it has been shown to be largely resilient against post-​mortem contamination by elements in the burial soil (e.g. Hoppe et al. 2003). A detailed description of the methods used by the project can be found in Chenery et al. (2010). For the Roman world, isotope mobility studies conducted prior to the Diasporas in Roman Britain project had focused on late Roman burials in southern Germany (Schweissing and Grupe 2003) and on Portus near the city of Rome (Prowse et al. 2007), although a number of studies have been added since (Perry et al. 2008, 2009, 2011; Killgrove 2010b; Prowse et al. 2010) and there will likely be more by the time this chapter is published. In a specifically Romano-​British context, a small study had previously examined individuals from Mangotsfield near Bristol and from Winchester (Budd et al. 2004: 134), while a paper by Evans et al. (2006) had analysed eighteen individuals from Lankhills, where Clarke (1979: 377–​389) identified a group of incomers on the basis of unusual objects and burial rite (cf. Swift 2010: 237). Evans et al. (2006) compared the isotopic

Mobility, Migration, and Diasporas in Roman Britain    209 signatures of nine individuals identified as ‘local’ and nine identified as ‘exotic’ on the basis of these archaeological criteria. The results clearly showed that the starting assumptions about origin were oversimplified, and that the relationship between social identity, as expressed in the burial rite, and geographical origin are much more complex. These conclusions were very much borne out by our larger project, which analysed 155 skeletons from five sites in Roman Britain (York, Catterick, Gloucester, Poundbury/​Dorchester, and Lankhills/​Winchester (Eckardt et al. 2009; Leach et al. 2009, 2010; Chenery et al. 2010, 2011; Müldner et al. 2011)). The key result is that all sites showed considerable diversity, with a range of values that may be compatible with local (within a 30 kilometre radius of the site) or other British origins, but also several individuals who came from either cooler and/​or more continental, or warmer, possibly more coastal, areas. Some attempt was made in the 2010 paper to evaluate levels of mobility, comparing the results of various epigraphic studies, which all suggest a proportion of c. 5 per cent stating foreign origin, with a major paper by Scheidel (2004), who argued that 40 per cent of Italian Roman citizens over the age of 45 would not have lived in the place where they were born (cf. Eckardt 2010: 103–​107). The isotope analyses of Prowse et al. (2007) and Schweissing and Grupe (2003) suggest high levels (approximately 30 per cent) of migration, and our results also indicate a high proportion of incomers. While roughly half the sampled individuals could be ‘local’ in the sense of consuming foods sourced within a 30 kilometre radius of the relevant site, and a significant proportion of the non-​locals are likely to have come from elsewhere in Britain, the project also identified some individuals who are likely to have spent their childhood outside Britain (Eckardt 2010: table 7.2). The small town and military fort of Catterick was the least diverse site, but nearby York (especially the beheaded males from the unusual cemetery at Driffield Terrace), as well as Lankhills (Winchester), showed larger numbers of suggested migrants (Eckardt et al. 2009; Leach et al. 2009, 2010; Chenery et al. 2010, 2011; Müldner et al. 2011). However, as pointed out in the original paper, the proportions given in table 7.2 (Eckardt 2010) have to be taken with a large pinch of salt. First, they are not based on a representative sample, since our selection of study sites was biased towards archaeologically visible, highly ‘Romanized’ urban populations and, within those, sample selection was often modified to include ‘unusual’ individuals, whose origin the project aimed to explore. Additional isotope data, especially from rural sites and from other parts of the Roman Empire, are therefore required to confirm the wider relevance of the results. Below we will review more recent work in what is a rapidly evolving area of research, concerning not just methodological questions but also issues that are perhaps specific to the Roman world.

Recent Isotope Research: Challenges and Issues Isotope analysis is an attractive tool to archaeologists and one that has had considerable exposure in the media, but, because of this, there is clearly a risk that complex data are oversimplified to the extent of creating the impression that strontium and oxygen isotope analysis can provide ‘an unequivocal postcode for the domicile of that individual as

210

210    Hella Eckardt and Gundula Müldner a child’ (Pollard 2011: 634). Although these dangers have already been pointed out above and in our original papers, it is worth expanding on some of the issues here, in the light of newly published data. In this respect, it is important to emphasize that isotope data are just like other archaeological evidence in that they are subject to reinterpretation as more information becomes available. For both oxygen and strontium isotope studies the question of how to define the ‘local range’—​that is, the range of values that can be considered as consistent with a local upbringing—​is a central one, and, while this range can be estimated using a number of different indicators, it is important to realize that it is essentially the result of culturally mediated behaviour (that is, food and drink acquisition strategies) and that changes in food preferences and agricultural production, the movement of foods, as well as the transport or differential treatment of drinking water, could cause fluctuations in the local range of a given site over time (e.g. Bruun 2010; Killgrove 2010a; Montgomery 2010; Chenery et al. 2011; Brettell et al. 2012). In this context, the issue of conversion equations, which are used to relate human oxygen isotope ratios to drinking water values (and thereby to certain geographical areas through modern rainfall maps), should also be mentioned. While the possibility of climate change is raised most often against these, this problem may be smaller than often assumed, at least for European populations from the later Holocene (see Chenery et  al. 2010:  153). However, recent discussions have also highlighted the problem that different laboratories prefer different equations (computing different drinking water values from the same skeletal δ18O) and that the often considerable statistical error inherent in these calculations is frequently understated (Daux et al. 2008; Chenery et al. 2010; Pollard et al. 2011b). Interpretations based on this approach may, therefore, not be as accurate as they might appear to a non-​specialist reader. For this reason, direct comparisons with human oxygen isotope data, whenever these are available, are to be preferred. One very significant advance since the start of the Diasporas in Roman Britain project has been the large increase in the number of available human isotope data from Britain, which can be used for making more robust local range estimates than was previously possible. In 2009, we had only 57 humans to compute a likely range of enamel oxygen isotope values (δ18Op) of individuals who spent their childhood in Britain (see Leach et al. 2009; Chenery et al. 2010); subsequent publications—​especially the synthetic paper by Evans et al. (2012)—​have seen the number of data points increase to more than 600. The revised range that can be calculated from these data is somewhat wider than our earlier estimate (two standard deviations range of 16.3 to 19.1‰ compared to the previous estimation of 16.8–​18.6‰; see Evans et al. (2012: figure 4). The implications of this are illustrated in Figure 10.2, which displays isotope data from the Lankhills cemetery with both the earlier and the revised estimate for a British skeletal oxygen isotope range. A number of individuals whose δ18Op had previously seemed too high to be easily consistent with origins in Britain and for whom a childhood in warmer, possibly more coastal, areas had to be considered (see

Mobility, Migration, and Diasporas in Roman Britain    211 Eckardt et al. 2009) are now just inside the British range estimate, and local origins for these now appear more plausible. It should be noted here that there is the danger of a circular argument, since almost one quarter of the data set contained in Evans et al.’s (2012) paper and from which the new estimate is computed are Diaspora project results. The new (2012) estimate is also probably somewhat inflated, as the synthetic data set is likely to include a number of non-​British migrants, even if some of the more obvious outliers have been removed (see Evans et al. 2012). Nevertheless, the suggestion that the skeletal oxygen isotope range for Britain is wider than had been assumed previously appears to be also borne out by other new research (Parker Pearson et al., forthcoming). Even if the new estimate is likely to be revised again in the future, this exercise therefore demonstrates the fluidity (within reason) of local range estimates in a rapidly expanding field of research, and we can safely say that it will always be problematic, and to an extent nonsensical, to draw lines meant to separate locals and non-​locals through a continuous field of data points. This is especially true for oxygen isotope values, which, unlike strontium, are subject to fractionation and change through a large number of biological processes (but see Bentley (2006) and Montgomery (2010) for some of the complexities of strontium isotope data). Several recent publications have, for example, drawn attention to how the oxygen isotope composition of drinking water can be modified by various cultural practices (for example, prolonged cooking, brewing), causing a shift from the isotope ratio of the local rainwater (which is usually assumed as baseline for local values in a given region), to higher values (e.g. Daux et al. 2008; Knudson 2009; Price et al. 2010; Brettell et al. 2012). Regular consumption of such culturally modified drinking water could make individuals appear ‘too warm’ for a given locale. For the Roman period, the regular consumption of wine, even in smaller quantities, might well have had such an effect, since the isotopic shift between the (rain-​)water feeding the vine and the finished product can be very considerable (see West et al. 2007). However, it should be noted that there are next to no mechanisms that would cause a shift to lower values—​only transport of drinking water from ‘cooler’ areas/​higher altitudes would achieve this. Despite these complexities, and although isotope analysis is therefore not the panacea for archaeological migration studies that it is sometimes portrayed to be, the power of isotopic methods is beyond doubt and should not be understated. They are particularly effective when they can identify individuals who plot well outside the local range, ideally as outliers from a main distribution of samples (Figure 10.2). In this way, isotope data can provide an independent, ‘biological’ indication of mobility, which can be gainfully contrasted with other, more culturally mediated types of evidence relating to the individual (such as burial rite or grave goods) in order better to understand the complex ways by which identity can be expressed in material remains—​and the Diasporas in Roman Britain project has produced a number of such examples. Recent investigations have also highlighted the great advantage of multi-​isotopic investigations in order to increase the resolution of the isotopic signal. Within the

212

212    Hella Eckardt and Gundula Müldner ‘Intrusive’ burials

0.7125

Other burials

0.7115

87Sr/86Sr

0.7105 0.7095 0.7085

Local Sr isotope range

0.7075 0.7065 0.7055 14.0

British O isotope range (2009 and 2012 estimate)

Analytical error (2σ) 15.0

16.0

17.0

18.0

19.0

20.0

δ18Op(‰)VSMOW

Figure 10.2  Strontium and oxygen isotope data of humans from Lankhills/​Winchester, contrasting burials identified as ‘intrusive’ according to Clarke’s criteria (1979) with other burials from the site; the two boxes indicate the local strontium isotope range for Winchester and estimates for the range of oxygen isotope values consistent with a childhood in Britain based on human skeletal phosphate (δ18Op) data available in 2009 (thin line) and 2012 (thick line). Source: after Evans et al. (2006) and Eckardt et al. (2009). © Department of Archaeology, University of Reading.

Diasporas in Roman Britain project, we had particular success with integrating palaeodietary (carbon and nitrogen isotope) data with the oxygen and strontium isotope values. Even though carbon and nitrogen isotopes are not usually thought of as useful for migration studies, especially in a European context, they reliably identified a number of long-​distance migrants, probably because of the considerable environmental and dietary differences that existed within the Roman Empire and its adjoining regions (Müldner 2013). Carbon and nitrogen isotope evidence can, therefore, be used further to constrain possible areas of origin. In some cases, and especially where individuals showed evidence of consumption of C4-​plants (millet)-​based protein, which was not available in Roman Britain in any quantity, carbon and nitrogen isotopes can even identify non-​locals when the strontium and oxygen isotope data are ambiguous, as was the case for one of the ‘Headless Romans’ of York and for a much less conspicuous male burial from Gravesend, Kent (Müldner et al. 2011; Pollard et al. 2011a). Lead isotopes may also add another dimension to the analysis. Montgomery et al. (2010) demonstrate that during the Roman period—​when the geological information normally contained in lead isotopes is lost owing to high levels of anthropogenic

Mobility, Migration, and Diasporas in Roman Britain    213 lead pollution and individuals from Britain all exhibit similar values controlled by an amalgamate of British ore lead—​lead isotopes can, nevertheless, be used to distinguish migrants from areas where lead from very different sources was used. For example, they allowed for the high-​status female burial from Spitalfields, London, to be identified as an incomer, despite her very generic strontium isotope values, as her lead signature is clearly not British. Human lead isotope data from Rome itself, which have recently become available and are very similar to those of the ‘Spitalfields Lady’, give at least one possible area of origin (Montgomery et al. 2010: 217–​219). Conversely, the lead isotope values for both the ‘Ivory Bangle Lady’ of York and the Catterick ‘Eunuch’ are compatible with British ore lead (Eckardt and Müldner, unpublished data), and, although it is as yet unclear which regions of the Roman Empire shared similar values (an origin in Rome at least can apparently be excluded), these two individuals may indeed have been of British origin, especially since none of the other isotopic systems provided unequivocal evidence to the contrary (Leach et al. 2010; Chenery et al. 2011). It is only by playing a host of analyses off against each other that such patterns emerge, and it is probably that future investigations will help further to refine interpretations.

Models for Interaction The discussion of identities in Roman Britain and the Roman world has been dominated by the ‘Romanization’ paradigm (Millett 1990; cf. Webster 2001: 209–​217), which stressed the unifying aspects of the Roman Empire and focused on the ways in which certain cultural traits were adopted and adapted by conquered populations. More recent models stress the varied responses to, and different experiences of, empire (Woolf 1998; Webster 2001: 217–​223; Mattingly 2004, 2006). Given the high levels of mobility within a complex colonial system, it has been suggested that diaspora theory may be a useful model that allows us to focus on migrant interactions with host communities and to deal with the forced and voluntary movement of both groups and individuals (Eckardt 2010: 107–​109, with detailed references). Lilley (2004) and Cohen (2008) provide useful summaries of the characteristics of diasporic communities, which are defined not just by initial (forced or voluntary) dispersal and a distinction from the host society, but also by a continuous social or spiritual link to the homeland. Most of the initial work on diasporas has focused on so-​called victim diasporas, in particular the Jewish and Afro-​Caribbean experience, but more recently scholars have begun to explore trade and colonial diasporas, which both have potential relevance to the Roman world. Trade diasporas have been described as a ‘nation of socially interdependent, but spatially dispersed, communities’ whose members are distinct from both the societies in which they originated and those in which they live (Cohen 2008: 83). Archaeological case studies on trade

214

214    Hella Eckardt and Gundula Müldner diasporas have largely focused on early modern, historical communities (e.g. Voss 2005), but the Roman world is full of examples of merchants setting up monuments far from home, and it seems likely that in many cases their origin—​as in the case of the famous Barates of Palmyra who set up the gravestone for his British wife in Southshields—​continued to be important to them (Noy 2010: 20–​21). Also of considerable relevance to the Roman period is the notion of colonial and imperial diasporas. Cohen defines the imperial diaspora as ‘marked by a continuing connection with the homeland, a deference to and imitation of its social and political institutions and a sense of forming part of a grand imperial design’ (Cohen 2008: 69; cf. Lilley 2004). Both indigenous groups and European/​British colonizers in Australia, the United States, and India can be viewed as colonial and imperial diaspora communities (e.g. Lawrence 2003; Casella 2005; Lilley 2006). Again, the applications to the Roman world are obvious. Finally, the theoretical and methodological development of diaspora studies in some ways mirrors how ‘Roman-​ness’ has been studied in the provinces. For example, the interactions between diasporic and host communities are clearly complex, and both are potentially characterized by considerable diversity, and this has echoes in our changed understanding of ‘Romans’ and ‘natives’, which has evolved from sharply defined binary opposites to more nuanced appreciations of diversity (e.g. Mattingly 2004, 2006). Even more importantly, earlier studies have tended to focus on the identification of cultural traits as a means to trace origin and to demonstrate cultural continuity and continued connections with the homeland (Eckardt 2010: 107–​108). However, such approaches make assumptions about the static nature of culture, oversimplify the processes of cultural interaction, and are devoid of cultural context. Diaspora communities are different not just from the host population but also from that of their original homeland. Thus Cool (2004, 2010) suggests that some of the practices observed at Brougham may reflect not simply intrusive practices potentially reflecting a trans-​ Danubian homeland, but also traits ‘picked up’ from other areas such as the Rhineland, which these individuals may have passed through on their way to Britain or which were part of a military culture they encountered elsewhere. Humans continue to create their own ‘correct’ ways of being, and their perceptions and practices will be affected by relocations and by their continued interactions with various groups and with each other. More recent work is thus increasingly concerned with the nature of culture change and how, through processes of creolization and hybridization, diasporic identities are created and expressed (cf. Meskell 2001: 286; Wilkie 2004: 113–​114). The concept of creolization has been extensively explored in Roman archaeology (Webster 2001), while that of hybridity is perhaps less well known. The latter is often associated with colonial and postcolonial theory and the creation and maintenance of identities in contexts of colonial oppression, but in its wider sense of ‘entanglement’ conveys the ‘creation of something new which is more than just an addition of its origins’ (Stockhammer 2012: 47; Van Dommelen and Terrenato 2007: 8–​9).

Mobility, Migration, and Diasporas in Roman Britain    215

Immigrants, Race, and the Politics of Outreach The identification of immigrants is an emotive topic in contemporary discourse, and a careful engagement with the wider public is an important aspect of any research project in this area. In addition to using isotope analysis, the Diasporas in Roman Britain project also applied craniomorphometric analysis (forensic ancestry assessment) to skeletons from Roman York (Leach et al. 2009; see also Eckardt 2010: 111–​112). This led to the identification of ‘mixed-​race’ individuals, most notably the so-​called ‘Ivory Bangle Lady’, a rich female burial with unusual grave goods from Sycamore Terrace (Leach et al. 2010). The identification of ‘race’ is a highly controversial and sensitive subject in the humanities, and indeed ‘race’ is recognized as a social construct by the biological sciences; biological anthropologists are, therefore, using racial categories mainly as convenient shorthand for certain phenotypic attributes that vary in frequency in different populations around the globe. Despite this, for some any attempt at measuring skulls is still strongly associated with historic attempts to demonstrate that variability in outward appearance correlates with characteristics such as intelligence or moral character, an approach that was not only racist but also scientifically deeply flawed (e.g. Arnold 2006; Gosden 2006). When conducting these analyses within a Roman context, an added complication is the question of how ethnicity and ‘race’ were viewed in antiquity and whether racism towards, for example, Africans existed (Sherwin-​White 1967; Snowden 1983; Isaac 2004, 2006; McCoskey 2012). Our project employed forensic ancestry assessment in an explicitly interdisciplinary approach to examine the interplay of biological differences and social constructs.

Figure 10.3  Reconstruction of the so-​called ‘Ivory Bangle Lady’ from York. Source: © Aaron Watson.

216

216    Hella Eckardt and Gundula Müldner However, the more subtle points of an academic paper can get lost when reconstructions such as that of the ‘Ivory Bangle Lady’ (Figure 10.3) clearly capture the public imagination and form part of a modern discourse about ‘black’ identity, in many ways crystallizing general debates about immigration. This was evident not only in the press coverage of the ‘Ivory Bangle Lady’, which was dominated by ideas of exoticness and of identifying ‘the first African’, but in the subsequent, often vitriolic, responses by readers of certain newspapers (e.g. ). There is increasing awareness of the impact archaeological interpretations of the past have in the present, as evident in recent calls for a ‘more multicultural’ approach to British cultural heritage, where explicit links between the past and present are made (Benjamin 2003, 2004). While some argue that reconstructions of the past should not be influenced by contemporary politics, others stress the importance of postcolonial and politically aware discourses. Thus Hingley (2010: 232–​236) has reviewed how contemporary anxieties of the nineteenth-​century British Empire influenced interpretations of ‘foreign’ troops on Hadrian’s Wall and, conversely, how in recent years the ‘diverse geographical origins of the soldiers, and the idea that they may have retained elements of their original ethnic character, have been used to create an inclusive role for Hadrian’s Wall’ (Hingley 2010: 238; cf. Hingley 2012). The multinational character of the Wall was highlighted in a 2006 art project Writing on the Wall and an exhibition The Archaeology of ‘Race’: Exploring the Northern Frontier in Roman Britain (Hingley 2010: 238–​240; Tolia-​Kelly 2010). This exhibition, which was shown in Segedunum Museum and Tullie House Museum in 2009 and is still on display at Durham University, aims to show the impact that the diverse origins and mobility of the individuals living on Hadrian’s Wall may have had on daily life. It also explores the complex role and representation of the ‘African’ emperor Septimius Severus: on the one hand, ‘celebrating black presence’ and, on the other, interpreting him as a ‘violent imperialist’ (Tolia-​Kelly 2010: 85 ). Bearing in mind these issues, the Diasporas in Roman Britain project has now developed a website for primary-​school children, working with the Runnymede Trust, a race equality and educational charity. The website is designed to promote the project findings of diversity to Key Stage 2 learners, their teachers and families, and to break down the astoundingly strong public perception of ‘the Romans’ as uniformly ‘Italian’. Throughout, the emphasis is on questioning ‘how do we know’, trying to teach children about archaeological techniques but also that knowledge and interpretations can and do change.

Conclusion The Roman Empire was characterized by high levels of mobility, and diaspora theory may be a helpful way of thinking about the interactions of migrants and locals. The

Mobility, Migration, and Diasporas in Roman Britain    217 concept of trade and imperial diasporas in particular has clear applications to the Roman world. Attempting to assess numbers of incomers in a given site or region is fraught with difficulty: the epigraphic data are heavily biased, and material culture is clearly used in complicated, multi-​layered ways rather than as a simple indicator of ethnicity or origin. The combination of archaeological evidence with both isotopic and craniometric data offers a new avenue for research, but, as with all the other approaches, there are serious methodological and theoretical issues to consider. While some may latch on to a headline figure of 20 or 30 per cent immigration, the sample studied by the Diaspora Project is not representative of the Romano-​British population as a whole, and the interpretation of the isotopic values may well change as more data become available and the impact of various cultural practices on human isotope signals in different time periods are better understood. To this end, there is also an urgent need for further study, in particular of rural populations, which are expected to have been more stationary and can therefore give a better estimate of typical local ranges than the highly diverse urban populations on which research has tended to focus. It is also clear that a multi-​proxy approach involving a combination of scientific techniques has the greatest potential, and future studies should include multiple isotopes for the best possible results. Ultimately, it is in any case not the identification of biological origin that is truly of interest, but its interplay with social identity. As today, people make choices about whether to cling to an old, national, or ethnic identity or to adopt a new one—​a decision that may be described as the ‘sport team choice’ (Handley 2011: preface). In many societies, including quite possibly the Roman world, origin, in the sense of area of birth, may not matter that much (Swift 2010: 268; Pollard 2011: 635) and we must avoid naive attempts to create a multicultural past because we live (or want to live) in a multicultural present. Nevertheless, it is important for archaeologists to be aware of the political context of our work, and, in our attempts to communicate complicated scientific findings to the wider public, and especially children, we have to make a deliberate effort to convey the research process and its uncertainties. Hopefully, by combining a variety of scientific approaches with theoretical awareness, we can begin to understand the historically specific ways in which ‘otherness’ may have been perceived and expressed in Roman Britain.

References Arnold, B. (2006). ‘“Arierdämmerung”:  Race and Archaeology in Nazi Germany’, World Archaeology, 38/​1: 8–​31. Benjamin, R. P. (2003). ‘Black and Asian Representation in UK Archaeology’, The Archaeologist, Institute of Field Archaeologists Newsletter, 48 (Spring), 7–​8. Benjamin, R. P. (2004). ‘Building a Black British Identity through Archaeology’, Archaeological Review from Cambridge, 19/​2: 73–​83. Bentley, R. A. (2006). ‘Strontium Isotopes: From the Earth to the Archaeological Skeleton—​a Review’, Journal of Archaeological Method and Theory, 13: 135–​187. Birley, A. (1979). The People of Roman Britain. London: Batsford.

218

218    Hella Eckardt and Gundula Müldner Brather, S. (2000). ‘Ethnische Identitäten als Konstrukte der frühgeschichtlichen Archäologie’, Germania, 78: 139–​177. Brather, S. (2004). Ethnische Interpretationen in der frühgeschichtlichen Archäologie: Geschichte, Grundlagen und Alternativen. Berlin: Walther de Gruyter. Brettell, R., Montgomery, J., and Evans, J. (2012). ‘Brewing and Stewing:  The Effect of Culturally-​Mediated Behaviour on the Oxygen Isotope Composition of Ingested Fluids and the Implications for Human Provenance Studies’, Journal of Analytical Atomic Spectrometry, 27/​5: 778–​785. Bruun, C. (2010). ‘Water, Oxygen Isotopes and Immigration to Ostia-​Portus’, Journal of Roman Archaeology, 23: 109–​132. Budd, P., Millard, A., Chenery, C., Lucy, S., and Roberts, C. (2004). ‘Investigating Population Movement by Stable Isotope Analysis: A Report from Britain’, Antiquity, 78: 127–​141. Carroll, M. (2006). Spirits of the Dead: Roman Funerary Commemoration in Western Europe. Oxford: Oxford University Press. Casella, E. C. (2005). ‘Prisoner of His Majesty: Post-​Coloniality and the Archaeology of British Penal Transportation’, World Archaeology, 37/​3: 453–​467. Chenery, C. A., Müldner, G., Evans, J., Eckardt, H., Leach, S., and Lewis, M. (2010). ‘Strontium and Stable Isotope Evidence for Diet and Mobility in Roman Gloucester, UK’, Journal of Archaeological Science, 37: 150–​163. Chenery, C., Müldner, G., and Eckardt, H. (2011). ‘Cosmopolitan Catterick? Isotopic Evidence for Population Mobility on Rome’s Northern Frontier’, Journal of Archaeological Science, 38: 1525–​1536. Clarke, G. (1979). Pre-​Roman and Roman Winchester. Part 2: The Roman Cemetery at Lankhills. Oxford: Clarendon Press. Clay, C. (2007). ‘Before there were Angles, Saxons and Jutes: An Epigraphic Study of the Germanic Social, Religious and Linguistic Relations on Hadrian’s Wall’, in L. Gilmour (ed.), Pagans and Christians:  From Antiquity to the Middle Ages. Papers in Honour of Martin Henig, Presented on the Occasion of his 65th birthday. Oxford:  Tempus Reparatum, 47–​63. Cohen, R. (2008). Global Diasporas: An Introduction. London: Routledge. Cool, H. (2004). The Roman Cemetery at Brougham, Cumbria:  Excavations 1966–​ 67. London: Society for the Promotion of Roman Studies. Cool, H. E.  M. (2006a). Eating and Drinking in Roman Britain. Cambridge:  Cambridge University Press. Cool, H. (2006b). ‘Sustenance in a Strange Land’, in P. Ottaway (ed.), A Victory Celebration: Papers on the Archaeology of Colchester and Late Iron-​Age-​Roman Britain presented to Philip Crummy. Colchester: Colchester Archaeological Trust, 75–​82. Cool, H. E. M. (2010). ‘Finding the Foreigners’, in H. Eckardt (ed.), Roman Diasporas: Archaeological Approaches to Mobility and Diversity in the Roman Empire. Journal of Roman Archaeology Supplementary Series 78. Portsmouth, RI: Journal of Roman Archaeology, 27–​44. Curta, F. (2007). ‘Some Remarks on Ethnicity in Medieval Archaeology’, Early Medieval Europe, 15/​2: 159–​185. Daux, V., Lécuyer, C., Héran, M.-​A., Amiot, R., Simon, L., Fourel, F., Martineau, F., Lynnerup, N., Reychler, H., and Escarguel, G. (2008). ‘Oxygen Isotope Fractionation between Human Phosphate and Water Revisited’, Journal of Human Evolution, 55/​6: 1138–​1147. Derks, T., and Roymans, N. (2009) (eds). Ethnic Constructs in Antiquity: The Role of Power and Tradition. Amsterdam: Amsterdam University Press.

Mobility, Migration, and Diasporas in Roman Britain    219 Eckardt, H. (2010). ‘A Long Way from Home: Diaspora Communities in Roman Britain’, in H. Eckardt (ed.), Roman Diasporas: Archaeological Approaches to Mobility and Diversity in the Roman Empire. Journal of Roman Archaeology Supplementary Series 78. Portsmouth, RI: Journal of Roman Archaeology, 99–​130. Eckardt, H. (2012). ‘Foreigners and Locals in Calleva’, in M. Fulford (ed.), Silchester and the Study of Romano-​British Urbanism. Journal of Roman Archaeology Supplementary Series 89. Portsmouth, RI: Journal of Roman Archaeology, 246–​256. Eckardt, H., Booth, P., Chenery, C., Müldner, G., Evans, J. A., and Lamb, A. (2009). ‘Oxygen and Strontium Isotope Evidence for Mobility in Roman Winchester’, Journal of Archaeological Science, 36: 2816–​2825. Evans, J., Stoodley, N., and Chenery, C. (2006). ‘A Strontium and Oxygen Isotope Assessment of a Possible Fourth Century Immigrant Population in a Hampshire Cemetery, Southern England’, Journal of Archaeological Science, 33: 265–​272. Evans, J. A., Chenery, C. A., and Montgomery, J. (2012). ‘A Summary of Strontium and Oxygen Isotope Variation in Archaeological Human Tooth Enamel Excavated from Britain’, Journal of Analytical Atomic Spectrometry, 27/​5: 754–​764. Fitzpatrick, A. P. (2011). The Amesbury Archer and the Boscombe Bowmen: Bell Beaker Burials at Boscombe Down, Amesbury, Wiltshire. Salisbury: Wessex Archaeology. Fulford, M. (2010). ‘Roman Britain: Immigration and Material Culture’, in H. Eckardt (ed.), Roman Diasporas: Archaeological Approaches to Mobility and Diversity in the Roman Empire. Journal of Roman Archaeology Supplementary Series 78. Portsmouth, RI: Journal of Roman Archaeology, 67–​78. Gardner, A. (2007). An Archaeology of Identity: Soldiers and Society in Late Roman Britain. Walnut Creek, CA: Leftcoast Press. Gosden, C. (2006). ‘Race and Racism in Archaeology: Introduction’, World Archaeology, 38/​1: 1–​7. Gray, C. L. (2011). ‘Foreigners in the Burial Ground: The Case of the Milesians in Athens’, in M. Carroll and J. Rempell (eds), Living through the Dead: Burial and Commemoration in the Classical World. Oxford: Oxbow Books, 47–​64. Handley, M. (2011). Dying on Foreign Shores: Travel and Mobility in the Late Antique West. Portsmouth, RI: Journal of Roman Archaeology. Härke, H. (2007). ‘Ethnicity, “Race” and Migration in Mortuary Archaeology:  An Attempt at a Short Answer’, in S. Semple and H. Williams (eds), Early Medieval Mortuary Practices. Oxford: Oxbow Books, 12–​18. Hingley, R. (2010). ‘Tales of the Frontier: Diasporas on Hadrian’s Wall’, in H. Eckardt (ed.), Roman Diasporas: Archaeological Approaches to Mobility and Diversity in the Roman Empire. Journal of Roman Archaeology Supplementary Series 78. Portsmouth, RI: Journal of Roman Archaeology, 227–​243. Hingley, R. (2012). Hadrian’s Wall: A Life. Oxford: Oxford University Press. Hoppe, K. A., Koch, P. L., and Furutani, T. T. (2003). ‘Assessing the Preservation of Biogenic Strontium in Fossil Bones and Tooth Enamel’, International Journal of Osteoarchaeology, 13/​ 1–​2: 20–​28. Isaac, B. (2004). The Invention of Racism in Classical Antiquity. Princeton:  Princeton University Press. Isaac, B. (2006). ‘Proto-​Racism in Graeco-​Roman Antiquity’, World Archaeology, 38/​1: 32–​47. Ivleva, T. (2010). ‘British Emigrants in the Roman Empire: Complexities and Symbols of Ethnic Identities’, in D. Mladenović and B. Russell (eds), TRAC 2010: Proceedings of the Twentieth Theoretical Roman Archaeology Conference. Oxford: Oxbow Books, 132–​153.

220

220    Hella Eckardt and Gundula Müldner Jones, S. (1997). The Archaeology of Ethnicity: Constructing Identities in the Past and Present. New York and London: Routledge. Jones, B., and Mattingly, D. (2002). An Atlas of Roman Britain. Oxford: Oxbow Books. Kakoschke, A. (2002). Ortsfremde in den römischen Provinzen Germania inferior und Germania superior. Möhnesee: Bibliopolis. Killgrove, C. (2010a). ‘Response to Bruun, 2010’, Journal of Roman Archaeology, 23: 133–​136. Killgrove, K. (2010b). ‘Identifying Immigrants to Imperial Rome Using Strontium Isotope Analysis’, in H. Eckardt (ed.), Roman Diasporas: Archaeological Approaches to Mobility and Diversity in the Roman Empire. Journal of Roman Archaeology Supplementary Series 78. Portsmouth, RI: Journal of Roman Archaeology, 157–​175. Knudson, K. J. (2009). ‘Oxygen Isotope Analysis in a land of Environmental Extremes: The Complexities of Isotopic Work in the Andes’, International Journal of Osteoarchaeology, 19: 171–​191. Lawrence, S. (2003). ‘Archaeology and the Nineteenth-​Century British Empire’, Historical Archaeology, 37/​1: 20–​33. Leach, S., Lewis, M., Chenery, C., Eckardt, H., and Müldner, G. (2009). ‘Migration and Diversity in Roman Britain: A Multidisciplinary Approach to Immigrants in Roman York, England’, American Journal of Physical Anthropology, 140: 546–​561. Leach, S., Eckardt, H., Chenery, C., Müldner, G., and Lewis, M. (2010). ‘A “Lady” of York: Migration, Ethnicity and Identity in Roman York’, Antiquity, 84: 131–​145. Lilley, I. (2004). ‘Diaspora and Identity in Archaeology: Moving beyond the Black Atlantic’, in L. Meskell and R. W. Preucel (eds), A Companion to Social Archaeology. Oxford: Blackwell, 287–​312. Lilley, I. (2006). ‘Archaeology, Diaspora and Decolonization’, Journal of Social Archaeology, 6/​ 1: 28–​47. McCoskey, D. E. (2012). Race—​Antiquity and its Legacy. London: I. B. Tauris. Mattingly, D. (2004). ‘Being Roman: Expressing Identity in a Provincial Setting’, Journal of Roman Archaeology, 17: 5–​25. Mattingly, D. (2006). An Imperial Possession:  Britain in the Roman Empire, 54 bc–​ad 409. London: Penguin Books. Meskell, L. (2001). ‘Archaeologies of Identity’, in I. Hodder (ed.), Archaeological Theory Today. Cambridge: Polity, 187–​213. Mihailescu-​ Bîrliba, L. (2011). Ex Toto Orbe Romano:  Immigration into Roman Dacia. Leuven: Peeters. Millett, M. (1990). The Romanization of Britain. Cambridge: Cambridge University Press. Millett, M. (1999). ‘Coloniae and Romano-​British Studies’, in H. Hurst (ed.), The Coloniae of Roman Britain: New Studies and a Review. Journal of Roman Archaeology Supplementary Series 36. Portsmouth, RI: Journal of Roman Archaeology, 191–​196. Montgomery, J. (2010). ‘Passports from the Past:  Investigating Human Dispersals Using Strontium Isotope Analysis of Tooth Enamel’, Annals of Human Biology, 37/​3: 325–​346. Montgomery, J., Evan, J., Chenery, C., Pashley, V., and Killgrove, K. (2010). ‘“Gleaming white and deadly”: Using Lead to Track Human Exposure and Geographic Origins in Britain’, in H. Eckardt (ed.), Roman Diasporas: Archaeological Approaches to Mobility and Diversity in the Roman Empire. Journal of Roman Archaeology Supplementary Series 78. Portsmouth, RI: Journal of Roman Archaeology, 199–​226. Müldner, G. (2013). ‘Stable Isotopes and Diet: Their Contribution to Romano-​British Research’, Antiquity, 87/​335: 137–​149.

Mobility, Migration, and Diasporas in Roman Britain    221 Müldner, G., Chenery, C., and Eckardt, H. (2011). ‘“The Headless Romans”:  Multi-​Isotope Investigations of an Unusual Burial Ground from Roman Britain’, Journal of Archaeological Science, 38: 280–​290. Noy, D. (2000). Foreigners at Rome: Citizens and Strangers. London: Duckworth. Noy, D. (2010). ‘Epigraphic Evidence for Immigrants at Rome and in Roman Britain’, in H. Eckardt (ed.), Roman Diasporas:  Archaeological Approaches to Mobility and Diversity in the Roman Empire. Journal of Roman Archaeology Supplementary Series 78. Portsmouth, RI: Journal of Roman Archaeology, 13–​26. Parker Pearson, M., Richards, M. P., Chamberlain, A., and Jay, M. (forthcoming) (eds). The Beaker People:  Isotopes, Mobility and Diet in Prehistoric Britain. Oxford: Oxbow Books. Pearce, J. (2010). ‘Burial, Identity and Migration in the Roman World’, in H. Eckardt (ed.), Roman Diasporas: Archaeological Approaches to Mobility and Diversity in the Roman Empire. Journal of Roman Archaeology Supplementary Series 78. Portsmouth, RI: Journal of Roman Archaeology, 79–​98. Perry, M. A., Coleman, D., and Delhopital, N. (2008). ‘Mobility and Exile at Second Century ad Khirbet edh-​Dharih: Strontium Isotope Analysis of Human Migration in Western Jordan’, Geoarchaeology, 23/​4: 528–​549. Perry, M. A., Coleman, D. S., Dettmann, D. L., and Halim al-​Shiyab, A. (2009). ‘An Isotopic Perspective on the Transport of Byzantine Mining Camp Laborers into South-​Western Jordan’, American Journal of Physical Anthropology, 140: 429–​441. Perry, M. A., Coleman, D. S., Dettman, D. L., Grattan, J. P., and Halim al-​Shiyab, A. (2011). ‘Condemned to Metallum? The Origin and Role of 4th–​6th Century ad Phaeno Mining Camp Residents Using Multiple Chemical Techniques’, Journal of Archaeological Science, 38/​ 3: 558–​569. Pohl, W. (1998). ‘Telling the Difference: Signs of Ethnic Identity’, in W. Pohl and H. Reimitz (eds), Strategies of Distinction:  The Construction of Ethnic Communities, 300–​800. Leiden: Brill, 1–​69. Pollard, A. M. (2011). ‘Isotopes and Impact: A Cautionary Tale’, Antiquity, 85: 631–​638. Pollard, A. M, Ditchfield, P., McCullagh, J. S. O., Allen, T. G., Gibson, M., and Boston, C. (2011a). ‘“These boots were made for walking”:  The Isotopic Analysis of a C4 Roman Inhumation from Gravesend, Kent, UK’, American Journal of Physical Anthropology, 146: 446–​456. Pollard, A. M., Pellegrini, M. and Lee-​Thorp, J. A. (2011b). ‘Technical Note: Some Observations on the Conversion of Dental Enamel Delta O-​18(p) Values to delta O-​18(w) to Determine Human Mobility’, American Journal of Physical Anthropology, 145/​3: 499–​504. Price, T. D., Burton, J. H., Sharer, R. J., Buikstra, J. E., Wright, L. E., Traxler, L. P., and Miller, K. A. (2010). ‘Kings and Commoners at Copan: Isotopic Evidence for Origins and Movement in the Classic Maya Period’, Journal of Anthropological Archaeology, 29/​1: 15–​32. Prowse, T., Schwarcz, H. P., Garnsey, P., Knyf, M., Macciarelli, R., and Bondioli, L. (2007). ‘Isotopic Evidence for Age-​Related Immigration to Imperial Rome’, American Journal of Physical Anthropology, 132: 510–​519. Prowse, T. L., Barta, J. L., von Hunnius, T. E., and Small, A. M. (2010). ‘Stable Isotope and Mitochondrial DNA Evidence for Geographic Origins on a Roman Estate at Vagnari (Italy)’, in H. Eckardt (ed.), Roman Diasporas: Archaeological Approaches to Mobility and Diversity in the Roman Empire. Journal of Roman Archaeology Supplementary Series 78. Portsmouth, RI: Journal of Roman Archaeology, 175–​197.

222

222    Hella Eckardt and Gundula Müldner Rowland, R. J. (1976). ‘Foreigners in Roman Britain’, Acta Archaeologica Academiae Scientiarium Hungaricae, 28: 443–​447. Roymans, N. (2004). Ethnic Identity and Imperial Power: The Batavians in the Early Roman Empire. Amsterdam: Amsterdam University Press. Scheidel, W. (2004). ‘Human Mobility in Roman Italy, I: The Free Population’, Journal of Roman Studies, 94: 1–​26. Schweissing, M. M., and Grupe, G. (2003). ‘Stable Strontium Isotopes in Human Teeth and Bone:  A  Key to Migration Events of the Late Roman Period in Bavaria’, Journal of Archaeological Science, 30: 1373–​1383. Sherwin-​White, A. N. (1967). Racial Prejudice in Imperial Rome. Cambridge:  Cambridge University Press. Snowden, F. M. (1983). Before Color Prejudice:  The Ancient View of Blacks. Cambridge, MA: Harvard University Press. Stockhammer, P. F. (2012). ‘Conceptualising Cultural Hybridization in Archaeology’, in P. F. Stockhammer (ed.), Conceptualising Cultural Hybridization: A Transdisciplinary Approach. Heidelberg: Springer-​Verlag, 43–​58. Swan, V. G. (1992). ‘Legio VI and its Men: African Legionaries in Britain’, Journal of Roman Pottery Studies, 5: 1–​33. Swan, V. G. (1999). ‘The Twentieth Legion and the History of the Antonine Wall Reconsidered’, Proceedings of the Society of Antiquaries of Scotland, 129: 399–​480. Swan, V. G. (2009). Ethnicity, Conquest and Recruitment: Two Case Studies from the Northern Military Provinces. Journal of Roman Archaeology Supplementary Series 72. Portsmouth, RI: Journal of Roman Archaeology. Swift, E. (2010). ‘Identifying Migrant Communities:  A  Contextual Analysis of Grave Assemblages from Continental Late Roman Cemeteries’, Britannia, 41: 237–​282. Tolia-​Kelly, D. (2010). ‘Narrating the Postcolonial Landscape:  Archaeologies of Race at Hadrian’s Wall’, Transactions of the Institute of British Geographers, ns 36: 71–​88. Van Dommelen, P., and Terrenato, N. (2007). ‘Local Cultures and the Expanding Roman Republic’, in P. Van Dommelen and N. Terrenato (eds), Articulating Local Cultures: Power and Identity under the Expanding Roman Republic. Journal of Roman Archaeology Supplementary Series 63. Portsmouth, RI: Journal of Roman Archaeology, 7–​12. Voss, B. (2005). ‘The Archaeology of Overseas Chinese Communities’, World Archaeology, 37/​ 3: 424–​439. Webster, J. (2001). ‘Creolizing Roman Britain’, American Journal of Archaeology, 105/​ 2: 209–​225. Webster, J. (2010). ‘Routes to Slavery in the Roman World:  A  Comparative Perspective on the Archaeology of Forced Migration’, in H. Eckardt (ed.), Roman Diasporas: Archaeological Approaches to Mobility and Diversity in the Roman Empire. Journal of Roman Archaaeology Supplementary Series 78. Portsmouth, RI: Journal of Roman Archaeology, 45–​65. West, J. B., Ehleringer, J. R., and Cerling, T. E. (2007). ‘Geography and Vintage Predicted by a Novel GIS Model of Wine δ18O’, Journal of Agricultural and Food Chemistry, 55: 7075–​7083. Whittle, A., and Bickle, P. (2013) (eds). The First Farmers of Central Europe: Diversity in LBK Lifeways. Oxford: Oxbow Books. Wierschowski, L. (1995). Die regionale Mobilität in Gallien nach den Inschriften des 1.  bis 3. Jahrhunderts n. Chr. Stuttgart: Franz Steiner Verlag.

Mobility, Migration, and Diasporas in Roman Britain    223 Wierschowski, L. (2001). Fremde in Gallien—​‘Gallier’ in der Fremde: Die epigraphisch bezeugte Mobilität in, von und nach Gallien vom 1. bis 3. Jh. n. Chr. Stuttgart: Franz Steiner Verlag. Wilkie, L. A. (2004). ‘Considering the Future of African-​American Archaeology’, Historical Archaeology, 38/​1: 109–​123. Woolf, G. 1998. Becoming Roman:  The Origins of Provincial Civilization in Gaul. Cambridge: Cambridge University Press.

224

Chapter 11

Multicu ltu ra l i sm on Hadrian’ s  Wa l l Claire Nesbitt

Introduction During much of the Roman occupation of Britain a great wall marked the northern frontier of the Roman Empire. This frontier, running the width of the province across the Tyne–​Solway isthmus, delimited the northern boundary of empire from ad 122 to c. ad 142, when occupation was briefly pushed north to the central belt of Scotland and demarcated by the Antonine Wall. The Wall became the frontier again from c. ad 160, when the Roman military withdrew from Scotland, and it remained in use as a frontier or barrier until the fifth century. Construction began in ad 122, during the reign of the emperor Hadrian after whom the Wall is now named, with a turf and stone wall, which was later replaced entirely with stone. The Wall has been a focus of study and a destination for antiquarians, archaeologists, and the public for around 200 years. Throughout that time the monument has been appropriated by different groups for different socio-​political and economic reasons (see Nesbitt and Tolia-​Kelly 2009; Witcher et al. 2010; Hingley 2012). In the modern popular imagination, Hadrian’s Wall represents a militarized zone stamped with the indelible impression of ‘Rome’ (see Witcher 2010). Indeed, archaeological investigation of the Wall has focused heavily on its military aspects (e.g. Divine 1969; Poulter 2009). However, as well as being a military frontier, the Wall zone also represents a contested landscape of polyvocalities (see Hingley and Hartis 2011). The Wall landscape has stories of a pre-​Roman heritage to tell, biographies of peoples who were not Roman and, perhaps, not native to the region. These individuals and groups are often eclipsed in archaeological studies by the shadow of the Roman military machine. This chapter will consider the multicultural nature of Hadrian’s Wall; that is to say, it will focus on the presence and influence of ‘otherness’ evidenced in the material culture and epigraphy left to us in the archaeological record. It will investigate what Hingley (2010: 229) has described as the ‘dichotomy between the

Multiculturalism on Hadrian’s Wall    225 Wall’s nature as a contested landscape and the inclusivity suggested by the material culture’ and it will focus on concepts of identity as expressed through materiality. Before we turn to the archaeological evidence, however, I would like first to establish what we mean by a ‘foreign’ presence in an empire as vast and diverse as Rome, and to spend some time thinking through how we understand ‘multiculturalism’ and ‘identity’ in the archaeological past.

What is Multiculturalism? Multiculturalism is a word that is often used but seldom explained. Its definitions vary considerably from Modood’s interpretation (2007) that ‘multiculturalism reflects the cultural and racial diversity of a society and acknowledges the freedom of all members of that society to preserve, enhance and share their cultural heritage’ (Modood 2007: 16), to Adler’s understanding (1976) that the multicultural citizen has indefinite boundaries of the self, an identity that is in permanent flux and always being reformulated in response to cultural input. Steinberg and Kincheloe (2009: 4) argue for several incarnations of multiculturalism. One extreme, taking a conservative perspective, expounds occidental superiority, which entails an assimilation to western culture by other societies, and is essentially a promotion of a monoculture rather than multiculturalism; another, a pluralist multiculturalism, focuses on race, class, and gender differences, promotes pride in group heritage, and avoids the use of oppression. So what do we mean when we speak of a multicultural society in the Roman Empire? The variety of cultural and racial identities found across the Roman Empire has long been recognized. Witcher has argued that ‘the key to any investigation of Roman Italy must surely lie in the articulation of these multiple identities—​national, regional and local; social, political and ethnic’ (Witcher 2000: 214; emphasis added). This might just as effectively be applied empire-​wide. The crucial factor in discussing multiculturalism in the Roman world is the definition of multiple individual and group identities within the context of a unifying administration and the fluidity of cultural expression that this allows.

Shaping Identities in a Frontier Zone Identity has been defined as an ‘individual’s identification with broader groups on the basis of differences socially sanctioned as significant’ (Diaz-​Andreu and Lucy 2005: 1). Identity is made up of layers of different social signifiers including gender, religion, age, and ethnicity. Here I follow Lucy’s premise (2005: 86) that some aspects of identity, such as ethnicity, are ways of behaving rather than inherent qualities and, as such, must be learned and can cross-​cut other aspects of identity. Ethnicity also ‘involves the social

226

226   Claire Nesbitt negotiation of difference and sameness, and often entails larger tensions between individuals, the group, and the state’ (Meskell 2007: 25). Who, then, do we expect to be able to identify in the archaeological record of the Wall zone? There are several different groups that we might expect to recognize in the material culture of Hadrian’s Wall. The indigenous communities who existed in the region at the time of the Wall’s construction, for example; also the Roman military and perhaps other groups and identities associated with the military. Alongside this we might expect to see a range of individuals and communities who came to the region through trade, employment, or slavery. Traditionally the ‘Britons’ are the indigenous people who occupied the Wall zone before the Roman conquest and probably continued to live there, though Hodgson et al. (2012) have argued for a widespread dislocation and abandonment of sites on the Northumberland coastal plain in the early Roman period. While the exact nature of the identities of the pre-​Roman Iron Age peoples in the frontier zone has seldom been explored in any detail, indigenous peoples have traditionally been assigned to tribal groups based on the evidence from classical sources, most notably Ptolemy’s Geographia and Tacitus’ Agricola (Breeze 1982: 28–​36). These have been identified as the Brigantes, the Carvetii, and, more recently, the Anavionenses (Hunter 2007b: 7–​8). Increasingly, there has been recognition that such tribal models potentially mask a more complex socio-​political landscape and range of identities in the late Iron Age, and that these were in a state of reconfiguration at the time of Roman conquest (Moore 2011). In his work on the expression of identity through material culture in frontier zones, Hunter (2001, 2007a) has considered the adoption of Roman material culture by native society as prestige objects or objects of power, particularly in relation to identity in northern Britain. Hunter (2001) concludes that social control of Roman artefacts ensured that foreign or exotic objects had a complex role in social display, bribery, and gift-​giving in the frontier zone. Hunter’s studies also demonstrate that identities in the region were in a state of transformation at the time of the Wall’s construction and use (Hunter 2007a). This process had distinct regional differences, however; in Scotland, trends in material culture were towards a distinctly local style, seen very much in opposition to external forces. Meanwhile, in the frontier zone, material that began as an expression of local identity was absorbed into a Romano-​British milieu and interpreted differently by different people (Hunter 2007a: 293–29​4). Concepts of identity, then, appear to have been fluid, and there may have been a blurring of cultural distinction in areas like the frontier zone, as demonstrated by the inclusion of Celtic-​style motifs on Roman objects (Hunter 2007a: 292). Cultural influence is not a one-​way process, and the frontier zone can be seen as a melting pot of peoples, ideas, and cultures all in a dynamic state of reshaping themselves into a Romano-​British ‘frontier’ culture. Moore (2011) has argued that, in the later pre-​Roman Iron Age, rather than being a traditional tribalized society, communities were reshaping their identities in response to the expansion of Rome, perhaps in order to establish themselves in society in a new way through their interactions with the invading peoples. In this way, groups along the frontier may have re-​evaluated their own identity and perhaps even reshaped it in response to Rome’s

Multiculturalism on Hadrian’s Wall    227 presence. Moore (2011: 347) cautions against confusing identities, as expressed in the Roman province, with those situated in a pre-​Roman context because of the influence of Roman power. In addition, it seems more likely that most of these individuals and communities that already existed in the Wall zone would have identified themselves more at the level of kin group, settlement community, or sets of social and exchange networks. Archaeological investigations of pre-​Roman sites in the vicinity of the Wall are relatively few in number, reflecting the tendency to focus on military infrastructure, rather than evaluating the social identity of the peoples who lived there. As a consequence, our understanding of how pre-​Roman indigenous people in the region constructed their identities is unclear, although almost certainly this was at a scale far smaller than at the level of a ‘tribe’ like the Brigantes. Anderson’s analysis (2011) of regional ceramics from the later pre-​Roman Iron Age and post-​conquest period suggests that, while there is an identifiable shift in focus from preparation to consumption involving more ceramic vessels, overall the ceramic record reveals that identity was expressed at the level of small-​scale groups. At the same time, it is easy to underestimate the impact of Roman occupation on the existing population. Breeze (1990: 86) has noted the myriad ways in which society must have been affected, through taxation, education, language, conscription, alternative farming practices, redistribution of the population in terms of settlement patterns, the structure of society, and, indeed, the size of the population. These factors could easily lead to a complete transfiguring of a sense of individual, local, and regional identity, which may have been played out in the abandonment and dislocation that Hodgson et al. (2012) have suggested.

Military Identities: Join the Army—​See the World? By the end of the second century ad each legion was recruited mainly from the province in which it was stationed (Dobson and Mann 1973: 193). That meant that recruitment of legionaries from Britain was usual from the same time (Swan and Monaghan 1993: 26), while auxiliary units were generally raised from single groups of peoples or from provinces (Dobson and Mann 1973: 193). The civil settlements that developed around the military garrisons probably gradually became pools of hereditary recruits who provided most of the soldiery for the third century and later still on the frontier when hereditary military service for the sons of servicemen became compulsory (Dobson and Mann 1973). The military is one of the more obvious and easily traceable ways to find a foreign presence on the frontier, though what that means in terms of cultural identity is not straightforward and will be discussed further below. Book XL of the Notitia Dignitatum, originally compiled in the early fifth century, lists the Item per lineam valli—​the route along the line of the Wall. The book lists the forts and

228

228   Claire Nesbitt the units serving at them and this, together with epigraphic evidence, has enabled us to people the landscape with the foreign military units. From this evidence we can identify units from all over Europe, the Middle East, and North Africa (Figure 11.1 and Table 11.1). It would be easy to group the units by their nationality and look for a cultural identity in the materiality of their locations that reflects this. However, as has been argued, identity is not as straightforward as that. Gardner (2002: 338) sees the boundedness of the military communities as ‘blurred by the presence of people who were not members in the organizational sense, but were nonetheless strongly tied to the soldiers, through relations of friendship, marriage, enslavement, clientship or kinship’. This introduces a tension between the complex and perhaps multiple identities that were enacted in the frontier zone (Gardner 2002: 338). Mattingly has argued that ‘military identity was far from static and was continually refashioned, inevitably influenced by the cultural background of the soldiery’ (Mattingly 2010: 167). James (2011: 28) notes that at the height of empire around half of Rome’s soldiers were provincials who had acquired citizenship by virtue of their military careers and argues for a particular military-​centred, frontier-​ focused ‘Romanness’ that was created out of and probably transcended the ethnic diversity. Great cultural variability reflected the diversity and humanity of the army, but there were also ‘significant common patterns of behaviour and social outlook that made the army an empire-wide and self-​aware identity group of commilitones—​“fellow soldiers”’ (Mattingly 2010: 199). What this means, in terms of the identities of the soldiers, is that, when there were familial or local connections with them in their day-​to-​day existence, their unit identity may have been second to other complex identities shaped by their interactions with other, non-​military companions. James (2002: 39) has viewed soldiers ‘not as automata, cogs in the machine, but as self-​aware social agents’ and challenges the military/​civilian dichotomy that is the established framework of research. Gardner echoes this, suggesting that ‘in the daily lives of such people, unit, rank and status identities are likely to have been more significant than an empire-​wide “militariness”’ (Gardner 2002: 338). So though we can identify people from all over the empire, their cultural identity may not have remained as rigidly military, national, or ethnic as we may imagine. It has been argued that, within a generation of a regiment leaving its area of origin, its ethnic title became meaningless (Mann 1963). However, Haynes (1999: 165) understands the military culture differently, preferring to see regimental communities as more than simply militarized; rather, the units were exposed to a range of symbolism, experiences, values, ideas, and artefacts that were not exclusively military. In fact, he, like Gardner, argues that civilian followers probably countered the military culture with other local traditions that were maintained. The dilution of a culture through connection with the military does not necessarily indicate an abandonment of national or ethnic traditions. Haynes (1999: 167) cites evidence for the survival of ethnic tradition in the vici from Birdoswald, where a third-​century inscription commemorates Decebalus, a boy with a distinctively Dacian name (RIB 1920). The fact that this name was still in use at the site 50–​100 years after the Dacians serving in Cohors I Aelia Dacorum arrived there suggests that at least some aspects of Dacian identity survived in the vicus beyond the

Figure 11.1  Map showing areas of the Roman Empire with military units on Hadrian’s Wall. Source: © Rob Witcher, illustration by Christina Unwin.

230

Table 11.1 Foreign units serving on Hadrian’s Wall and their provincial origins Provincial origin of unit

Unit name

Britannia

I Cornoviorum

Hispania

Ala I Hispanorum Asturum Ala II Asturum I Aelia Hispanorum equitata II Asturum equitata I Fida Vardullorum equitata

Gallia

Ala Augusta Gallorum Petriana I Aquitanorum equitata IV Gallorum equitata V Gallorum equitata

Gallia Belgica

I Batavorum equitata II Nerviorum VI Nerviorum I Tungrorum II Tungrorum equitata

Germania

I Baetasiorum I Lingonum equitata II Lingonum equitata IV Lingonum equitata I Nervana Germanorum equitata Numerus Hnaudifridi I Ulpia Traiana Cugernorum I Vangionum equitata

Frisia

Cuneus Frisionum Aballavensium Cuneus Frisiorum Vercovicianorum I Frisiavonum

Raetia

Raeti gaesati Cohors Raetorum Vexillatio gaesatorum Raetorum

Pannonia

Ala I Pannoniorum Sabiniana (continued)

Multiculturalism on Hadrian’s Wall    231 Table 11.1 Continued Provincial origin of unit

Unit name I Pannoniorum equitata II Pannoniorum equitata

Dalmatia

I Delmatarum equitata II Delmatarum equitata

Dacia

I Aelia Dacorum

Thracia

I Thracum II Thracum equitata

Syria

I Hamiorum sagittariorum

Mesopotamia

Numerus barcariorum Tigrisiensium

Mauretania

Numerus Maurorum Aurelianorum

Source: © Claire Nesbitt.

first generation of non-​native troops (Haynes 1999: 167). This should not be surprising: belonging to a military cohort and the unity that this embodies can operate completely independently of ethnic or cultural heritage, as James (2011: 28) has suggested. The regimental customs of the unit, while having their origins in localized traditions, could be adopted wholeheartedly by new (and foreign) recruits as part of a recognized unit culture and tradition to which they could aspire and be proud to belong. Perhaps, rather than looking at what distinguished the multinational identities in the frontier zone, we ought to consider what it was that united them, the sense of being ‘Roman’.

Shades of Romanness Much consideration has been given to concepts of Romanization since Haverfield’s initial work (1912) on the subject. Millett (1990: 212) has conceived a framework within which Romanization is seen ‘not as a passive reflection of change, but rather as an active ingredient used by people to assert, project and maintain their social status’. Woolf (1998: 7) argues that ‘there was no standard Roman civilization against which provincial cultures might be measured’. He adds that Romanization did not culminate in a uniform culture throughout the empire (Woolf 1998: 7). In Woolf ’s view, ‘becoming Roman did not involve becoming more alike the other inhabitants of the empire, so much as participating in a cultural system structured by systematic differences, differences that both sustained and were a product of Roman power’ (Woolf 1998: 242).

232

232   Claire Nesbitt The Rome of the ancient world can be described as an idea, a concept, rather than a locus that can be pointed to or visited. As an empire Rome was necessarily multicultural, absorbing so many nations and peoples. This meant, inevitably, that the Senate and Imperial palace were themselves international, so that ‘Romanness’ was no longer determined by nationality or ethnicity but by citizenship and sharing in the ‘idea’ of Rome. In this sense it is possible to be completely foreign and completely Roman at the same time, to have multiple identities that cross traditional boundaries, because the growth of the Roman Empire, as Woolf (1997: 347) has argued, was not ‘the expansion of one national or ethnic culture at the expense of others’, but rather a new social formation that incorporated new cultural logic and new configurations of power. Because of the ill-​defined and blurred nature of the cultural and ethnic implications of the term ‘Roman’, the Roman attitude to other peoples focused more on civilization and citizenship than it did on nationality or ethnicity. Citizenship was a crucial part of ‘Romanness’ or Romanitas; being a Roman citizen—​Civis Romanus—​meant belonging to the Roman state, and it entailed certain rights and privileges. Citizenship, however, was a complex concept; there were several different levels of citizenship, and these are not always clearly defined. Full Roman citizens (Cives Romani) were protected by Roman law (ius civile). One of the best examples of the importance of citizenship can be drawn from Christian scripture. The Acts of the Apostles describes the events that unfolded when the Apostle Paul was preaching in Jerusalem. Paul was arrested by the Roman authorities because the monotheistic doctrine of Christianity was incompatible with the polytheism of the Roman Empire, meaning that Christians refused to undertake the required sacrifice to Roman gods. He was to be detained and flogged for breaking the law. Paul’s father had been a Roman citizen and Paul inherited this status. This posed a problem for his captors who had planned to flog him: Paul said to the centurion who was standing there: ‘Can you legally flog a man who is a Roman citizen, and moreover has not been found guilty?’ When the centurion heard this, he went and reported it to the commandant. ‘What do you mean me to do?’ he said. ‘This man is a Roman citizen.’ The commandant came to Paul. ‘Tell me, are you a Roman citizen?’ he asked. ‘Yes,’ said he. The commandant rejoined, ‘It cost me a large sum to acquire this citizenship.’ Paul said, ‘But it was mine by birth.’ (Acts 22: 22–​8)

Two things stand out from this passage. The first is the impact that the realization of Paul’s citizenship had on his captors and the way it influenced their treatment of him. The second thing that emerges is an insight into the nature of acquiring citizenship; we have, on the one hand, Paul, who tells us citizenship was his by right of birth and, on the other hand, the commandant, who bought his citizenship, no doubt at a high price. Individuals could be granted citizenship for acts of outstanding service to Rome, and provinces could be granted citizenship—​usually a limited version, such as Latini. Full citizenship was often granted to auxiliary soldiers at the end of their military service of twenty-​five years. Several military diplomas survive that set out the conditions of the

Multiculturalism on Hadrian’s Wall    233 soldiers’ retirement. With citizenship, whether inherited, purchased, or granted, peoples from all over the empire, of any ethnic background, could rise through the ranks of Roman society. They were not classified by their nationality, their heritage, or the colour of their skin; they were citizens of Rome, and this was the common factor that unified the multicultural empire. In terms of culture in the Roman Empire, then, Mann has argued that ‘individual societies do not exist, instead we should think in terms of overlapping social networks of varying scales relating to different types of social power, ideological, economic, military or political’ (Mann 1986, cited in Shennan 1989: 11). However, equally it can be argued that ‘the process of ethnic identity creation only comes to have its power in a situation in which pre-​existing forms of identity creation and maintenance—​kinship for example—​ are being destroyed’ (Shennan 1989: 16). If Romanness was desired and welcomed, it had the potential to dilute cultural traditions, to foster inclusivity and a cosmopolitan shared identity. If, on the other hand, Rome was an unwelcome concept, it is more likely that we would recognize traditional cultures being expressed, revived, and even reinvented. It is clear, then, that identity can be a changing concept and one that can be manipulated by individuals throughout life and in response to different circumstances. What emerges from our understanding of identity is that multiculturalism may not be as easy to see in the archaeological record as we may hope and is not as simple as identifying well-​defined cultural or ethnic groups. If we are to seek to understand different cultures and identities through material culture, it is important to keep in mind these fluid concepts of identity, the projection of self-​identity, and how that is embedded in materiality.

Materiality of Multiculturalism What, then, of other groups and cultural identities along Hadrian’s Wall and how might we see identity in the materiality of the Wall zone? Lazzari (2005: 127) has argued that materiality is a dialogue between people and objects that entails continual reflection, and there is a cyclical relationship between the way people make things and the way things make people. Miller suggests that ‘the importance of an engagement with the concept of materiality for the archaeologist is that it helps us to grant to those we study a perspicacity equivalent to that which we seek for ourselves’ (Miller 2005: 213). In terms of how we understand multiculturalism, materiality is crucial, as it is often the lens through which we see people as ‘other’. The difficulty in using materiality to understand the past is that materiality ‘is not reducible to a set of given conditions or practices common to all cultures and all times, it is a set of cultural relationships’ (Meskell 2005: 6). In attempting to disentangle the meaning in the materiality, we are trying to interrogate specific moments of creation, exchange, use, and discarding of objects (Meskell 2005). Meskell (2005) maintains that we conceive and construct the material world ourselves; within this world we can be ‘constrained or triggered by objects and features, consciously or unconsciously. They may be produced or appropriated with specific intentions and yet

234

234   Claire Nesbitt influence future actions in an unpredictable way’ (Fahlander and Oestigaard 2008: 4). In this way, material objects have the potential to be active in the sense of stimulating or determining social action (Fahlander and Oestigaard 2008:  4). Through studying objects that belonged to, were created by, or were buried with individuals, then, we can begin to understand the creation and projection of social identities (Dellino-​Musgrave 2005: 220). The case studies below will focus on two different aspects of identity: the identity that is lived through material culture during life and the identity that is projected by material culture at death.

Materiality of Life The materiality of the Wall has suggested a diversity of groups were present there, represented by military units that included Frisians, Batavians, Dacians, Pannonians, Raetians, and Thracians, to name a few (see Breeze 2006). One heavily contested element of the multicultural nature of the Wall zone is the presence of communities and individuals from North Africa. In ad 208, the emperor Septimius Severus arrived in Britain to respond to troubles on the northern frontier. He was accompanied by a military contingent from outside the province of Britannia, including Legio VI Victrix—​a legion that almost certainly included soldiers from North Africa (Swan and Monaghan 1993: 26). The presence of North African individuals has also been claimed on the basis of changes in material culture in parts of the province. Most notable is the appearance, in York, of a new style of ceramic ware, known as ‘Late Ebor Ware’, manufactured in York from local fabric but similar in style to pottery manufactured in North African provinces (Africa Proconsularis, Numidia, and Mauretania: Swan 1992: 6). The vessels, largely representing cookware, represent a departure from more common Romano-​ British cooking vessels, having uncharacteristic rounded bottoms and lids. Swan has argued that the manufacture, in significant quantities, of North African-​style vessels suggests that there were sufficient people who knew how to use them and desired to use them (Swan 1999; 2009: 54); indeed she has gone so far as to see these as ceramics made for Africans by Africans, probably soldiers who were garrisoned in the province. Elsewhere, the presence of African-​style cooking vessels at Segedunum, Wallsend, has also led to speculation that a unit of Moors may have been stationed there in the third century ad (Breeze 2006: 134). The implications of such material may be more complex, however. Fulford (2010: 77) has emphasized the need to be cautious in directly associating particular forms of pottery with particular ethnic groups; he argues that, while Ebor ware has close parallels in Tunisian ceramic assemblages, the style was one commonly seen around the Mediterranean. Its dissemination need not be seen as coming directly from Tunisia and, even if we accept the presence of African potters, it does not follow that we can assume an ethnic customer base (Fulford 2010: 75–7​7). The possible presence in Britain of Africans is also indicated by other evidence: a unit of North African soldiers (the Numerus Maurorum Aurelianorum) is attested to

Multiculturalism on Hadrian’s Wall    235 at Burgh-​by-​Sands on Hadrian’s Wall in ad 253–​8 (RIB 2042), and a tombstone recording a North African freedman, Victor (RIB 1064), has also been discovered on the Wall (see section on Expressions of Individual Identity). Further evidence of a North African influence, this time in building techniques, may be apparent in structures along the Wall, with tubuli lingulati, ceramic tubing used in roof vaulting, discovered at Pardshaw, Cumbria. This technique of roofing is common in North African building and is known in Britain only from the baths at Caerleon and Chester and a sewer at York (Mason 1990: 221). These three sites are all known to have undergone refurbishments during the Severan period, when the North African influence is believed to have arrived in Britain with the emperor, and all are military in origin. How far these trappings of North African material culture actually represent the presence of North African individuals has been a matter of debate (Swan 1992; Fulford 2010), although it is notable that these appeared at the same time as historical records tell us military units associated with North Africa arrived in Britain. We may be seeing North Africans units preferring the cookware or architectural forms that they were familiar with. Alternatively, if we view armies as ‘major engines of integration’ forging ‘Romanness’ from ethnic diversity (James 2011: 28) there is evidence that military units borrowed from a range of ethnic styles and created their own military culture, which cannot be equated with a single geographical, ethnic, or cultural origin. In their static role as material culture, they may be simply a different style of cooking pot, but, as active materiality, these ceramics could represent part of a dialogue of expressed identity for the people who owned and used them. It is, however, problematic to equate specific cultural packages with ethnicity. Rather than simply reflecting ‘North African’ origin, the use of new cultural packages may reflect these external groups’ need to express new identities in contrast to existing ones, perhaps marking their particular military unit or even groups within those units defining their social status and group identity. Swan’s identification, therefore, of what seem to be North African or Mediterranean eating and cooking habits emphasizes the complex set of identities (and indeed contemporary multiple identities) being expressed even within a small area of northern Britain; these individuals may have not seen themselves as African or even come from North Africa, but food ways from that region had become integral to their newly formed identity. Whether or not we can associate certain styles of pottery or construction techniques with particular ethnic groups, these new forms of dining and building methods emphasize the complex identities being formed in northern Britain, particularly within the units on Hadrian’s Wall. The diverse nature of the identities and cultural antecedents of many of the military units in the Wall zone is also illustrated by a further example of ceramic evidence. Ceramics known as ‘Housesteads ware’ were originally identified by Jobey (1979) as probably being associated with Frisian units from north of the Rhine. Van Driel-​Murray (2009: 818–​819), however, notes that examples of the pottery do not reflect the typical Frisian domestic assemblage but, rather, are finer forms, which may reflect continued use of strong Germanic ritual traditions centred around hospitality and guest-​friendship. This could suggest that Housesteads ware represents not simply a community of

236

236   Claire Nesbitt immigrant Frisians manufacturing and using objects that are familiar and remind them of home, but also objects that have cultural symbolism that then become clear identity signifiers.

Materiality of Death In contrast to the accumulation of possessions and use of objects in life, which can be transient, expedient, and perhaps even mindless, the objects people are buried with represent how they saw themselves in life, or perhaps more accurately, how they or their loved ones wished them to be seen in death. Treatment of the body at death can, in some ways, be seen as ‘the ultimate expressions of humans’ perceptions of themselves in society and cosmos’ (Fahlander and Oestigaard 2008: 1). In many ways, therefore, the expression of identity can be seen more evidently in death and its memorialization than in life. This is evident in the burial choices of individuals and communities in the frontier zone. Many burials have been excavated along the Wall, but systematic excavation of cemeteries has been limited. Of those that have been excavated, only Brougham provides sufficient material culture to withstand broader analysis. Other cemeteries that have been examined, such as Beckfoot in Cumbria (Caruana 2004), comprising mainly cremations with limited grave goods, and the Botchergate cemetery in Carlisle (Breeze 2006), indicate little evidence of cultural identities. In 2009, excavations at the cemetery at Birdoswald were undertaken. Initial results from Birdoswald suggest that there is little evidence for cultural packages that may reflect particular provinces or regions (see Haynes, this volume). To date, only one cemetery in the frontier zone, Brougham, Cumbria (Cool 2004), provides us with evidence concerning the expression of identity through the materiality of death. The cemetery at Brougham was in use between c. ad 200 and 310, and there is evidence of two distinct burial grounds; Cool has suggested that the second cemetery could represent groups with a different set of identities establishing a new burial ground if they felt that existing ones did not allow them to bury their dead in their traditional or correct way (Cool 2004: 463). Many of the elements of the burial rite used in the cemetery have been argued as being Pannonian in style and as indicating the presence of foreign groups (Cool 2004: 464). Though this is not attested by epigraphy, it is thought that a new military unit that arrived at Brougham in the early third century ad (the Numerus Equitum Stratonicianorum) may be the people represented by this change in burial activity. To begin with it is known that, in Pannonia, as well as Noricum, Dalmatia, Upper Moesia, and Dacia, graves were purified using a ritual burning event; in the Brougham cemetery, dark staining in grave 227 may represent such a burning (Cool 2004: 465). Secondly, there are some finds in association with the burials at Brougham, including iron bucket pendants and cremated horses, that are unparalleled in funerary rites known elsewhere in Roman Britain. Cool (2004: 464) suggests that the bucket pendants imply trans-​Danubian connections, while the horse cremations may suggest contact with northern Germany. Other items of personal adornment discovered

Multiculturalism on Hadrian’s Wall    237 at Brougham include red-​striped beads, also found at Augst and Avenches in modern Switzerland. Cool (2004: 464) suggests that again these reflect a Danubian tradition and are often a feature of the Maslomecz-​Gruppe type culture. The only object from the cemetery that may offer us an individual identity is a jar that was included with a burial. The jar is inscribed with a graffito of the name ‘Bata’, which is Illyrian in origin. Cool (2004: 466) has argued that, if indeed the cemetery was that of the Numerus Equitum Stratonicianorum and their families, it is likely the unit came from the Pannonian area of the Danube and probably included cavalrymen from Barbaricum. The drawback of burial archaeology is, of course, that the dead do not bury themselves. They are buried by others, and there is potential for great difference between their view of the deceased and that of others (Fahlander and Oestigaard 2008: 11). While, on the one hand, we can suggest that material culture is used in life and in death to manipulate an individual’s persona (Cool 2010: 27), it is also true that the identity of an individual is not straightforward, and each identity is a complex collection of varying situated roles and identities that transect one another in time and space (Fahlander and Oestigaard 2008: 11). In many ways, epigraphic evidence, which was sometimes prepared by the deceased before death, can offer us a more reliable interpretation, perhaps not of who they were, but of whom they wanted us to believe they were.

Expressions of Individual Identity The above examples illustrate the multicultural nature of the Army, but to what extent can we determine the multicultural nature of individuals within the Wall zone? Examples of individual identity that survive are almost exclusively epigraphic and from tombstones. In Britain, the people most likely to record their life on a tombstone were probably immigrants; the epigraphic habit was not a native British tradition, and it was a trend that took off slowly among the civilian population, with most inscriptions representing the military and their families (Noy 2010: 18; see also Hope, this volume). Hope (1997: 257) has argued that tombstones were a means by which foreigners or outsiders could seek acceptance into a community. Derks has also argued ‘inscribed monuments erected by individual members of local communities provide unparalleled access to subjective feelings of belonging even at the level of the individual’ (Derks 2009: 240). The tombstones appear also to transcend social differences with evidence of tombstones for women (for example, that of Pussitta from Raetia, modern Switzerland (RIB 984)), children (for example, Salmanes, aged 15, commemorated on a tombstone from Auchendavy on the Antonine Wall (RIB 2182)) and former slaves such as Victor, a Moor found at South Shields (RIB 1064), whose short life was recorded by this inscription: D M VICTORIS NATIONE MAVRVM ANNORVM XX LIBERTVS NVMERIANI EQITIS ALA I  ASTVRVM QVI PIANTISSIME PROSEQVTVS EST [RIB 1064].

238

238   Claire Nesbitt To the spirits of the departed and Victor, of the Moorish nation, twenty years old, freedman of Numerianus, a trooper of the First Wing of Asturians, who most devotedly conducted [his burial].

It was a common practice in the Roman world to express identity away from one’s place of origin (Salway 1965). This even appears to have included those who came from elsewhere in the province, some of whom were serving in units with origins supposedly elsewhere in the empire—​for example, Nectovelius son of Vindex, serving in a Thracian cohort, is described as nationis Brigans (RIB 2142), or Verecunda, commemorated by her husband as cives Dobunna (RIB 621), and another woman from Ilkley remembered as civis Cornovia (RIB 639) (see also Ivleva, this volume). Burial was usually the duty of the deceased’s heir. A Roman tombstone fulfilled two functions: it commemorated the dead by simply recording the name, sometimes with his or her achievements, and it also stated in writing the commemorator’s discharge of his duty. It was the Roman way of indicating the fulfilment of a particularly Roman obligation (Meyer 1990: 78). In this sense, what may be reflected are the agendas of the bereaved, rather than the deceased (Cool 2004: 437). However, tombstone evidence is unique in its ability to reveal the individuals of the frontier landscape. Tombstones are a specific form of evidence, which inevitably mean that our understanding of the individual is led by how the individual wanted to be represented. There are one or two epigraphic examples that provide detailed evidence of the complex multiple identities individuals could have on the Wall. The first is the tombstone of Regina discovered at South Shields (Figure 11.2). The tombstone reads: D M REGINA LIBERTA ET CONIVGE BARATES PALMYRENVS NATIONE CATVALLAVNA AN XXX. (RIB 1065) To the spirits of the departed and Regina, freedwoman and wife of Barates of Palmyra, a Catuvellaunian by race, thirty years old. (Noy 2010: 18)

Below this, in Palmyran script, is the following text: ‘Regina freedwoman of Barate(s), alas!’ (Noy 2010: 18). This stone is regularly held up as an example of a foreign presence on the Wall, of intermarriage between native Briton and foreign immigrants, and the nature of ethnicity and race in the Wall region (Hingley 2010: 235; Noy 2010: 18; Pearce 2010: 87). But how much can we discern about multiculturalism on the frontier from the short inscription? The lady in question, Regina, is described as a freedwoman and wife, suggesting that she had been in captivity and been bought, perhaps by the husband, Barates. The inscription describes her as a 30-​year-​old Catuvellaunian—​the civitas associated with the area of modern Hertfordshire. The choice to include this regional heritage and the decision to self-​identify with this region may reflect a desire to associate with a provincial status conferred on this particular social group by the Roman administration, rather than to hark back to the longevity of an Iron Age identity (Moore 2011). Mattingly (2010: 218) has suggested that the inscription referring to her ethnic origin as Catuvellaunian would

Figure 11.2  Regina’s tombstone. Source: © Claire Nesbitt.

240

240   Claire Nesbitt be unusual in a civilian context but is more common in a military context, situating her within the practices and beliefs of the military community. As Cool (2010: 29) has argued, ‘being native in Roman Britain was no more straightforward than being Roman was from an ethnic point of view’. Her status as a freedwoman also reveals the fluidity of social standing in the Roman Empire. Whatever the motivation for the description of Regina, the tombstone represents a mixed-​race couple from very different parts of the world, unified in a frontier landscape through an empire that dissolved many socio-​cultural barriers. Equally interesting is Barates’ own heritage. A tombstone that may belong to the same Barates was discovered at Corbridge. This inscription reads: D(IS) M(ANIBUS) [BA]RATHES PALMORENVS VEXIL(I)A(RIUS) VIXIT AN(N)OS LXVIII [RIB 1171]. To the Sprits of the departed and Barathes the Palmyrene, flag bearer/​maker lived 68 years. (Noy 2010: 19)

It is uncertain whether this individual was associated with a particular unit, although the presence of the Numerus Barcariorum Tigrisiensium, a unit of Tigris bargemen, from the Middle East, at Arbeia, South Shields, may suggest he was part of this unit. That he chose to honour his wife’s memory in Palmyrene as well as Latin is revealing, marking a potentially strong statement of ethnic or cultural identity (Noy 2010: 22). Latin was the usual epigraphic language in Roman Britain, and there is nothing to suggest a large community of Palmyrans who would understand the inscription. Perhaps the subtext represents a tender desire to say farewell in his own language, in the Palmyran way. Despite the complexities of their origins, these inscriptions emphasize the multicultural identities, even within a married couple, not only of the Hadrian’s Wall community but of the multi-​layered and complex nature of individual identities within the broader Wall zone.

Conclusions Hadrian’s Wall has often been seen as a barrier between peoples and cultures, and a contested landscape. However, as Hingley (2010: 229) has suggested, the material culture tells a different story, a story of inclusivity. In fact it was a place where people of many different cultures came together, brought to the frontier by military service, slavery, trade, or family ties; almost every corner of the empire can be seen represented on the Wall. The interesting aspect of the presence of these people is not so much the culture they brought with them, but rather the culture they developed and adopted, the culture they shared and the ways in which that culture was expressed to form new identities. The military materiality of the Wall belies a greater, more

Multiculturalism on Hadrian’s Wall    241 cosmopolitan reality, of a place where peoples of a diverse range of ethnic, cultural, and religious backgrounds unified under a common Roman identity. The materiality of multiculturalism on the Wall is rich and will no doubt continue to unfold as further excavations happen and more discoveries are made. The archaeology reveals quite clearly the presence of the ‘other’, of peoples far away from their homes and perhaps determined to maintain a cultural identity. However, at the heart of multiculturalism in the frontier zone lies one unifying factor, that of being, or aspiring to be, Roman.

Abbreviation RIB  R  . G. Collingwood and R. P. Wright, The Roman Inscriptions of Britain Volume I: Inscriptions on Stone. Oxford: Clarendon Press, 1965–​to date.

References Adler, P. S. (1976). ‘Beyond Culture Identity’, in L. Samovar and R. Porter (eds), Intercultural Communication. Belmont, CA: Wadsworth Publishing, 362–​378. Anderson, A. W. (2011). ‘Traditions and Transitions: Later and Roman Iron Age Communities in the North-​East of England’. Unpublished Ph.D. thesis, Durham University. Breeze, D. J. (1982). The Northern Frontiers of Roman Britain. London: Batsford. Breeze, D. J. (1990). ‘The Impact of the Roman Army on the Native Peoples of North Britain’, in H. Vetters and M. Kandler (eds), Internationalen Limeskongresses 1986 in Carnuntum: Der römische Limes in Österreich, Heft 36. Berlin: Wein, 85–​97. Breeze, D. (2006). Handbook to the Roman Wall. Newcastle: Society of Antiquaries. Caruana, I. (2004). ‘The Cemetery at Beckfoot Roman Fort’, in R. J. A. Wilson and I. Caruana, (eds), Romans on the Solway: Essays in Honour of Richard Bellhouse. Kendal: Cumberland and Westmorland Antiquarian and Archaeological Society, 134–​173. Cool, H. E. M. (2004). The Cemetery at Brougham, Cumbria: Excavations 1966–​67. Britannia Monograph Series 21. London: Society for the Promotion of Roman Studies. Cool, H. E. M. (2010). ‘Finding the Foreigners’, in H. Eckardt (ed.), Roman Diasporas: Archaeological Approaches to Mobility and Diversity in the Roman Empire. JRA Supplementary Series 78. Portsmouth, RI: Journal of Roman Archaeology, 27–​44. Dellino-​Musgrave, V. (2005). ‘British Identities through Pottery in Praxis: The Case Study of a Royal Navy Ship in the South Atlantic’, Journal of Material Culture, 10: 219–​243. Derks, T. (2009). Ethnic Constructs in Antiquity:  The Role of Power and Tradition. Amsterdam: Amsterdam University Press. Diaz-​Andreu, M., and Lucy, S. (2005). ‘Introduction’, in M. Diaz-​Andreu, S. Lucy, S. Babić, and D. N. Edwards (eds), The Archaeology of Identity. London: Routledge, 1–​12. Divine, D. (1969). The North-​West Frontier of Rome:  A  Military Study of Hadrian’s Wall. London: MacDonald. Dobson, B., and Mann, J. C. (1973). ‘The Roman Army in Britain and Britons in the Roman Army’, Britannia, 4: 191–​205.

242

242   Claire Nesbitt Fahlander, F., and Oestigaard, T. (2008). The Materiality of Death: Bodies, Burials, Beliefs. BAR International Series 1768. Oxford: Archaeopress. Fulford, M. (2010). ‘Roman Britain: Immigration and Material Culture’, in H. Eckardt (ed.), Roman Diasporas: Archaeological Approaches to Mobility and Diversity in the Roman Empire. JRA Supplementary Series 78. Portsmouth, RI: Journal of Roman Archaeology, 67–​78. Gardner, A. (2002). ‘Social Identity and the Duality of Structure in Late Roman-​Period Britain’, Journal of Social Archaeology, 2/​3: 323–​351. Haverfield, F. (1912). The Romanization of Britain. Oxford: Clarendon Press. Haynes, I. (1999). ‘Military Service and Cultural Identity in the auxilia’, JRA Supplementary Series 34. Portsmouth, RI: Journal of Roman Archaeology, 165–​174. Hingley, R. (2010). ‘Tales of the Frontier: Diasporas on Hadrian’s Wall’, in H. Eckardt (ed.), Roman Diasporas: Archaeological Approaches to Mobility and Diversity in the Roman Empire. JRA Supplementary Series 78. Portsmouth, RI: Journal of Roman Archaeology, 227–​243. Hingley, R. (2012). Hadrian’s Wall: A Life. Oxford: Oxford University Press. Hingley, R., and Hartis, R. (2011). ‘Contextualizing Hadrian’s Wall:  The Wall as “Debatable lands”’, in O. Hekster and T. Kaizer (eds), Frontiers in the Roman World: Proceedings of the Ninth Workshop of the International Network Impact of Empire (Durham, 16–​19 April 2009). Leiden: Brill, 79–​96. Hodgson, N., McKelvey, J., and Muncaster, W. (2012). The Iron Age on the Northumberland Coastal Plain. Newcastle upon Tyne: TWM Archaeology and the Arbeia Society. Hope, V. M. (1997). ‘Words and Pictures: The Interpretation of Romano-​British Tombstones’, Britannia, 28: 245–​258. Hunter, F. (2001). ‘Roman and Native in Scotland:  New Approaches’, Journal of Roman Archaeology, 14: 289–​309. Hunter, F. (2007a). ‘Artefacts, Regions and Identities in the Northern British Iron Age’, in C. C. Haselgrove and T. Moore (eds), The Later Iron Age in Britain and Beyond. Oxford: Oxbow Books, 286–​296. Hunter, F. (2007b). Beyond the Edge of the Empire:​Caledonians, Picts and Romans. Ross-​shire: Groam House Museum. James, S. (2002). ‘Writing the Legions: The Development and Future of Roman Military Studies in Britain’, Archaeological Journal, 159: 1–​58. James, S. (2011). Rome and the Sword:  How Warriors and Weapons Shaped Roman History. London: Thames and Hudson. Jobey, I. (1979). ‘Housesteads Ware:  A  Frisian Tradition on Hadrian’s Wall’, Archaeologia Aeliana, 5 [ser. 7]: 127–​143. Lazzari, M. (2005). ‘The Texture of Things: Objects, People, and Landscape in North-​West Argentina (First Millennium ad)’, in L. Meskell (ed.), Archaeologies of Materiality. Oxford: Blackwell, 126–​161. Lucy, S. (2005). ‘Ethnic and Cultural Identities’, in M. Diaz-​Andreu, S. Lucy, S. Babić, and D. N. Edwards (eds), The Archaeology of Identity. London: Routledge, 86–​109. Mann, J. C. (1963). ‘The Role of the Frontier Zones in Army Recruitment’, in G. Novak (ed.), Quintus Congressus Internationalis Limitis Romani Studiosorum 1961. Acta et Dissertationes Archaeologicae, Arheološki Radovi i Rasprave 3. Zagreb: Jugoslavenska Akademija Znanosti i Umjetnosti, 145–​150. Mann, M. (1986). The Sources of Social Power, Volume I: A History of Power from the Beginning to ad 1760. Cambridge: Cambridge University Press.

Multiculturalism on Hadrian’s Wall    243 Mason, D. J.  P (1990). ‘The Use of Earthenware Tubes in Roman Vault Construction:  An Example from Chester’, Britannia, 21: 215–​222. Mattingly, D. J. (2010). Imperialism, Power and Identity. Princeton: Princeton University Press. Meskell, L. (2005). ‘Introduction:  Object Orientations’, in L. Meskell (ed.), Archaeologies of Materiality. Oxford: Blackwell, 1–​17. Meskell, L. (2007). ‘Archaeologies of Identity’, in T. Insoll (ed.), The Archaeology of Identities: A Reader. Abingdon: Routledge, 23–​43. Meyer, E. A. (1990). ‘Explaining the Epigraphic Habit in the Roman Empire: The Evidence of Epitaphs’, Journal of Roman Studies, 80: 74–​96. Miller, D. (2005). ‘Afterword’, in L. Meskell (ed.), Archaeologies of Materiality. Oxford: Blackwell, 212–​219. Millett, M. (1990). The Romanization of Britain:  An Essay in Archaeological Interpretation. Cambridge: Cambridge University Press. Modood, T. (2007). Multiculturalism. Cambridge: Polity Press. Moore T. (2011). ‘Detribalizing the Later Prehistoric Past: Concepts of Tribes in Iron Age and Roman Studies’, Journal of Social Archaeology, 11/​3: 334–​360. Nesbitt, C., and Tolia-​Kelly, D. P. (2009). ‘Hadrian’s Wall: Embodied Archaeologies of the Linear Monument’, Journal of Social Archaeology, 9/​3: 368–​390. Noy, D. (2010). ‘Epigraphic Evidence for Immigrants at Rome and in Roman Britain’, in H. Eckardt (ed.), Roman Diasporas: Archaeological Approaches to Mobility and Diversity in the Roman Empire. JRA Supplementary Series 78. Portsmouth, RI: Journal of Roman Archaeology, 14–​26. Pearce, J. (2010). ‘Burial, Identity and Migration in the Roman World’, in H. Eckardt (ed.), Roman Diasporas: Archaeological Approaches to Mobility and Diversity in the Roman Empire. JRA Supplementary Series 78. Portsmouth, RI: Journal of Roman Archaeology, 79–​98. Poulter, J. (2009). Surveying Roman Military Landscapes across Northern Britain: The Planning of Roman Dere Street, Hadrian’s Wall and the Vallum, and the Antonine Wall in Scotland. BAR British Series 492. Oxford: Archaeopress. Salway, P. (1965). The Frontier People of Roman Britain. Cambridge:  Cambridge University Press. Shennan, S. (1989). Archaeological Approaches to Cultural Identity. London: Unwin Hyman. Steinberg, S. R., and Kincheloe, J. L. (2009). ‘Smoke and Mirrors: More than One Way to be Diverse and Multicultural’, in S. R. Steinberg (ed.), Diversity and Multiculturalism: A Reader. New York: Peter Lang, 3–​22. Swan, V. G. (1992). ‘Legio VI and its Men: African Legionaries in Britain’, Journal of Roman Pottery Studies, 5: 1–​34. Swan, V. G. (1999). ‘The Twentieth Legion and the History of the Antonine Wall Reconsidered’, Proceedings of the Society of Antiquaries of Scotland, 129: 399–​480. Swan, V. G. (2009). Ethnicity, Conquest and Recruitment: Two Case Studies from the Northern Military Provinces. JRA Supplementary Series 72. Portsmouth, RI: Journal of Roman Archaeology. Swan, V. G., and Monaghan, J. (1993). ‘Head-​Pots: A North African Tradition in Roman York’, Yorkshire Archaeological Journal, 65: 21–​38. van Driel-​Murray, C. (2009). ‘Ethnic Recruitment and Military Mobility’, in Á. Morillo, N. Hanel, and E. Martín (eds), Limes XXth International Congress of Roman Frontier Studies. Madrid: Ediciones Polifemo, 813–​822.

244

244   Claire Nesbitt Witcher, R. E. (2000). ‘Globalisation and Roman Imperialism: Perspectives on Identities in Roman Italy’, in E. Herring and K. Lomas (eds), The Emergence of State Identities in Italy in the First Millennium BC. London:  University of London Accordia Research Institute, 213–​225. Witcher, R. (2010). ‘The Fabulous Tales of the Common People, Part 2’, Public Archaeology, 9/​ 4: 211–​238. Witcher, R., Tolia-​Kelly, D. P., and Hingley, R. (2010). ‘Archaeologies of Landscape: Excavating the Materialities of Hadrian’s Wall’, Journal of Material Culture, 15: 105. Woolf, G. (1997). ‘Beyond Romans and Natives’, World Archaeology, 28/​3: 339–​350. Woolf, G. (1998). Becoming Roman. Cambridge: Cambridge University Press.

Chapter 12

B ritons on t h e  Mov e Mobility of British-​Born Emigrants in the Roman Empire Tatiana Ivleva

Introduction: Identifying The Mobile Identities The movement of people and objects is attested for all periods of human history. For the Roman Empire the evidence for such movement is abundant, owing to the conquests of various territories, which resulted in a wide range of individuals and communities being on the move, with both voluntary and forced migration being common. Within Roman Britain itself the movement of immigrants from the continent has been the topic of a number of publications covering the origin of migrants, their distribution, and the ways one might identify them (cf. Thompson 1972; Rowland 1976; Birley 1988; Swan 1992; Wilmott 2001; Leach et al. 2009, 2010, to name but a few). This chapter seeks to go beyond the issues of internal migration within the Roman province of Britannia and avoids addressing the movement of foreigners within Britain. It focuses instead on movement away from Britain and discusses the presence of Britons elsewhere in the Roman Empire. The movement to a new territory influences the ways individual migrants or communities of settlers see themselves, (re)forming along the way the myriad of identities that already existed within both newcomer and host societies. The way one perceives the other in migrant groupings undergoes identity stress when new forms of identification are constructed, manipulated, or adjusted to circumstances (Oltean 2009: 92–​93). Consequently, mobility and the transformation of one’s identity go hand in hand, and discussion of the changes within the personal self cannot be avoided in studying individuals on the move.

246

246   Tatiana Ivleva The concept of ‘identity’ can usefully be considered in terms of two categories: universalization and duality. The first category is based on an Aristotelian approach, whereby identity is defined according to the principle that ‘a thing is itself ’ (Aristotle, Metaphysics VII. 17), emphasizing the universality and sameness in things. The second category allows identity to be considered in the same terms as the theoretical concept ‘duality of structure’ (Giddens 1979): individual identity—​selfhood—​is formed within the personal self and as an opposition to the perception of others. Here the emphasis is placed on the duality of the nature of ‘identity’, where ‘selfhood’ is perceived by the self and by others, allowing the understanding, formation, negotiation, fragmentation, fluctuation, and so on of the self. This is where identity, better understood in its plural form—​identities—​is categorized as fluid, dynamic, and unstable; it is constantly changing, depending on situations in which agents find themselves. While the first category has one level, ‘sameness’, the second category implies various levels of identification—​ that is, an individual or a self has many identities, based on gender, ethnicity and culture, age, status, class, and religion, opposing and contrasting them with definitions imposed by ‘the others’. One of the most studied levels of identity is ethnic identity. Ethnicity has usually been considered to be based on ‘racial’ characteristics: the same origin, language, or descent (cf. Brather (2004: 77–​88) on the notion of ‘race’ in discussions of ethnicity), but it has now been widely acknowledged that ethnicity is more ‘an idea than a thing’, based primarily on social relationships and similar ways of behaving, and is something that can be learnt, rather than something one is born into (Lucy 2005: 86). Any ethnic affiliations can be changed by agents through mobility or social associations; ethnicity is, therefore, highly mutable and dependent on the contexts in which agents find themselves (Brather 2004: 568; Lucy 2005: 97). As a result, ethnic identity is created and (re)invented within a variety of cultural repertoires embedded within, and formed by, social practices and formulated through dialectic opposition of the self and the other. Ethnic realization is born within particular groups at the moment when cultural differences are recognized, providing the motive for the universalization of these differences and making them into ‘practice’. Yet, there is a precondition that allows such groups to justify their communal closure and to find a common ground for the group’s formation: the unifying principle (that is, the ‘sameness’), embedded within the familiarity in the use of objects or interaction between agents known as habitus (Bourdieu 1998). In other words, recognition of the similarities that allows a group to form ‘an ethnicity’ derives from a social practice based on ‘shared ways of doing things’ (Lucy 2005: 101). Ethnic identity is, therefore, a product of understood differences and preconditioned sameness. Taking into consideration the explanation proposed here of the notion ‘ethnic identity’, this chapter seeks to outline, in general, the possible ways one can identify British emigrants on the continent and to understand, in particular, how ethnicity was formed and operated within the groups that moved away from Roman Britain. It attempts to chart the changes over time in terms of the personal and communal identification and to provide explanation for such changes. Furthermore, it explores the consequences that

Britons on the Move    247 movement might have had on the use, as well as on the changing roles, of objects made in Britain and brought over to the continent.

Multiplicity of Identities in  the Roman Empire: Being Roman and Being British One cannot approach the study of mobility in the Roman Empire without asking questions relating to issues of identity and ethnic identification. The existence of a multiplicity of identities in, and the multiculturalism of, the Roman Empire has been widely recognized; the majority of scholarship in this area has been devoted to showing the fragmentation of Roman identity and has perceived the Roman Empire as a heterogeneous society containing a variety of individual and group responses to ‘being Roman’ (Wallace-Hadrill 2007; Hingley 2009; Revell 2009). ‘Roman’ cannot be regarded as a fixed entity, since the different individuals and groups dwelling within the boundaries of the Roman Empire may have understood and experienced ‘being Roman’ in a variety of ways: ‘being Roman’ always meant something different (Revell 2009: p. xii). In a similar vein, in thinking about ‘being British’ we might suppose that, for individuals or groups coming from Roman Britain and settling on the continent, this notion was inverted. It has been proposed that within Roman Britain itself there were ‘no such social groups as “Britons”, the peoples were an assortment of tribes’ (Mattingly 2004: 10). This seems indeed to have been the case, especially when we take into consideration the epigraphic record of Roman Britain: a total of ten inscriptions have been recorded, dating roughly to late first–​late second centuries, which mention the origin of an individual from a particular British tribe or town. These ten individuals were interprovincial migrants and belonged to various British tribes: six were citizens of the Canti (RIB 192), Cornovi (RIB 639, female), Dobunni (RIB 621, female), Dumnonii (RIB 188), and Catuvellauni (RIB 1065, female) and the city of Lindum (RIB 250, female). Three indicated their origin (natione) as belonging to the Belgae (RIB 156), Briganti (RIB 2142), and Catuvellauni (RIB 1962) tribes, while one simply named his origin as (colonia) Victrix (RIB 3005). These inscribed stones were erected either by the relatives of individuals who had died somewhere other than in the territory of their tribe or by individuals who were fulfilling vows in a foreign region of their home province. All of them found it important to emphasize their origin—​an action that indicates both the significance of tribal above provincial forms of identification and a possible continuation of tribal divisions and differences in Britain under Roman rule. By trying to understand the refusal to denote provincial origin and the continuation in the use of tribal identification, one needs to explore the significance of the terms Britannus and Britto—​labels associated with the pan-​tribal community. It is likely that both terms were coined and artificially imposed by the new dominant power in Britain

248

248   Tatiana Ivleva after ad 43 and actually do not derive from any self-​awareness on the part of the indigenous population (Matthews 1999). The labels were probably imposed by the Roman administration for an administrative convenience or perhaps intended to speed up the process of inclusion of the natives into the Roman orbit: a process called ‘superficial homogenization’ (Matthews 1999: 29). Such homogenization is recorded in other communities who supplied recruits for the Roman army. The main purpose was the promotion of a special type of military identity—​a regional one. For instance, the Romans continuously cultivated tribal associations among the Batavians, a tribe from Germania Inferior, placing a particular emphasis on their militaristic nature (van Driel-​Murray 2003: 201; Roymans 2004: 223). The Romans might also have reinvented and manipulated British ethnic identity by consistently referring to the people who originated from, or were born in, the province of Britannia as ‘Britons’. For instance, Dio Cassius (62.4) puts the following phrase in Boudicca’s mouth prior to the major battle between Roman and British forces in ad 60/​61: ‘for I [Boudicca] consider you all my kinsmen inasmuch as you inhabit a single island and are called by one common name.’ This is, clearly, an example of Roman rhetoric and propaganda rather than an exhibition of pan-​tribal British identity; such notions of artificial ethnicity may not have had much relevance for the peoples of Britain. In this sense, Rome created a new ethnic unity among the fragmented groupings (cf. Hingley (2009), commenting on the formation of artificial Batavian ethnicity), but one may ask how successful this attempt was, taking into consideration the promotion of tribal affiliations discussed above.

Britons Abroad: Forms of Identification If tribal affiliation was emphasized when an individual moved to another tribal region, how did Britons settling elsewhere in the Roman Empire denote their origin? In a more general sense, do we have any examples of such Britons and how can we identify them? The theme of the presence of foreigners in the various provinces of the Roman Empire has usually been tackled from an epigraphic perspective (cf. Noy 2001; Wierschowski 2001; Kakoschke 2002, 2004; Oltean 2009). The epigraphic record is the obvious source of evidence to turn to here: funerary, votive, and other types of inscriptions can be seen to have played an important role in reflecting existing identities—​ social, cultural, and ethnic—​as well as having created new ones (Hope 2001). When left by emigrants, inscriptions can indicate the choices they made when stating their origin, the places they settled in, and their reasons for migration overseas. Together with inscriptions, military diplomas (Roman citizenship certificates issued to auxiliary soldiers who had completed twenty-​five years of military service) can be used to determine the ways in which Britons drafted into the Roman army indicated their origin, as well as provide us with information on their status within a unit, and their social

Britons on the Move    249 and family relations. While the epigraphic record provides us with glimpses of various aspects of an emigrant’s life such as their age, occupation, or ethnic origin, it can rarely be taken at face value; the information provided was often ‘cleaned up’. Inscriptions and diplomas were a medium for invented identities: what was included, and in what form, might have been determined by the circumstances and the desires of a client who made the choices regarding what was appropriate to communicate (Bodel 2001: 34). Another problem when dealing with inscriptions and military diplomas is dating: while some can easily be dated by means of specific references in the texts, others can be dated only approximately, on the basis of the known development of linguistic formulae (cf. Holder 1980). Another type of evidence that can be used to trace British emigrants is that of dress accessories, because of their regionality and their ability to serve as a medium for the expression and negotiation of a person’s various identities, not least their origin (cf. Swift 2000; Rothe 2009; see also Cool, this volume). The significance of brooches, the most common and regionally specific dress accessory, as identity-​markers and their double functionality (that is, being passive, functional tools, they also acted as active participants in constructing the identities of the wearer) have been considered elsewhere (cf. Jundi and Hill 1998; Pudney 2011). Brooches were personal items used to secure clothing, and, while crossing the Channel, emigrants from Britain most likely wore them or had them with them among their personal belongings. Moreover, British-​made brooches were distinctive in their design, decoration, and form compared to local products in other parts of the empire, which makes them stand out within the homogenous material culture of continental sites, as I have argued elsewhere (Ivleva 2011: 133). British-​made brooches found on the continent have a similar dating problem as the epigraphic record: the precise date range when brooches were in use will always be uncertain (Snape 1993: 6). Yet the contexts where these brooches were located can provide a relative time span when particular types were in use (cf. dating of British-​made brooches in Bayley and Butcher 2004; see also Snape 1993; Mackreth 2011). One needs, however, also to take into account that brooches that are found in a context dating to a period when their popularity was on the wane may represent ‘heirlooms’. This problem is undoubtedly significant when one discusses a migrant population, considering that a family on the move may have curated brooches over a long period of time as a reminder of home ties (Revell, pers. comm; cf. also Gilchrist 2013). Moreover, these artefacts also have limitations regarding how representative they are of the population and present specific problems. For example, an object without context does not allow any conclusions concerning a person’s religious belief, status, or age. Personal accessories are particularly valuable as sources to study the projection and negotiation of personal identities—​not just ethnicity—​but such actions depend on the circumstances when a particular object was worn, something that is not visible in archaeological record. A total of 40 persons of British descent have been identified through the epigraphic record yet only 27 mention the individual’s origins directly—​which is extremely low in comparison to other ethnic groups such as Dacians (150 cases: Oltean 2009: 96) or Germans (174 cases: Kakoschke 2004: 198). Such low numbers are not representative

250

250   Tatiana Ivleva of the real level of mobility: not everyone was able to commission a funerary or votive monument or received a military diploma. A second factor might be the irrelevance of naming individual origins for soldiers serving among their own countrymen (Oltean 2009: 91): for a ‘Briton’ in a British auxiliary unit, it would have been unnecessary specifically to name his origin, whereas if he had served in another ethnic unit he would most likely have wanted to emphasize his ethnic background. A total of 242 British-​made brooches have been recorded from 102 sites across the empire—​the majority being found in the western part of the Roman Empire, in the militarized areas of Germania Inferior and Superior and on civilian sites in Gallia Belgica; the provenance of 19 brooches is unknown (Figure 12.1). That these brooches were actually ‘made in Britain’—​as opposed to the brooches reproduced by local craftspeople from templates—​can be supported by the fact that they occur in too small a quantity on the continent. While brooches with typical British characteristics appear to be relatively numerous in Britain, overseas they are found in limited numbers: 1 or at most 3 per cent of the total number found on any given site.

Figure 12.1  Distribution of British brooches. Source: Brooches’ distribution partly after Morris (2010: 86, figure 4.35 and appendix 6); map by author. © Tatiana Ivleva.

Britons on the Move    251

Identities in Words: Epigraphic Narrative A total of just twelve British soldiers have been identified who had served in British auxiliary and numeri units, which constitutes a small minority of all recorded soldiers who are known to have served in these troops (Ivleva 2012). Of these twelve British soldiers, the origin of five is recorded: a unit’s prefect and also legionary soldier from Lindum (Lincoln) (AE 1973: 459 and CIL III. 6679); a Dobunnian infantryman (CIL XVI. 49; Kennedy (1977) argued that he was Dobunnian through his mother, but Mullen and Russell (2009) show that his name is well attested in British epigraphic record); two footsoldiers from Ratae Corieltauvorum (Leicester), and the Belgae tribe (CIL XVI. 160 and AE 1944: 58 respectively), and, lastly, another infantryman who claimed to be Britto (RMD I. 47). The origin of the other seven has been identified through (a) linguistic analysis of their names (CIL V. 7717: Catavignus (his name has the suffix–​ign common in Insular Celtic; cf. Evans (1967: 209) and Sims-​Williams (2004: 155, n. 921); CIL III. 3256: Virssuccius and Bodiccius (for discussion, see Birley 1980: 103)); (b) the recruitment period (AE 1999: 1258 and CIL III. 10331: two ignotii, recruited at the same time and to the same unit as foot-​soldiers who originated from the Belgae tribe and Ratae Corieltauvorum); (c) the recorded recruitment pattern (AE 1994: 1487: possibly a son or a grandson of a British soldier who had followed his father or grandfather into military service (that is, following the pattern of hereditary military service whereby recruitment was from among the sons of veterans who had settled in the proximity of a fort) (cf. Dobson and Mann 1973: 202)); (d) the find-​spot of the military diploma (AE 2005: 954, found in southern Britain and which probably records a British veteran returning from Pannonia (cf. Tully 2005)). Those who were born in Britain were also selected to fill gaps in the legionary and auxiliary units stationed in the province and abroad. A variety of evidence comes from different parts of the empire and records the existence of at least seventeen men who emphasized their origin from Britain: five legionaries (CIL III. 11233: origin recorded as Claudia Camulodunum (Colchester); CIL VI. 3594: origin recorded as cognomen Britto; CIL VIII. 21669: from Lindum (Lincoln); CIL VI. 3346: originated from Glevum (Gloucester); CIL VIII. 2877: considered to be of British descent owing to his service in five British legions (cf. Malone 2006: 117)); three troopers in the Imperial horse guard in Rome (CIL VI. 3279, 3301, and 32861: the origin of all three is recorded as natione Britto/​ Britan(n)icianus); one centurion in a British detachment in Mauretania Tingitana (AE 1920: 47 and 48: the linguistic analysis of his name shows his possible British origin (cf. Raybould and Sims-​Williams 2009: 22)); eight auxiliaries in various cohorts and in fleets (AE 1951: 47: recorded as ex Br(e)itonibus; AE 2003: 1218: Trinovantian by origin; CIL III. 14214: origin recorded as Britto; CIL XIII. 8314: recorded as civi Brittoni; AE 1956: 249: citizen of Dumnonii; ILJug 02, 679: origin recorded as natione Britto; RMM 20: recorded as Britto; AE 2007: 1772: from the Cornovi tribe). The origin of two legionary soldiers has

252

252   Tatiana Ivleva been identified based on their religious beliefs: they made dedications to British Mother Goddesses while on service in legio XXX Ulpia in Xanten (CIL XIII. 8631 and 8632; for discussion see Ivleva 2011: 139). The number of civilians of British descent known to have settled abroad is low in comparison with the number of British servicemen: only seven are known—​one trader (AE 1922: 116: priest of the Imperial cult at the coloniae at Eboracum (York) and Lindum (Lincoln)); five whose occupation is unknown (CIL XIII. 1981 (a male): origin recorded as natione Britto; CIL XIII. 6221 and AE 1915: 70: both from Deva (Chester); AE 1939: 53 (a male): described as Britannus natione; Martial 11.53 (a female): ‘born amongst the woad-​stained Britons’; and one freedman in Rome, CIL VI. 2464: recorded as ‘taken from Brittannia’ (sic)). In total, thirty-​seven men and one woman have been identified, although two more females can be added to this list: Catonia Baudia (CIL VI. 3594) and Lollia Bodicca (CIL VIII. 2877), ‘travelling’ wives of the legionary soldiers who followed their partners to their posts within the empire. Both have quite a remarkable cognomen, one that resembles the name of British rebel Queen Boudicca. Considering that their husbands were of British descent, it is plausible that these women hailed from one of the British tribes. These forty people born in Britain had various professions, although the majority served in the Roman army: in the legions and auxiliary units posted overseas, in the fleet garrisoned on the continent, and in Rome as the emperor’s bodyguards (Figure 12.2). Quite surprisingly, only one British trader has been detected epigraphically, although there must have been British-​born indigenous traders (as opposed to British-​born immigrant traders) involved in cross-​Channel trade. It is unlikely that all trading activities between

17

8

4

3

3 2

Legionaries Auxiliaries

Sailors

Imperial horse guard

1

1

Centurion

Trader

1 Civilians (profession unknown)

Freedman

Women

Figure 12.2  Britons abroad: profession and status. Source: © Tatiana Ivleva.

Britons on the Move    253 Britain and the continent lay in the hands of people born on the continent, as the epigraphic record would seem to suggest (Hassall 1978: 43). Those who were born in Britain and later moved to the continent were not necessarily of native British stock: at least three legionary soldiers might have been sons or grandsons of immigrants to Britain in the mid and late first century ad—​namely, Titus Statius Vitalis (CIL III. 11233), who hailed from Colchester, and Marcus Minicius Marcellinus and Marcus Iunius Capito (CIL III. 6679; CIL VIII. 21669), who both came from Lincoln. Both Colchester and Lincoln were colonies for retired legionary veterans, and these men were probably descendants of legionary veterans who had settled in Britain upon their retirement (Birley 1980: 104–​105). One legionary soldier, Marcus Ulpius Quintus from Gloucester (CIL VI. 3346), might have been a son or grandson of an auxiliary veteran who had either come from the continent or been drafted from a British tribe to serve in Britain (Dobson and Mann 1973: 203; Birley 1980: 105). The geographic spread of inscriptions mentioning British emigrants is not confined to a particular province: they are distributed across the whole Roman Empire from North Africa to Germania Inferior and from Gallia to the Roman frontiers on the Danube (Figure 12.3). While the presence of some Britons in particular territories was due to the orders of Roman officials, others seem to have settled in particular places in a search of a better life. The example of the latter is Atianus, recorded as natione Britto, who settled in Lyon (CIL XIII. 1981). Neither his profession or the reason for his presence in Lyon

Figure  12.3 Distribution of the military diplomas (star), funerary (circle), and votive (diamond-​shape) inscriptions mentioning British emigrants. Source: Map by author. © Tatiana Ivleva.

254

254   Tatiana Ivleva is recorded on his funeral monument, but Lyon was a hub for commercial activity and attracted wealthy merchants and craftsmen, so he might have gone there to open a warehouse selling British goods or to help in establishing trading contacts between the two provinces (for the existence of such contacts, see Morris 2010). There is also evidence for female migration, with three women being identified: Catonia Baudia and Lollia Bodicca on the basis of their names and ‘British’ husbands (see above), and Claudia Rufina, who was referred to as ‘being born among the woad-​stained Britons’ by her friend, the poet Martial (11.53). It was relatively common for women—​whether wives, partners, or sisters—​to follow their military husbands, partners, or brothers to their postings (Allason-​Jones 1999:  48). These three British women, therefore, conform to this picture and should not be seen as an exception. All three enjoyed a privileged status: they were wives of legionary centurions. Moreover, Catonia Baudia and Lollia Bodicca are recorded on the funerary inscriptions that they themselves erected to commemorate their husbands, another hint as to their status and wealth. Claudia Rufina is praised by Martial (4.13, 8.60, and 11.53) on numerous occasions as an educated woman as well as adoptee of a Roman way of life. The epigraphic material also shows a considerable degree of variation in the nomenclature of origin, which varied from naming a tribe (e.g. Dobunni, Belgae, Cornovi, Trinovantes, Dumnonii) or specific place (e.g. Lindum, Ratae Corieltauvorum, Claudia Camulodunum, Glevum, Deva) to the formula natione Britto/​Britan(n)icianus. I have argued elsewhere (Ivleva 2011: 142–​144) that when these inscriptions and diplomas are divided by century a pattern seems to emerge in terms of the changing way in which origin is referred to. Inscriptions dated to the late first century ad usually record the name of the tribal and city origin, with an emphasis on the individual’s citizen status, which might signify the importance of indicating that one was Roman (that is, having citizenship) and at the same time belonging to a specific British tribe. Inscriptions dated to the second century ad show a difference in the choices made when referring to origin: while some individuals continued to name as their place of origin either a British city or a tribe as their place of origin (that is, thereby emphasizing their tribal affiliation), others preferred to identify themselves through geographical provenance as natione Britto/​Britan(n)icianus on inscriptions or as Britto on military diplomas. Such changes relate to a wider shift in forms of identification detected in the recording of affiliations in the second century ad, where the tribal affiliations were being ‘replaced by formulae using geographical provenance or political-​administrative inscription in a certain civitas’ (Derks 2009: 269). While this process appears to be relatively common, one may pose the question: what could have prompted some inhabitants of Britain settling on the continent to choose the geographical over the tribal provenance? As has already been mentioned, there were no such inter-​provincial grouping as ‘Britons’—​the term is likely to have been a Roman construct designed to denote all the inhabitants of the newly acquired province in ad 43 without paying attention to inner tribal divisions. By choosing this Roman-​imposed label, Britons abroad may have been expressing this new form of Roman-​imposed identity and constructed ethnicity. This invented label might have been used by second-​generation emigrants—​those who were not born in Britain but

Britons on the Move    255 whose parents belonged to one of the British tribes—​because they did not have a precise ethnic identification: being born at particular place on the continent would not necessarily have made them a member of a continental tribal entity (Ivleva 2011: 142–​143). The use of the Roman-​imposed identification, therefore, became a necessity: by choosing to refer to one’s origin as British, one distinguished oneself from other groups of migrants or from the dominant group in the territory where British migrants and their families settled down. The epigraphic record left by migrants in the third century ad and later indicates that the tendency for designating origins then shifted the other way: emigrants preferred to name their province instead of their tribe or city. This situation may have resulted from being incorporated into a new identity group in the third century ad as a result of everybody being given Roman citizenship by the edict of Caracalla in ad 212. This broke tribal ties, and the supra-​regional identity suppressed the regional one, resulting in the ultimate ‘e pluribus unum’ when, from a variety of tribes, one ‘province’ of emigrants emerged.

Archaeological Narrative of Mobile Identities: British-​made Objects and Britons Abroad The presence of 242 British-​made brooches on the continent can be related to a variety of activities of the people who brought these objects with them. Brooches ‘travelled’ because of their function as clothes fasteners: after all, the people (whatever their origin) who travelled from Britain to the continent needed something to hold their clothes together. This makes brooches useful tools in determining the places where such migrants settled down, be they traders, military men (veterans or soldiers), the followers of the first two (households, slaves, partners, wives, and children) or craftspeople. Because brooches were brought overseas by various groups, emigrants from and immigrants to Britain alike, this leads to the consideration that British brooches do not provide evidence for the ethnicity of their users and wearers: other identities—​such as status, gender, and, to a lesser extent, ethnicity—​might have projected through their use (see also, Cool, this volume). Moreover, brooches may have changed meaning depending on their usage, on the context in which they were worn or discarded, and on their viewers or admirers. Owners may also have had particular associations with them, possibly treating brooches as tourist knick-​knacks rather than objects of personal use. In general, it is possible to associate particular sites with the presence of British-​born emigrants through in-​depth analysis based on an object’s biography, site location, the history of a settlement, epigraphic analysis, and the study of the context in which an object has been found (Ivleva 2011: 137–1​42). It is likely that most British-​made brooches were taken overseas by British recruits of British auxiliary units, or by recruits of other

256

256   Tatiana Ivleva legionary and auxiliary forces of different ethnic origin who served in Britain, or by veterans who, after being discharged, returned home from Britain (Ivleva 2011: 142). Similarly, as the epigraphic record indicates that women were also on the move (often as the partners of British recruits or the partners of returning veterans of different ethnic origins), it is also probable that some of the British brooch types were taken to the continent on the clothes of such women (for how brooches were worn by Romano-​British women, see Croom (2004); but see also Allason-​Jones (1995) for discussion of the ‘sexless’ nature of brooches). This conclusion is in many ways similar to that derived from the epigraphic analysis of the presence of British-​born individuals, where the majority had a military connection. The variety of contexts in which British-​made brooches appear on the continent reflects the diversity of their meanings and the associations that emanated through their usage. This is an indication of the mobile identities of the brooches themselves. The analysis of the sites where brooches were found in burials has shown that they were probably brought by veterans returning from Britain to their own homelands (Ivleva 2011: 145). Within Britain, brooches were found in both inhumation and cremation burials, albeit in small numbers, and usually performed a double role: they were placed both for their functionality (that is, they were used to fasten a piece of cloth containing the remains of the deceased or to fasten a piece of clothing covering the deceased body) and for their associations with the dead person (that is, they were placed in a wooden box or a cloth or leather bag positioned next to the cremated remains) (Philpott 1991). As for the continental examples—​where all burials were cremations—​some brooches were positioned on top of the remains, which suggests that they were also used as cloth fasteners, while some appear to have been placed as votive offerings. Notably, most of the brooches had their pins intact (slightly corroded but still with the spiral attached), suggesting that they were deposited not as broken objects of no further use but as functional items intended to secure pieces of clothing. The brooches’ functionality was, therefore, an important factor; yet, one may ask why these particular brooches were put into graves: that is, why the relatives of the deceased chose British brooches to follow their beloved ones into the afterlife. The deliberate inclusion of brooches suggests that they had important connotations for the deceased whose remains they were supposed to secure as well as for the relatives, whose choice of a particular brooch may have been a conscious act. While the evidence indicates that brooches were rare as grave goods in Roman Britain, it does not mean that brooches were not placed with the bodies of the deceased. Rather their absence as intact objects may indicate that they were placed together with the body of the deceased and consequently were completely burned and, therefore, did not survive to enter the archaeological record. However, for the users of British-​made brooches on the continent, these objects may have had other associations: made in Britain, brought across the Channel to the continent because of their functionality, not destroyed but kept intact, they could have been used by other members of a family or community because of their limited availability, exoticness, and uniqueness. Yet they officially ended their lives being buried and being a protector of a dead individual’s remains. Therefore, it may not have been

Britons on the Move    257 their precious looks or their functional value for the living, but their particular associations with the deceased that were important. Brooches in continental burials are confined to areas where there is evidence for the presence of veterans having returned from Britain. Brooches, therefore, could have been valued by their owners and, later, by the relatives of the deceased for their associations with the past, indicating the (dead) owner’s experience in Britain. If we think first about British-​made brooches as the embodiment of a ‘British’ past and second about the care taken to avoid them being destroyed and the deliberateness with which they were placed in burials, it could follow that their inclusion in graves was a manifestation of memory relating to the deceased’s connection with Britain, either as a soldier who had served in Britain or as a Briton (be he or she male or female) who had died in a foreign land. A similar conclusion can be proposed for the placement of British-​made brooches in sanctuaries. Analysis has shown that these brooches too were brought by families of returning veterans or by veterans themselves. While they wore British-​made brooches in the setting of foreign cultures, the owners projected their past as people who had lived in Britain. When they deliberately refused to use the brooches any more as clothes fasteners, the projection of a foreign past and the experience in a foreign land was brought to an end. In this sense, brooches were subject to a twofold action: as personal offerings to gods and as closures of past activities—​when such personal items were given away to the gods, the past was symbolically buried and vows were fulfilled. Notably, the majority of brooches recorded as rubbish and accidental losses were found on sites where there is evidence of the stationing of troops coming from Britain. The rarity of British-​made objects on the continent did not influence the decision of some brooch owners intentionally to deposit these objects in rubbish pits, which meant the functional death of the object. Such an action had consequences for the projection of any form of identity, be it gender, ethnic, or cultural; the intentional death of an object stands for the death of meaning with which this item is associated, following up on the death of identities desired or wished for or (un)intentionally projected. However, such actions might have been influenced by the ready availability of brooches on the site, which would have allowed owners to continue to transmit whatever identities they wished and to serve as a reminder of an ethnic origin. One should take into account that the presence of objects made in Britain on continental sites with a homogenous material culture would have allowed them to stand out in the material record of that site. The realization that a brooch was different and exotic might have provided the grounds for the emergence of new meanings, possibly not existing in Britain itself, as is evident from the utilization of British-​made brooches as embodiments of the past among the veterans returning from Britain. Within British emigrant groups, the realization of brooches’ uniqueness could have reinforced the sense of being different, leading to the realization of belonging to another culture, of being of different ethnic stock. Ethnicity, therefore, becomes a by-​product of the relationship between the owner and the object: the particularity of the artefact might enhance the expressions of ethnic identity. The variety of treatment of British-​made brooches suggests that they were valued by migrants for particular reasons and played an important part in the processes of

258

258   Tatiana Ivleva remembrance and evocation of the past. That the idea of the past played a role when brooches were put in specific contexts abroad indicates a desire to forget, to reinvent, to evoke, or to project the past; it also emphasizes the value of memory within the groups travelling from Britain.

Conclusions Although the evidence is limited, it makes it possible to pinpoint the location of few Britons abroad. It is clear that both the past and memory of the land of their birth were important to these Britons, although it should be emphasized that for any moved individuals the past and homeland are important: an increase in the demonstration of one’s origin is particularly noticeable in moved communities (Oltean 2009: 94–​95). Britons were no different from any other migrants, and some were rather keen to make their ethnic origin explicit through written language, whether the decision lay in naming their tribe or in employing the adopted Roman construct Britonnes. For most Britons wearing a British-​made brooch abroad would have been a necessary and obvious thing to do, since it would have been brought among their personal possessions. Whether wearing a British brooch would reinforce the sense of ‘being’ from Roman Britain is a difficult issue, since a variety of other identities and messages could have been projected as well, yet the objects’ uniqueness and distinct style might have provided ground for the growing realization of ‘being’ different. At the beginning of this chapter I have stated that ethnic realization is born within particular groups at the moment when cultural differences are recognized, providing the motive for the universalization of these differences and making them into ‘practice’. In both epigraphic and archaeological evidence, we see that universalization of ethnic consciousness in a community living abroad might have taken place, because groups could have realized their uniqueness through the use of different objects and through exploitation of the imposed and invented label. In this sense, ‘being British’ abroad becomes more of an invented identity, something that can be evoked and reinforced through the use of brooches (and other British-​made objects) and when naming an origin.

Abbreviations AE   L’Année Épigraphique CIL     Corpus Inscriptionum Latinarum ILJug 02  A  . Šašel, and J. Šašel (1978). Inscriptiones latinae quae in Iugoslavia inter annos MCMLX et MCMLXX repertae et editae sunt. Ljubljana: Narodni Muzej. RIB      R  . G. Collingwood and R. P. Wright, Roman Inscriptions of Britain. Oxford: Clarendon Press, 1965–​to date.

Britons on the Move    259 RMD   M. M. Roxan, Roman Military Diplomas. London: Institute of Archaeology, 1978–​to date. RMM     B. Pferdehirt, Römische Militärdiplome und Entlassungsurkunden in der Sammlung des Römisch-​Germanischen Zentralmuseums. Mainz:  Römisch-​ Germanisches Zentralmuseum, 2004.

References Primary Sources Aristotle, The Metaphysics, trans. H. Lawson-​Tancred. London: Penguin, 1998. Dio Cassius, Roman History, trans. E. Cary. Bury St Edmunds: St Edmundsbury Press, 1989. Martial,. Epigrams, trans. D. R.  Schackleton Bailey. Bury St Edmunds:  St Edmundsbury Press, 1994.

Secondary Sources Allason-​ Jones, L. (1995). ‘Sexing “Small Finds”’, in P. Rush (ed.), Theoretical Roman Archaeology: Second Conference Proceedings. Aldershot: Avebury, 22–​32. Allason-​Jones, L. (1999). ‘Women and the Roman Army in Britain’, in A. Goldsworthy and I. Haynes (eds), The Roman Army as a Community:  Including Papers of a Conference held at Birkbeck College, University of London on 11–​ 12 January, 1997. Journal of Roman Archaeology Supplementary Series No 34. Portsmouth, RI:  Journal of Roman Archaeology, 41–​51. Bayley, J., and Butcher, S. (2004). Roman Brooches in Britain:  A  Technological and Typological Study based on the Richborough Collection. London:  The Society of Antiquaries in London. Birley, A. (1980). The People of Roman Britain. Berkeley and Los Angeles:  University of California Press. Birley, E. (1988). ‘Pannonians in Roman Britain’, Zeitschrift für Papyrologie und Epigraphik, 73, 151–​155. Bodel, J. (2001). ‘Epigraphy and the Ancient Historian’, in J. Bodel (ed.), Epigraphic Evidence: Ancient History from Inscriptions. London: Routledge, 1–​57. Bourdieu, P. (1998). Practical Reason: On the Theory of Action. Cambridge: Polity Press. Brather, S. (2004). Ethnische Interpretationen in der frühgeschichtlichen Archäologie: Geschichte, Grundlage und Alternativen. Berlin: De Gruyter. Croom, A. (2004). ‘Personal Ornament’, in M. Todd (ed.), A Companion to Roman Britain. Oxford: Blackwell, 288–​298. Derks, T. (2009). ‘Ethnic Identity in the Roman Frontier: The Epigraphy of Batavi and Other Lower Rhine tribes’, in T. Derks and N. Roymans (eds), Ethnic Constructs in Antiquity: The Role of Power and Tradition. Amsterdam: Amsterdam University Press, 239–​283. Dobson, B., and Mann, J. C. (1973). ‘The Roman Army in Britain and Britons in the Roman Army’, Britannia, 4: 191–​205. Evans, D. E. (1967). Gaulish Personal Names: A Study of Some Continental Celtic Formations. Oxford: Clarendon Press.

260

260   Tatiana Ivleva Giddens, A. (1979). Central Problems in Social Theory: Action, Structure and Contradiction in Social Analysis. Houndmills: Macmillan. Gilchrist, R. (2013). ‘The Materiality of Medieval Heirlooms:  From Biographical to Sacred Objects’, in H. P. Hahn and H. Weiss (eds), Mobility, Meaning and the Transformations of Things. Oxford: Oxbow Books, 170–​183. Hassall, M. (1978). ‘Britain and the Rhine Provinces: Epigraphic Evidence for Roman Trade’, in J. du Plat Taylor and H. Cleere (eds), Roman Shipping and Trade: Britain and the Rhine Provinces. London: The Council for British Archaeology, 41–​48. Hingley, R. (2009). ‘Cultural Diversity and Unity: Empire and Rome’, in S. Hales and T. Hodos (eds), Material Culture and Social Identities in the Ancient World. New York: Cambridge University Press, 54–​75. Holder, P. A. (1980). Studies in the Auxilia of the Roman Army from Augustus to Trajan. BAR International Series 70. Oxford: Archaeopress. Hope, V. (2001). Constructing Identity: The Roman Funerary Monuments of Aquileia, Mainz and Nimes. BAR International Series 960. Oxford: Archaeopress. Ivleva, T. (2011). ‘British Emigrants in the Roman Empire: Complexities and Symbols of Ethnic Identities’, in D. Mladenović and B. Russell (eds), TRAC 2010: Proceedings of the Twentieth Theoretical Roman Archaeology Conference, Oxford 2010. Oxford: Oxbow Books, 132–​153. Ivleva, T. (2012). ‘British Military Units and the Identity of British-​Born Recruits in the Roman Army, between the First and Third Centuries ad’, Orbis Terrarum: Internationale Zeitschrift für historische Geographie der alten Welt 10 [2008–​2011], 59–​92. Jundi, S., and Hill, J. D. (1998). ‘Brooches and Identities in First Century ad Britain: More than Meets the Eye?’ in C. Forcey et al. (eds), TRAC 97: Proceedings of the Seventh Annual Theoretical Roman Archaeology Conference Nottingham 1997. Oxford: Oxbow Books, 125–​137. Kakoschke, A. (2002). Ortsfremde in den römischen Provinzen Germania inferior und Germania superior: Eine Untersuchung zur Mobilität in den germanischen Provinzen anhand der Inschriften des 1. bis 3. Jahrhunderts n. Chr. Möhnesee: Bibliopolis. Kakoschke, A. (2004). ‘Germanen’ in der Fremde:  Eine Untersuchung zur Mobilität aus den römischen Provinzen Germania inferior und Germania superior anhand der Inschriften des 1. bis 3. Jahrhunderts n. Chr. Möhnesee: Bibliopolis. Kennedy, D. (1977). ‘The ala I and cohors I Britannica’, Britannia, 8: 249–​255. Leach, S., Lewis, M., Chenery, C., Müldner, G., and Eckardt, H. (2009). ‘Migration and Diversity in Roman Britain:  A  Multidisciplinary Approach to the Identification of Immigrants in Roman York, England’, American Journal of Physical Anthropology, 140: 546–​561. Leach, S., Eckardt, H., Chenery, C., Müldner, G., and Lewis, M. (2010). ‘A Lady of York: Migration, Ethnicity and Identity in Roman Britain’, Antiquity, 84: 131–​145. Lucy, S. (2005). ‘Ethnic and Cultural Identities’, in M. Diaz-​Andreu and S. Lucy (eds), The Archaeology of Identity:  Approaches to Gender, Age, Status, Ethnicity and Religion. London: Routledge, 86–​109. Mackreth, D. F. (2011). Brooches in Late Iron Age and Roman Britain. Oxford: Oxbow Books. Malone, S. J. (2006). Legio XX Valeria Victrix: Prosopography, Archaeology and History. BAR International Series 1491. Oxford: Archaeopress. Matthews, K. J. (1999). ‘Britannus/​Britto:  Roman Ethnographies, Native Identities, Labels, and Folk Devils’, in A. Leslie (ed.), Theoretical Roman Archaeology and Architecture: Third Conference Proceedings. Glasgow: Cruithne, 14–​33. Mattingly, D. (2004). ‘Being Roman: Expressing Identity in a Provincial Setting’, Journal of Roman Archaeology, 17: 5–​25.

Britons on the Move    261 Morris, F. (2010). North Sea and Channel Connectivity during the Late Iron Age and Roman Period (175/​150 BC–​AD 430). BAR International Series 2157. Oxford: Archaeopress. Mullen, A., and Russell, P. (2009). Celtic Personal Names of Roman Britain (accessed 3 June 2014). Noy, D. (2001). Foreigners at Rome: Citizens and Strangers. London: Duckworth. Oltean, I. A. (2009). ‘Dacian Ethnic Identity and the Roman Army’, in W. S. Hanson (ed.), The Army and Frontiers of Rome:  Papers offered to David J.  Breeze on the Occasion of his Sixty-​Fifth Birthday and his Retirement from Historic Scotland. Journal of Roman Archaeology Supplementary Series 74. Portsmouth, RI: Journal of Roman Archaeology, 90–​103. Philpott, R. (1991). Burial Practices in Roman Britain:  A  Survey of Grave Treatment and Furnishing, AD 43–​410. BAR British Series 219. Oxford: Archaeopress. Pudney, C. (2011). ‘Pinning down Identity:  The Negotiation of Personhood and the Materialization of Identity in the Late Iron Age and Early Roman Severn Estuary’, in D. Mladenović and B. Russell (eds), TRAC 2010: Proceedings of the Twentieth Annual Theoretical Roman Archaeology Conference, Oxford 2010. Oxford: Oxbow Books, 115–​131. Raybould, M. E., and Sims-​ Williams, P. (2009). Introduction and Supplement to the Corpus of Latin Inscriptions of the Roman Empire Containing Celtic Personal Names. Aberystwyth: CMCS. Revell, L. (2009). Roman Imperialism and Local Identities. Cambridge: Cambridge University Press. Rothe, U. (2009). Dress and Cultural Identity in the Rhine–​Moselle Region of the Roman Empire. BAR International Series 2038. Oxford: Archaeopress. Rowland, R. J. (1976). ‘Foreigners in Roman Britain’, Acta Archaeologica Academiae Scientiarum Hungaricae, 28: 443–​447. Roymans, N. (2004). Ethnic Identity and Imperial Power: The Batavians in the Early Roman Empire. Amsterdam: Amsterdam University Press. Sims-​Williams, P. (2004). The Celtic Inscriptions of Britain: Phonology and Chronology, c. 400–​ 1200. Oxford: Blackwell. Snape, M. (1993). Roman Brooches from North Britain:  A  Classification and a Catalogue of Brooches from Sites on the Stanegate. BAR British Series 235. Oxford: Archaeopress. Swan, V. (1992). ‘Legio VI and its Men: African Legionaries in Britain’, Journal of Roman Pottery Studies, 5: 1–​33. Swift, E. (2000). Regionality in Dress Accessories in the Late Roman West. Montagnac: Monique Mergoil. Thompson, L. A. (1972). ‘Africans in Northern Britain’, Museum Africum, 1: 28–​38. Tully, G. D. (2005). ‘A Fragment of a Military Diploma for Pannonia found in Northern England?’, Britannia, 36: 375–​382. van Driel-​Murray, C. (2003). ‘Ethnic Soldiers: The Experience of the Lower Rhine Tribes’, in T. Grünewald and S. Seibel (eds), Kontinuität und Diskontinuität: Germania Inferior am Beginn und am Ende der römischen Herrschaft. Berlin: De Gruyter, 200–​217. Wallace-​Hadrill, A. (2007). ‘The Creation and Expression of Identity: The Roman World’, in S. Alcock and R. Osborne (eds), Classical Archaeology. Oxford: Blackwell, 355–​380. Wierschowski, L. (2001). Fremde in Gallien, ‘Gallier’ in der Fremde:  Die epigraphisch bezeugte Mobilität in, von und nach Gallien vom 1. bis 3. Jh. n. Chr. (Texte, Übersetzungen, Kommentare). Stuttgart: Steiner. Wilmott, T. (2001). ‘Cohors I Aelia Dacorum: A Dacian Unit on Hadrian’s Wall’, Acta Musei Napocensis, 38/​1: 103–​123.

262

Chapter 13

Britain , G au l , and Germa ny Cultural Interactions Tom Moore

Introduction The changes that took place in Britain as a result of the influence of Rome, both before and after the conquest, cannot be seen in isolation. Britain’s place on trade routes, its position as a key focus for the Roman military, and its economic interaction with its nearest neighbours meant it had close relationships with its adjacent provinces in Gaul and Germany. Many of these links, particularly between southern Britain and Gaul, stemmed from earlier interaction in the late Iron Age; indeed, many of the changes that took place at this time underpinned the cultural changes that have in the past been described as ‘Romanization’. With widespread recognition that the term ‘Romanization’ poses a problematic idea of ‘Roman’ versus native culture, inadequately explaining the nature of change in Roman Britain, it has been suggested by some that interactions with Gaul and Germany were more significant than those with the Mediterranean in defining cultural transformations in Roman Britain. Some have gone so far as to suggest processes of ‘Gallicization’ in the first century ad (e.g. King 1999: 189), with the changes at the end of Roman Britain part of a corresponding ‘Germanization’ (Mattingly 2006: 224). Focusing on key changes in diet, architecture, and burial rites, this discussion aims to assess the nature and extent of cultural interactions between these provinces. In particular, it will be asked whether terms such as Gallicization are helpful in reconceptualizing the processes of cultural change before and after the Roman conquest, or whether these too present their own problems in describing the changes that took place.

Britain, Gaul, and Germany: Cultural Interactions    263

Cross-​C hannel Interactions before the Roman Conquest To assess the significance of interactions between Roman Britain, and Gaul and Germany, we need first to assess the context and importance of these interactions prior to Britain’s inclusion in the empire. Interaction between Britain and the near continent became increasingly pronounced from the second century bc onwards with the appearance of Gallo-​Belgic pottery and Dressel 1 wine amphorae, first in central southern Britain—​most notably at Hengistbury Head, Hampshire—​and later in south-​ east England. This interaction coincided with a range of cultural changes in southern Britain. One of the most significant of these was the appearance of new burial rites in the early first century bc, with a move away from excarnation and inhumation, to cremation burial in southern and eastern England (Figure 13.1). These burial rites range from the earlier, relatively simple cremations with urn and occasional grave goods, as at Westhampnett, Hampshire (Fitzpatrick 1997: 201), to more elaborate ‘Welwyn-​type’ burials with a range of imported grave goods from the Mediterranean and Gaul, including wine amphorae and metal drinking paraphernalia. These burials were normally situated in rectangular enclosures varying in size from a few metres to c. 50 metres across.

km

100

Ri

ve

rS

ev

er

n

0

River Thames

Figure 13.1  Distribution of cremation burials in late Iron Age Britain. Source: after Fitzpatrick (1997); © T. Moore.

264

264   Tom Moore Some were extremely elaborate, including wooden burial chambers (Folly Lane, Herts.), tumuli (Lexden, Essex), and burials situated in large enclosures (Stanway, Essex). Despite the presence of some high-​status imports from the Mediterranean, the form and content of these burials emphasize a connection to the near continent rather than influence directly from Rome. Similarly enclosed cremation burials are found in areas of modern-​day Luxembourg and the Champagne region of France (Crummy et al. 2007: 453), while similar burials and ceramic forms to those at Westhampnett also occur in Normandy (Fitzpatrick 1997: 208). Some can be seen to be remarkably similar in form and content; for example, those at Welwyn and Folly Lane in Hertfordshire and at Clemency and Goblange-​Nospelt in Luxembourg (Figure 13.2; Metzler et al. 1991; Niblett 1999: 397). The objects in these burials also emphasize continental links; the ‘warrior burial’ from Stanway, for example, has brooches most closely paralleled in central eastern France (Crummy et al. 2007: 446), while a mid-​first-​century bc inhumation burial from North Bersted, Sussex, has a helmet type normally found only on the continent (Taylor et al. 2014). Even some of the Mediterranean imports in such graves may have come not directly from Roman traders, but indirectly as prestige goods from elites in Gaul (Millett 1990: 38). Changes in burial corresponded with other changing attitudes to the body, including a marked increase in the use of toilet instruments (ear scoops, nail cleaners, and tweezers) which also had their origins on the continent (Hill 1997; Eckardt and Crummy 2008). To explain this phenomenon, from an early stage archaeologists noted Caesar’s claim (Gallic War 5.12) that people from northern Gaul had previously migrated to southern Britain. Allied to the archaeological evidence of the appearance of cremation burial, archaeologists in the cultural–​historical tradition saw this as supporting evidence for a large-​scale migration of people from the continent prior to Caesar’s invasion (Hawkes and Dunning 1931; Cunliffe 1988: 147). The similarity in names of the social entities (or ‘tribes’) in central southern Britain and on the near continent, the Atrebates, in modern-​ day Hampshire and north-​eastern France and the Catuvellauni of Essex and Cataluni of north-​eastern France (Crummy et al. 2007: 452), has also been used as evidence of movement of peoples. More recently, most archaeologists would argue that many of the changes witnessed in the late Iron Age took place around the time of, or after, Caesar’s invasion and that, while burial rites changed, many other aspects of society remained similar to those that had gone before, suggesting little evidence for a large-​scale movement of people. Although we may not be witnessing mass migration prior to the Roman conquest, there were clearly close relationships between communities in southern Britain and northern France. Classical sources provide some evidence of the complex inter-​relations that existed across the Channel. Julius Caesar in his account of the Gallic War observes, for example, that Diviciacus, king of the Suessiones, held land on both sides of the Channel (Gallic War 2.4). It has also been suggested that Commius, king of the Gallic Atrebates, a one-​time ally of Caesar, might have been installed in central southern Britain to control the Atrebates there, as may have been the case with Tasciovanus and the Catuvellauni (Creighton 2000: 64; 2006: 22). The appearance of coins bearing the name ‘Commios’ in central southern Britain in the mid-​first century bc might

(a)

(b)

Figure  13.2  Comparison of late Iron Age cremation burials from (a) Welwyn Garden City, Hertfordshire, and (b) Clemency, Luxembourg. Source: © British Museum and Musée National d’Histoire et d’Art, Luxembourg.

266

266   Tom Moore corroborate this (Creighton 2000). These elite relationships before Rome’s final submission of Britain in the mid-​first century ad were part of a much wider interplay between Rome and its neighbours in the manipulation of allies and client kings—​one in which Gaul and southern Britain were inextricably linked (Creighton 2000, 2006). Evidence for these complex political relationships is potentially present in the coinage that appears in southern Britain from the second century bc. Gallo-​Belgic type coins, which emerged on the continent imitating Greek coinage, were adopted and then imitated in Britain from the early first century bc. These gold coins were unlikely to be for everyday transactions, and their high value, alongside similarities of types across the Channel, may indicate that they represented connections between elites both within and beyond Britain (Cunliffe 2012: 17) or the movement of mercenaries between Britain and Gaul (James 2001: 191). The distribution of imported coinage types in some areas, such as the Isle of Wight (Wellington 2001), indicates that some communities were involved in a range of connections along and across the Channel. The adoption of coinage was only part of the broader changes in society. This included the emergence of so-​called oppida, such as Verulamium and Camulodonum—​large dyke complexes encompassing vast areas of landscape, largely lacking in dense settlement but often displaying evidence of elite burial, ritual activity, and craft production (Moore 2012). The distinct differences between oppida in Britain and Gaul perhaps indicates that, despite the political and cultural interactions evident above, there were significant differences in social organization and power structures between the two areas. Only one such site shares significant similarities with those on the continent. At Silchester, Hampshire, an organized pre-​conquest street layout suggests relatively intense occupation in a defined enclosure more similar to the planned, nucleated settlements in northern France—​such as Villeneuve-​Saint-​Germain in Picardy (Haselgrove 1996; Fulford and Timby 2000: 549)—​than the sprawling dyke systems in Britain. At Silchester it has even been suggested that the nature of the artefact assemblage, changes in dietary preferences (including increasing consumption of cattle and pig), and the planned layout indicate it was a planted settlement, with large numbers of settlers from northern Gaul (Fulford and Timby 2000: 563). However, while the presence of Gaulish immigrants and traders (for example, at Hengistbury: Fitzpatrick 2001: 89)—​particularly specialists or elites—​is certainly a possibility (Millett 1990: 34), there seems no reason to argue for wholesale migration or that it was only immigrants who were using new forms of material culture (cf. Fitzpatrick and Timby 2002: 171). Many aspects of ceramic use and treatment of the dead continued (as at other oppida such as Bagendon and Verulamium), reflecting existing practices, suggesting the majority of people were indigenous and that more complex social changes were underway than just immigration.

Explaining Interaction Before the Roman Conquest Earlier explanations for the relationship between southern Britain and the continent in the late Iron Age tended to emphasize the economic nature of these relations and

Britain, Gaul, and Germany: Cultural Interactions    267 the expansion of Rome in this process as fundamental. Evidence of Dressel 1 amphorae (imported first into central southern and then south-​eastern England), alongside Strabo’s description (Geography 5.2.2) of the exports from Britain, indicated to many that there was increased trade with the continent that was indirectly influenced by the expansion of Rome into Gaul (Cunliffe 1988: 103). With indigenous elites in south-​east England acting as ‘middlemen’, it was argued that this region became a core zone that exploited the peripheries of eastern and western England for commodities—​such as slaves, precious metals, and foodstuffs—​to supply Rome (Haselgrove 1982). More recently, the economic nature of this interaction has been questioned, with a recognition that the amount of imports arriving in Britain was relatively small and that the connections across the Channel were related to social and cultural interaction rather than trade (Hill 2007). Creighton (2000, 2006) has argued that the division between pre-​and post-​conquest Britain overlooks the ongoing interaction between elites in southern Britain and Rome from the time of Caesar’s invasion in the mid-​first century bc, suggesting that much of the material record—​such as the elaborate burials—​reflects the complex political and identity changes underway either side of the Channel. There is also increasing recognition that interaction across the Channel preceded the late Iron Age. For example, recognition that round-houses, often thought of as a peculiarly British architectural tradition, did exist in northern France implies that cultural interaction between communities on both sides of the coast had much longer antecedents (Lefort and Marcigny 2009; Cunliffe 2012: 16; Webley 2015). This indicates that the expansion of Rome was not necessarily the prime instigator of change: Britain was already participating in a range of cultural zones. The nature of interaction was also complex: burial practices across the Channel—​while often strikingly similar in particular instances—​obscure a variety of rites that existed, even in those areas where cremation was adopted (Hamilton 2007). Despite their similarities, these burials reflect a combination of both longer-​lived burial traditions and continental phenomena. We should not underestimate the longer-​term impact of internal social and economic changes (as a result of population increase, agricultural intensification, and the dynamics of existing social systems), which were also important in stimulating the change in social structures, forms of identity, and burial rites taking place at this time (Hill 2007). What we are seeing, therefore, is not simply the adoption of foreign traditions, but the creation of new practices. None of these can be put down to one particular source—​ external or internal, Rome or Gaul; rather, they are the result of a combination of elements adopted and manipulated by individuals and communities. The varied practices and grave goods in the individual burials at Stanway, for example, emphasize that individuals were reconceptualizing identities at a range of levels: gender, status, ethnicity. The use of toilet instruments reflects this process: a continental form of material culture was adopted in the late Iron Age, but then certain types continue to be used (often in different ways) in Roman Britain—​in contrast to the areas in which they originally emerged (Eckardt and Crummy 2008). The identities being expressed appear to be focused on elite connections and social networks, and there is no reason necessarily to see many of these individuals as

268

268   Tom Moore immigrants from northern Gaul. We should be wary of seeing the practices adopted as ethnic attributes, such as ‘Gallic’ (let alone ‘Roman’). Gaul and Germany were themselves artificial constructs, which only emerged from the works of Roman authors and, later on, of Imperial administrators, which often had little to do with the realities of pre-​Roman identities. Writers such as Julius Caesar, in his Gallic War, and Tacitus, in his Germania, often defined the habits and nature of these regions in opposition to each other, as civilized versus barbaric, for example (O’Gorman 1993; Wells 1999: 102). Such concepts, and the regions they portrayed, often have little relationship to the evidence from archaeology, which indicates the varied identities, settlement patterns, and languages of these communities, and they often show no clear division between east and west of the Rhine (Woolf 1998: 49; 2011; Wells 1999: 102). We might even question whether identity was expressed at the level of the civitas or ‘tribe’ (for example, Catuvellauni or Atrebates), when such labels were also largely a creation of the post-​ conquest era (Moore 2011). Individuals, like those at Stanway, therefore, may have been imitating other elite individuals as part of social connections and expressions of status, but this need not be seen as an expression of ethnicity. When discussing interaction between Britain, Gaul, and Germany, we must recognize the complex variety of identities and range of communities involved in these areas (both before and after the Roman conquest) and the dangers in defining attributes as reflecting ethnic groups—​many of which are likely to be constructs of our own and Roman writers’ imaginations.

Continental Influence in  Roman Britain Villa Architecture The close links between southern Britain and Roman Gaul continued after the conquest, at least between certain sectors of the elite. Perhaps one of the clearest signs of these interactions is in the architectural changes that took place in early Roman Britain. The row houses and winged corridor villas, which developed in the late first and early second century ad, are a shared style that occurs across Britain and northern Gaul. Many such early villas, constructed in the late first century ad, such as Ditches, Gloucestershire, and Gorhambury, Hertfordshire, developed along similar lines and relatively contemporaneously to examples in northern Gaul (Figure 13.3; Trow et al. 2009: 55, 70; Taylor 2012: 181). As with their continental counterparts, they also often emerged from existing late Iron Age settlements, potentially representing the continuity of those groups and individuals who had participated in connections across the Channel prior to the conquest. The origin of these building forms is uncertain, but they are clearly not a direct imitation of Italian architectural forms and appear to have developed in Gaul (Haselgrove 1995: 71). The most likely explanation for their adoption in Britain is a process similar

Herouville, Calvados

Mansfield Woodhouse, Notts.

10m Boos-Le Bois Flahault, Seine-Maritime

Gayton Thorpe, Norfolk

Gadebridge Park, Herts. Herouville, Calvados (late phase)

Behen, Somme

Ditches, Glos. C

C

Boos-Le Bois Flahault, Seine-Maritime (late phase) Shakenoak II, Oxon.

C

C

Figure 13.3  Similarities in villa layouts from northern Gaul and Britain. Source: © T. Moore.

270

270   Tom Moore to the one that took place before the conquest. They were not necessarily the buildings of immigrants from Gaul, rather, indigenous elites were attempting to imitate the trappings of their counterparts on the opposite side of the Channel (Trow et al. 2009: 70). The development of stone architecture in early Roman Britain was largely alien to indigenous building traditions, and it has been suggested that Gallic masons were imported for such constructions (Blagg 1984: 260); there is certainly tantalizing evidence to support this, with epigraphic evidence of at least one stonemason at Bath (RIB 151) who hailed from northern Gaul (Mattingly 2006: 299). Powerful individuals (and their architects) were likely to be far more familiar with the elite forms of architecture in northern Gaul than the peristyle houses of Italy; it is natural that it is this template they should imitate. At sites like Ditches, cultural connections may have been indirect through elites located around more politically connected towns, such as Verulamium; but such individuals were potentially well aware of the wider cultural and social connections into which they were embedded. By contrast, the owners of the palatial Italian-​style courtyard villa of the late first century ad at Fishbourne, Sussex (Perring 2002: 33), were clearly making more direct connections with the Roman imperial administration. The adoption of such villa designs was not just one of functional adaptation to a northern climate or imitation of Gallic examples. The layout of such villas also represented a mode of social display that emphasized power relations within and between communities, perhaps harking back to indigenous modes of social relations, as well as new ones (Smith 1997). The similarities in villa form may, therefore, reflect similarities in social relations that had developed concurrently in southern Britain and northern Gaul immediately before the Roman conquest. Despite these early similarities, however, villa architecture in Britain did not continue to emulate those on the near continent (Taylor 2012: 182), which emphasizes the way villa development existed within its own peculiarly British social and economic context, while also sitting within the broader context of household organization seen elsewhere in the Roman world (Perring 2002: 206).

Religion and Romano-​Celtic Temples Links between Britain and Gaul in the late Iron Age have also been suggested in ritual practices. This has largely been on the basis of Caesar’s assertion that the Druids of Gaul had their training in Britain (Gallic War 6.13). Whether there were close links in ritual and religious practices between these areas is open to debate (Webster 1999); in reality, the complex regional nature of the Iron Age archaeological record across north-​western Europe implies quite varied and localized practices. However, the appearance of shrine structures in late Iron Age Britain marked a significant change in religious practices—​ one that seems to have spread from the continent (see also Kamash, this volume). Dedicated shrines emerged in Britain in the first century bc, with the most well-​ understood example at Hayling Island, Hampshire, on the south coast of England. This consisted of a rectangular enclosure with circular inner structure, which saw the votive deposition of coinage and metalwork (King and Soffe 2001). In similar fashion to its

Britain, Gaul, and Germany: Cultural Interactions    271 counterparts of the continent, the Iron Age structure was replaced after the conquest by a Romano-​Celtic temple (these are sometimes referred to as Gallo-​Roman temples: e.g. Derks 1998). These temples were part of a common architectural form found across the north-​western Roman provinces: they were normally a rectangular (sometimes circular) cella often situated within a larger temenos (Figure 13.4). As they were restricted to Gaul, Germany, and Britain, it has been suggested that this form of temple represents a fusion between ‘Celtic’ and Roman ritual practices (King 2007). In Britain, there has been a attempt to draw the antecedents for these sanctuaries back to the Iron Age (e.g. King 2007), although the evidence for Iron Age shrines in Britain continues to rely on a handful of examples, many of which may not be shrines at all (Millett 1995). The idea

0

150 km

Figure 13.4  Approximate distribution of Romano-​Celtic temples. Source: after Derks (1998); © T. Moore.

272

272   Tom Moore that these structures represented a fusion of religious practices has also been critiqued, not least on the lack of evidence for an overarching ‘Celtic’ religion in these areas. Despite these problems, the reason behind the commonality of temple design requires explanation (King 2007: 13). The lack of evidence for antecedents in Britain suggests the inspiration for these came from elsewhere. The earliest securely dated and recognizable temple at Hayling Island, in its form and practices, implies it was inspired by the much longer tradition of rectangular sanctuary structures, which occurred in northern Gaul from the third century bc, at sites such as Gournay-​sur-​Aronde and Ribemont-​ sur-​Ancre (King and Soffe 2001: 121). The more widespread adoption of Romano-​Celtic temples after the conquest has been seen by some, therefore, as part of the adoption of what were perceived by communities in Britain as ‘Romanized’ practices from elsewhere, where they had already been hybridized in Gaul and Germany (Millett 1995: 98). The same may be true of another similarly restricted phenomenon: that of Jupiter columns—​tall, decorated columns surmounted with a figure riding a horse over a half-​human/​half-​snake creature and found predominantly in Germany, eastern Gaul, and, to some extent, in Britain. Wells (1999: 220) has seen the iconography depicted on these columns and their distribution as representing a fusion of indigenous and Roman beliefs transformed into a peculiarly north-​western imagery (see also Woolf 2001). We need to be wary, however, of seeing ritual practices either as imported by immigrant communities or as continuity from Iron Age practices. The varied and fluid nature of Iron Age ritual practices is well recognized and the complex adoption and adaptation of practices, despite overarching similarities, are increasingly recognized in the Roman provinces (e.g. Goldberg 2009). Late Iron Age ritual practice across Gaul and Britain was in a state of flux, and the appearance of aspects such as Romano-​Celtic temples was part of widespread changes. Many of the gods depicted, the religious processes, and the architectural forms—​such as the Hayling Island temple—​that developed as part of the so-​called Romano-​Celtic beliefs owe as much to post-conquest developments and importations from the continent as from indigenous belief systems (Webster 1995). Indeed, Derks (1998: 243) regards the Romano-​Celtic temple itself as a new architectural and ritual development, with many in Gaul, as in Britain, showing little evidence of continuity from the Iron Age. The adoption of the Romano-​Celtic temple in parts of Britain might be better seen, therefore, as part of a wider repertoire of ritual architecture that emerged in the north-​western Roman provinces as part of a dialogue between existing Iron Age practices and a variety of external forces, including the Roman state. As with the adoption of particular villa architecture, this particular form of temple may have been regarded in some areas as expressions of personal benefaction and engagement within wider religious spheres beyond the local.

Diet and FoodWays Food ways are a particularly sensitive way in which groups differentiate themselves and express new identities in colonial contexts, and it is through this arena that much of

Britain, Gaul, and Germany: Cultural Interactions    273 what has, in the past, been regarded as early ‘Romanization’ can be seen (Meadows 1994: 136). However, like much else in early Roman Britain, the source and inspiration for changing paraphernalia and food choices were not necessarily from the Mediterranean, but from Britain’s nearest neighbours. The importation of Gaulish ceramics began well before the Roman conquest, beginning with Amorican vessels to sites like Hengistbury Head in the early first century bc and then—​far more widely in the early first century ad—​imports of terra rubra and nigra finewares from northern France, followed by the import of terra sigilatta via the Rhine from central France. These Gallic ceramics had a far wider range of distribution, before and after the Roman conquest, than imports direct from Italy (Millett 1990: 33). These new forms of pottery included platters and cups that had not existed in the Iron Age ceramic repertoire. This signalled new forms of eating and drinking that were more personalized and indicate different foodstuffs being consumed. The adoption of new dining wares was matched by changes in consumption patterns—​at least on some high-​ status sites in southern Britain—​with greater quantities of cattle, pig, and fowl being eaten (for example, at Skeleton Green, Puckeridge (King 1988); Ditches (Reilly 2009); Silchester (Fulford and Timby 2000; Albarella 2007: 395)). Such consumption preferences were more similar to those in Gaul than the continued focus on sheep on most late Iron Age sites elsewhere in Britain (Albarella 2007). These changes have often been interpreted as largely reflecting new markets open to the emergent pottery industries in Gaul, who were now firmly ‘Romanized’ (Fitzpatrick and Timby 2002: 169). However, this rather underestimates the choices being made by indigenous communities as part of a broader adoption of new culinary habits by certain sectors of the community. Major changes in diet appear to have continued on a more widespread basis after the Roman conquest (see also Maltby and Van der Veen, this volume). King’s analysis (1984, 1999) of faunal assemblages in Britain indicates an increasing proportion of cattle being consumed, particularly on military and urban sites. Traditionally, this has been argued as a process of ‘Romanization’ led by the elite and then imitated more widely in society (King 1984: 193). This preference for beef in Britain corresponds with a similar pattern across the northern provinces—​one that contrasts sharply with consumption patterns in Italy and the Mediterranean. It also marks a major change from Iron Age norms of a predominantly sheep-​based diet, although the latter appears to have continued in more rural areas (King 1984; 1999: 189). The parallel of these dietary patterns with Gaul and Germany, rather than Italy, led King (1984: 198) to see this as a process of the ‘Gallicization’ or ‘Germanization’ of British diets, with the Roman army being the engine for change, as many of its units had been based in Gaul and Germany, or contained members who originated there. This dietary preference subsequently became regarded as ‘Romanized’ after the conquest of Britain and was adopted by those of high status and in the more ‘Romanized’ areas of the new province. As with all such patterns, there is significant diversity within and between provinces (King 1999: 190), but the variation in faunal assemblages emphasizes the changing nature of what it meant to be ‘Roman’ and indicates that the continued influences—​particularly on the upper echelons of society in Britain—​were often from military sources, and by extension from Gaul and Germany,

274

274   Tom Moore rather than from the Mediterranean. Vivien Swan (2009) has also suggested that particular cooking vessels, and their associated culinary habits, were imported into Britain in the first and second centuries ad via Gaulish influences, which themselves were a ‘Gallicized’ development of forms that had their origins further east. Like King (1984, 1999), Swan (2009: 51) sees the vector of this material culture and culinary change in Britain as Gauls serving in the military, who brought their culinary preferences with them. There are potential problems in regarding the dietary changes and culinary habits in early Roman Britain as a process of ‘Gallicization’. Such a term implies cultural imitation of ethnic identity: that diet or food ways were defined by (and reflected) ethnic identities, which were then adopted wholesale by different communities. As King’s study (1999) illustrates, changes in dietary practices varied within and between sectors of society, and changes in diet were more complex than the adoption of, or resistance to, a cultural package. In both late Iron Age and Roman Britain, certain serving wares and types of food were adopted but were often used in traditional ways, or foodstuffs were prepared in new forms (Meadows 1994: 137). Food ways and ceramic traditions were increasingly diverse across the Roman Empire and were themselves in a state of flux, through trade, adoption, and fusion (Fulford 2010). What studies by King (1999) and Swan (2009) illustrate more convincingly is that particular sectors in society could adopt elements of traditions from other parts of the empire and utilize them within new contexts, as part of new identities—​ be they elites, the military, or a host of other social groups.

Migration and Cultural Change Many of the changes regarded as ‘Romanizing’ in pre-​conquest Britain have been explained in the past by the migration of already Romanized peoples from northern Gaul, either on a large scale (Hawkes and Dunning 1931) or by smaller groups and individuals (Fulford and Timby 2000: 563; Crummy et al. 2007). The evidence for systematic migration prior to the conquest is largely lacking, but to what extent can some of the cultural changes in Britain—​once it was part of the Roman Empire—​be explained by migration from Gaul and Germany? It is increasingly recognized that the Roman Empire saw significant population mobility and that the urban population (at least) of Britain was relatively cosmopolitan (Eckardt 2010). It has been suggested that individuals from Gaul, because of the existing cultural and linguistic similarities with southern Britain, were in the vanguard of exploiting opportunities in the new province (Mattingly 2006: 109; Fulford 2010: 69). There is certainly evidence for the presence of individuals from Gaul, such as monuments to individuals from the Bituriges-​Cubi of central Gaul in York (RIB 678; Cool 2010) and from the Bellovaci of northern Gaul in London (RIB 3014; Mattingly 2006: 298), but assessing the number of immigrants is far harder. The army too was clearly a significant agent of cultural influences from Gaul and Germany, be it in the form of dress (Swift 2000), architecture (Creighton 2006: 91), or food ways (King

Britain, Gaul, and Germany: Cultural Interactions    275 1999; Swan 2009). However, this was not necessarily a simple process of the transfer of cultural signifiers with particular groups and units; many such elements had already been incorporated into more complex fusions of military identity (Fulford 2010). The influence of individuals cannot be underestimated, but migration as a cause of many of these changes was only one element of interaction with Gaul and Germany; influence was far more complex than through people bringing identities with them (see also Ivleva and Eckardt, this volume, for discussions on mobility in the empire).

Late Roman Britain As with the claimed ‘Gallicization’ of identities in late Iron Age and early Roman Britain, some have also perceived changes in dress in the late Roman western empire, largely influenced from origins in the Germanic world (Swift 2000: 119), as a corresponding ‘Germanization’ of existing practices (Mattingly 2006: 228). One of the strongest routes by which identities changed in the late Roman Empire was via the Roman army, with the use of peoples with German origin as mercenaries. Significant movements of groups of people—​particularly groups attached to the army but including women and children—​occurred in the late empire, but there were also earlier movements in smaller numbers. The cemetery at the fort at Brougham, for instance, with its unusual burial rites and brooch types, has been used as an argument for the presence of people with trans-​Danubian or German origins (Cool 2004). By the fourth century ad, a variety of evidence indicates both incoming groups and individuals, alongside changes in local forms of dress (see Cool, this volume, for a discussion of the role of dress and identity). At Lankhills, Winchester, some burials with particular types of bracelets in the late Roman cemetery have been used to suggest a Pannonian origin for some individuals, or at least the expression of foreign identities (Swift 2000: 75; Pearce 2010: 85). The process was not one way: the occasional presence of belt fittings of ‘British type’ in Germany emphasizes that, while such an item might have a symbolic meaning in one place, it could have different meanings in other contexts (see Eckardt and Crummy 2006: 94). Such interactions did not just happen in the late Roman period, however. We should not underestimate the array of influences that took place across the Roman period, many of which occurred before these specific Germanic influences in the late empire. For example, Bruhn (forthcoming) has argued that the adoption of glass bangles in northern Britain from the late first century ad could have been inspired by the particular preference for bangles in the lower Rhineland; the localized adoption of material culture through specific channels of interaction (for example, the Roman army) could then have been ‘reinvented’ in a new regional context and should not necessarily be regarded as an ‘ethnic’ trait. Many of these developments were as much about changes in the expression of military identity and the emergence of localized identities, as adopting notions of ‘German’ identities. The adoption of the cross-​bow brooch, for example, which appears to have

276

276   Tom Moore originated in Hungary, was a symbol of military or bureaucratic status rather than ethnic origin (Swift 2000). Notably, however, as before, such symbols were being adopted from the peripheries of the empire, particularly Germany (Swift 2000; Pearce 2010: 85) rather than from the Mediterranean. As with changing modes of dress and diet earlier, we must remain cautious in seeing this process as ‘Germanization’. Many of these changes in material culture, such as the supposedly Pannonian bracelets from burials at the cemetery of Lankhills, often argued as representing ethnic identity, might better be seen in terms of age or gender identities (Gowland 2007; Theuws 2009), while many of the rites expressed appear to relate to a range of different practices, rather than one definable cultural package (Pearce 2010: 90). Rather like earlier influences from Gaul, many of the new items of dress adopted from the German world became part of Romano-​British provincial dress, moving from association with the military and ‘foreignness’, to localized expressions of place and identity.

Conclusions: Romanization, Gallicization, or Germanization? The impact of interaction with Gaul and Germany has a long history in understanding the changes from the late Iron Age in to the Roman period. However, it is no longer regarded as a product of mass migration or just an economic relationship. As attitudes to the complex nature of identity change in Roman Britain have emerged, scholars such as Richard Reece (1988: 11) increasingly saw this as a process not of Romanization, but of people adopting a multitude of identities and practices from elsewhere in the empire: ‘becoming more Gaulish, more Rhinelandish, more Spanish’. While helpful in recognizing the varied external influences Britain was subject to from the first century bc onwards—​and moving away from seeing these changes as ‘Romanizing’—​defining these processes as ‘Gallicization’ is problematic. In using such terms we are in danger of once again reducing the process of cultural change in the Roman period to a simplistic process of acculturation: the transferring of one culture for another (James 2001: 191). Alternatively, and equally problematically, is to continue to see these changes as processes of ‘hybridization’, the fusing of two discreet cultures: indigenous Iron Age with either Roman or Gallic. Recognition by some that Gallic or German culture had already changed and been informed by its earlier interaction with Rome goes some way to accepting that these identities and influences were not, in themselves, reflective of discreet indigenous practices. In both Roman Britain and in the late empire, particular material culture and architectural forms—​while originating in areas such as Gaul or Germany—​could be perceived as part of a broader ‘Roman’ cultural experience or take on connotations of ‘inclusivity’ in an exotic or high-​status set of practices—​even if much of those practices might have been relatively alien to elites in Rome itself. However, we need to be cautious in regarding ‘Gallicization’ (or even ‘Gallo-​Belgicization’ (Creighton

Britain, Gaul, and Germany: Cultural Interactions    277 2000: 82) as potentially ‘perceived Romanization’ by people in Britain. In so doing we are in danger of effectively seeing aspects such as dietary change or the adoption of Romano-​Celtic temples as ‘Romanization by proxy’: forms of culture that were essentially ‘Gallic’ but perceived and engaged with as ‘Roman’ by their adopters. We have to be careful, therefore, in defining normative comparanda (such as villa or temple forms) and regarding these as ‘the same’ outside of their local context; while forms of architecture or types of grave goods may be similar between Britain and Gaul, the ways in which they were used in everyday practice may have been significantly different. The evidence discussed above suggests instead that cultural interactions and influences between Britain, Gaul, and Germany were not simply a process of adoption. Links—​when they existed—​seem to be through particular media by particular groups and often (as seen in the burial rites and dining practices of the immediate pre-​conquest period) adapted and utilized within an array of existing traditions and localized requirements. For some periods this was largely restricted to emergent elites who had close political (and perhaps economic) connections with Gaul, but not all elites engaged with this process. As Creighton (2006: 77) recognizes, the province of Britain (and even pre-conquest Britain) was ‘full of individuals with differing experiences’, based on backgrounds from different areas of the world and different social agendas, ensuring this was a dynamic set of relationships and not one done at the level of ethnic group. We should also not necessarily regard this as a one-​way process of elite adoption. It is perhaps better to regard (parts of) southern Britain and northern Gaul in what James (2001: 191) has described as a ‘zone of cultural convergence’ (or, rather, a range of zones), which existed both prior to, and after, Britain’s incorporation into the empire. Through cross-​pollination of cultural preferences and media for exchange and clientage, shared practices of burial and villa architecture existed, but they were more complex than simply the elite imitating their Gallic neighbours. Thus, while we might envisage certain individuals in Britain adopting aspects of identities and expressions of power from late Iron Age and early Roman Gaul, these were part of a range of identities being expressed and transfigured in a colonial encounter. Later, in the Roman province, the adoption of ritual practices and architectural forms that owed their origins to Gaul may have been understood and practised in quite different and locally oriented ways. There is a danger in seeing this as merely an elite process that then ‘trickled down’ through emulation to the rest of the community. The adoption of aspects of material culture or architectural forms—​while often processes of elite emulation—​were also influenced by other factors; for example, the discreet distribution of Romano-​Celtic temples in Britain reflecting localized social and ritual practices rather than more Gallicized practices. ‘Gallic’ and ‘British’ identities were far from static, and, in reality, it is questionable how much such identities meant to, or defined, inhabitants of these provinces. The varied burial rites, settlement patterns, and material culture of southern Britain and northern Gaul emphasize diverse and discrepant identities, cross cut by concepts of status, gender, and age, on both sides of the Channel (Mattingly 2004). It is increasingly recognized in Roman studies that identities—​including ethnicity—​are fluid and multifaceted (e.g. Eckardt 2012: 251). Concepts of Gallic and Germanic identities were

278

278   Tom Moore themselves constructs, and it is increasingly apparent that regional, so-​called ethnic identities within the broader sphere of ‘Germany’ were being (re)conceptualized in the context of Roman power and administration (Roymans 2004). Many identities, such as ‘Batavian’, may have meant far more to people who claimed them (and those external to them) than any notion of Germania; but such identities themselves also changed their meaning, significance, and expression between the first century bc and fourth century ad. The relationships between Gaul, Germany, and Britain and how cultural transformations were played out, is likely to have been equally complex. Interactions and influences from Gaul and Germany were just part—​if a significant and recurring one—​of the multitudes of relationships in which different individuals, groups, and regions in Britain were engaged from the first century bc to the fifth century ad and cannot be isolated from the multitude of indigenous and external influences on expressions of power, identity, ritual practices, and social structure (see also Esmonde Cleary, this volume).

Acknowledgements I am grateful to Richard Hingley, Rebecca Gowland, and Claire Nesbitt for their comments on earlier versions of this chapter and to staff at the Musée national d’histoire et d’art, Luxembourg, for their assistance in obtaining, and granting permission to use, the image of the tomb at Clemency.

Abbreviation RIB  R  . G. Collingwood and R. P. Wright, The Roman Inscriptions of Britain. Volume I: Inscriptions on Stone. Oxford: Clarendon Press, 1965–​to date.

References Primary Sources Caesar, The Gallic War, trans. C. Hammond. Oxford: Oxford University Press, 1996. Tacitus, The Agricola and The Germania, trans. H. Mattingly. London: Penguin, 1948. Strabo, Geography, trans. H. L. Jones. Cambridge, MA: Loeb, 1923.

Secondary Sources Albarella, U. (2007). ‘The End of the Sheep Age: People and Animals in the Later Iron Age’, in C. Haselgrove and T. Moore (eds), The Later Iron Age in Britain and Beyond. Oxford: Oxbow Books, 389–​402.

Britain, Gaul, and Germany: Cultural Interactions    279 Blagg, T. F.  C. (1984). ‘An Examination of the Connexions between Military and Civilian Architecture’, in T. F.  C. Blagg and A. C. King (eds), Military and Civilian in Roman Britain:  Cultural Relationships in a Frontier Province. BAR British Series 136. Oxford: Archaeopress, 249–​264. Bruhn, J. (forthcoming). ‘Glass Bangles in Northern Britain: The Regional Uptake of Hybrid Cultural Objects in the Roman Period’. Cool, H. (2004). The Roman Cemetery at Brougham, Cumbria: Excavation 1966–​67. Britannia Monograph Series 21. London: Society for the Promotion of Roman Studies. Cool, H. (2010). ‘Finding the Foreigners’, in H. Eckardt (ed.), Roman Diasporas: Archaeological Approaches to Mobility and Diversity in the Roman Empire. Journal of Roman Archaeology Supplementary Series 78. Portsmouth, RI: Journal of Roman Archaeology, 27–​44. Creighton, J. (2000). Coins and Power in Late Iron Age Britain. Cambridge:  Cambridge University Press. Creighton, J. (2006). Britannia: The Creation of a Roman Province. London: Routledge. Crummy, P., Benfield, P., Crummy, N., Rigby, V., and Shimmin, D. (2007). Stanway: An Élite Burial Site at Camulodunum. Britannia Monograph Series 24. London: Society for Promotion of Roman Studies. Cunliffe, B. (1988). Greeks, Romans and Barbarians: Spheres of Interaction. London: Batsford. Cunliffe, B. (2012). ‘Calleva in Context’, in M. Fulford (ed.), Silchester and the Study of Romano-​ British Urbanism. Journal of Roman Archaeology Supplementary Series 90. Portsmouth, RI: Journal of Roman Archaeology, 14–​22. Derks, T. (1998). Gods, Temples and Ritual Practices: The Transformation of Religious Ideas and Values in Roman Gaul. Amsterdam: Amsterdam University Press. Eckardt, H. (2010). ‘A Long Way from Home: Diaspora Communities in Roman Britain’, in H. Eckardt (ed.), Roman Diasporas: Archaeological Approaches to Mobility and Diversity in the Roman Empire. Journal of Roman Archaeology Supplementary Series 78. Portsmouth, RI: Journal of Roman Archaeology, 99–​130. Eckardt, H. (2012). ‘Foreigners and Locals in Calleva’, in M. Fulford (ed.), Silchester and the Study of Romano-​British Urbanism. Journal of Roman Archaeology Supplementary Series 90. Portsmouth, RI: Journal of Roman Archaeology, 247–​256. Eckardt, H., and Crummy, N. (2006). ‘“Roman” or “Native” Bodies in Britain: The Evidence of Late Roman Nail Cleaners’, Oxford Journal of Archaeology, 25/​1: 83–​103. Eckardt, H., and Crummy, N. (2008). Styling the Body in Late Iron Age and Roman Britain:  A  Contextual Approach to Toilet Instruments. Monographies Instrumentum 36. Motagnac: Instrumentum. Fitzpatrick, A. (1997). Archaeological Excavations on the Route of the A27 Westhampnett Bypass, West Sussex, 1992, ii. The Cemeteries. Wessex Archaeology Report 12. Salisbury:  Wessex Archaeology. Fitzpatrick, A. (2001). ‘Cross-​Channel Exchange, Hengistbury Head, and the End of the Hillforts’, in J. Collis (ed.), Society and Settlement in Iron Age Europe. Actes du XVIIIe coloque de l’AFEAF Winchester, April 1994. Sheffield: J. R. Collis Publications, 82–​97. Fitzpatrick, A., and Timby, J. (2002). ‘Roman Pottery in Iron Age Britain’, in A. Woodward and J. D. Hill (eds), Prehistoric Britain: The Ceramic Basis. Oxford: Oxbow Books, 161–​172. Fulford, M. (2010). ‘Roman Britain: Immigrations and Material Culture’, in H. Eckardt (ed.), Roman Diasporas: Archaeological Approaches to Mobility and Diversity in the Roman Empire. Journal of Roman Archaeology Supplementary Series 78. Portsmouth, RI: Journal of Roman Archaeology, 67–​78.

280

280   Tom Moore Fulford, M., and Timby, J. (2000). Late Iron Age and Roman Silchester: Excavations on the Site of the Forum Basilica 1977, 1980–​86. Britannia Monograph Series 15. London: Society for the Promotion of Roman Studies. Goldberg, M. (2009). ‘The Dichotomy in Romano-​Celtic Syncretism:  Some Preliminary Thoughts on Vernacular Religion’, in M. Driessen, S. Heeren, J. Hendriks, F. Kemmers, and R. Visser (eds), TRAC 2008: Proceedings of the Eighteenth Annual Theoretical Roman Archaeology Conference, Amsterdam, 2008. Oxford: Oxbow Books, 187–​202. Gowland, R. (2007). ‘Beyond Ethnicity: Symbols of Social Identity from the Fourth to Sixth Centuries in England’, in S. Semple and H. Williams (eds), Early Medieval Mortuary Practices. Anglo-​Saxon Studies in Archaeology and History 14. Oxford: Oxbow Books, 56–​65. Hamilton, S. (2007). ‘Cultural Choices in the “British Eastern Channel Area” in the Late Pre-​ Roman Iron Age’, in C. Haselgrove and T. Moore (eds), The Later Iron Age in Britain and Beyond. Oxford: Oxbow Books, 81–​106. Haselgrove, C. (1982). ‘Wealth, Prestige and Power: The Dynamics of Late Iron Age Political Centralisation in South-​East England’, in C. Renfrew and S. Shennan (eds), Ranking, Resource and Exchange. Cambridge: Cambridge University Press, 79–​88. Haselgrove, C. (1995). ‘Social and Symbolic Order in the Origins and Layout of Roman Villas’, in J. Metzler, M. Millett, N. Roymans, and J. Slofstra (eds), Integration in the Early Roman West:  The Role of Culture and Ideology. Luxembourg:  Dossiers d’Archéologie du Musée National d’Histoire et d’Art IV, 65–​76. Haselgrove, C. (1996). ‘Roman Impact on Rural Settlement and Society in Southern Picardy’, in N. Roymans (ed.), From the Sword to the Plough. Amsterdam: Amsterdam University Press, 127–​188. Hawkes, C., and Dunning, G. (1931). ‘The Belgae of Gaul and Britain’, Archaeological Journal, 87: 150–​335. Hill, J. D. (1997). ‘The End of One Kind of Body and the Beginning of Another Kind of Body? Toilet Instruments and Romanization’, in A. Gwilt and C. Haselgrove (eds), Reconstructing Iron Age Societies. Oxford: Oxbow Books, 96–​107. Hill, J. D. (2007). ‘The Dynamics of Social Change in Later Iron Age Eastern and South-​Eastern England c.300 bc–​ad 43’, in C. Haselgrove and T. Moore (eds), The Later Iron Age in Britain and Beyond. Oxford: Oxbow Books, 16–​40. James, S. (2001). ‘Romanization and the Peoples of Britain’, in S. Keay and N. Terrenato (eds), Italy and the West: Comparative Issues in Romanization. Oxford: Oxbow Books, 187–​209. King, A. (1984). ‘Animal Bones and Dietary Identity of Military and Civilian Groups in Roman Britain, Germany and Gaul’, in T. F. C. Blagg and A. C. King (eds), Military and Civilian in Roman Britain:  Cultural Relationships in a Frontier Province. BAR British Series 136. Oxford: Archaeopress, 187–​218. King, A. (1988). ‘A Comment on the Animal Bones’, in T. W. Potter and S. D. Trow (eds), ‘Puckeridge-​Braughing, Hertfordshire: The Ermine Street Excavations 1971–​2’, Hertfordshire Archaeology, 10: 154–​155. King, A. (1999). ‘Diet in the Roman World: A Regional Inter-​Site Comparison of the Mammal Bones’, Journal of Roman Archaeology, 12: 168–​202. King. A. (2007). ‘Romano-​ Celtic Temples in Britain:  Gallic Influence or Indigenous Development?’, in R. Haeussler and A. King (eds), Continuity and Innovation in Religion in the Roman West. Volume 1. Journal of Roman Archaeology Supplementary Series 67. Portsmouth, RI: Journal of Roman Archaeology, 13–​18.

Britain, Gaul, and Germany: Cultural Interactions    281 King, A., and Soffe, G. (2001). ‘Internal Organisation and Deposition at the Iron Age Temple on Hayling Island, Hampshire’, in J. Collis (ed.), Society and Settlement in Iron Age Europe: Actes du XVIIIe coloque de l’AFEAF Winchester, April 1994. Sheffield:  J. R.  Collis Publications, 111–​124. Lefort, A., and Marcigny, C. (2009). ‘La Probable Agglomération portuaire du second Âge du Fer d’Urville-​Nacqueville: Un état de la documentation’, Bulletin de l’AMARAI, 22: 39–​81. Mattingly, D. (2004). ‘Being Roman: Expressing Identity in a Provincial Setting’, Journal of Roman Archaeology, 17: 5–​25. Mattingly, D. (2006). An Imperial Possession: Britain in the Roman Empire. London: Penguin. Meadows, K. (1994). ‘You are What You Eat: Diet, Identity and Romanisation’, in S. Cottam, D. Dungworth, S. Scott, and J. Taylor (eds), TRAC 94:  Proceedings of the Fourth Annual Theoretical Roman Archaeology Conference, Durham. Oxford: Oxbow Books, 133–​140. Metzler, J., Waringo, R., Bis, R., and Metzler-​Zens, N. (1991). Clemency et les tombes de l’aristocratie en Gaule Belgique. Luxembourg: Musée National d’Histoire et d’Art. Millett, M. (1990). The Romanization of Britain. Cambridge: Cambridge University Press. Millett, M. (1995). ‘Re-​Thinking Religion in Romanization’, in J. Metzler, M. Millett, N. Roymans, and J. Slofstra (eds), Integration in the Early Roman West: The Role of Culture and Ideology. Luxembourg:  Dossiers d’Archéologie du Musée National d’Histoire et d’Art IV, 93–​100. Moore, T. (2011). ‘Detribalizing the Later Prehistoric Past: Concepts of Tribes in Iron Age and Roman Studies’, Journal of Social Archaeology, 11: 334–​360. Moore, T. (2012). ‘Beyond the Oppida: Polyfocal Complexes and Late Iron Age Societies in Southern Britain’, Oxford Journal of Archaeology, 31/​4: 391–​417. Niblett, R. (1999). The Excavations of a Ceremonial Site at Folly Lane, Verulamium. Britannia Monograph Series 14. London: Britannia. O’Gorman, E. (1993). ‘No Place like Rome: Identity and Difference in the Germania of Tacitus’, Ramus, 22/​2: 135–​154. Pearce, J. (2010). ‘Burial, Identity and Migration in the Roman World’, in H. Eckardt (ed.), Roman Diasporas: Archaeological Approaches to Mobility and Diversity in the Roman Empire. Journal of Roman Archaeology Supplementary Series 78. Portsmouth, RI: Journal of Roman Archaeology, 79–​98. Perring, D. (2002). The Roman House in Britain. London: Routledge. Reece, R. (1988). My Roman Britain. Cirencester: Cotswold Studies. Reilly, K. (2009). ‘Animal Bone’, in S. Trow, S. James, and T. Moore (eds), Becoming Roman, Being Gallic, Staying British: Research and Excavations at Ditches ‘Hillfort’ and Villa 1984–​ 2006. Oxford: Oxbow Books, 187–​210. Roymans, N. (2004). Ethnic Identity and Imperial Power: The Batavians in the Early Roman Empire. Amsterdam Archaeological Studies 10. Amsterdam: Amsterdam University Press. Smith, J. T. (1997). Roman Villas: A Study in Social Structure. London: Routledge. Swan, V. (2009). Ethnicity, Conquest and Recruitment:  Two Case Studies from the Northern Military Provinces. Journal of Roman Archaeology Supplementary Series 72. Portsmouth, RI: Journal of Roman Archaeology. Swift, E. (2000). The End of the Western Roman Empire:  An Archaeological Investigation. Stroud: Tempus. Taylor, A., Weale, A., and Ford, S. (2014). Bronze Age, Iron Age and Roman Landscapes of the Coastal Plain and a Late Iron Age Warrior Burial at North Bersted, Bognor Regis, West Sussex:

282

282   Tom Moore Excavations 2007–2010. Reading: Thames Valley Archaeological Services Monograph Volume 19. Taylor, J. (2012). ‘The Idea of the Villa: Reassessing Villa Development in South-​East Britain’, in N. Roymans and T. Derks (eds), Villa Landscapes in the Roman North: Economy, Culture and Lifestyles. Amsterdam: Amsterdam University Press, 179–​194. Theuws, F. (2009). ‘Grave Goods, Ethnicity, and the Rhetoric of Burial Rites in Late Antique Northern Gaul’, in N. Roymans and T. Derks (eds), Ethnic Constructs in Antiquity: The Role of Power and Tradition. Amsterdam: Amsterdam University Press, 283–​320. Trow, S., James, S., and Moore, T. (2009). Becoming Roman, Being Gallic, Staying British: Research and Excavations at Ditches ‘Hillfort’ and Villa 1984–​2006. Oxford: Oxbow Books. Webley, L. (2015). ‘Rethinking Iron Age Connections across the Channel and North Sea’, in D. Garrow, H. Lamdin-​Whymark, and F. Sturt (eds), Continental Connections: Cross-​Channel Relationships from the Lower Palaeolithic to the Iron Age. Oxford: Oxbow Books [ebook]. Webster, J. (1995). ‘Interpretatio:  Roman Word Power and the Celtic Gods’, Britannia, 26: 153–​162. Webster, J. (1999). ‘At the End of the World: Druidic and Other Revitalization Movements in Post-​Conquest Gaul and Britain’, Britannia, 30: 1–​20. Wellington, I. (2001). ‘Iron Age Coinage on the Isle of Wight’, Oxford Journal of Archaeology, 20/​1: 39–​57. Wells, P. (1999). The Barbarians Speak. Princeton: Princeton University Press. Woolf, G. (1998). Becoming Roman:  The Origins of Provincial Civilization in Gaul. Cambridge: Cambridge University Press. Woolf, G. (2001). ‘Representation as Cult: The Case of the Jupiter Columns’, in W. Spickermann, H. Cancik, and J. Rüpke (eds), Religion in den Provinzen Roms. Tubingen: Mohr Siebeck, 117–​134. Woolf, G. (2011). Tales of the Barbarians:  Ethnography and Empire in the Roman West. London: Wiley-​Blackwell.

PA RT  I I

SOCIETY AND T H E  I N DI V I DUA L

284

Chapter 14

Inscrip tions a nd Identi t y Valerie M. Hope

IMP CAES DIV[I NERVAE F] /​NERVAE TRAIA[NO AVG] /​GER PONTIF MAXIM[O TRIB] /​POTEST P P /​COS III /​LEG II AVG For the emperor Caesar Nerva Trajan Augustus, conqueror of Germany, son of the deified Nerva, pontifex maximus, with tribunician power, father of his country, consul for the third time, the Second Augustan Legion made this. (RIB 330; Caerleon, dedicatory panel for a building or structure) DEA DIA/​NA SACRATI/​SSIMA VOTV/​M SOLVIT /​VETTIVS BE/​ NIGNVS LIB To the most sacred goddess Diana, Vettius Benignus, a freed slave, fulfilled his vow. (RIB 138; Bath, small votive altar) D M /​IVLIE BRICE AN XXXI /​SEPRONIE MARTINE AN VI /​ SEPRONIVS MARTINVS F C To the spirits of the departed Julia Brica, aged 31, and Sepronia Martina, aged 6, Sepronius Martinus had this set up. (RIB 686; York, tombstone with full figure portrait images of the deceased above the epitaph)

286

286   Valerie M. Hope

Inscriptions on Stone Latin inscriptions capture the imagination by presenting us with the range of Roman society: children and adults, men and women, soldiers and civilians, emperors and slaves, gods and mortals. Inscriptions can record the names of the powerful, but more often represent ordinary people doing relatively everyday things such as remembering the dead or thanking a god. The term ‘inscription’ is wide ranging in its modern application; it can incorporate words cut on stone, scratched or painted graffiti, labels on pots, stamps on tiles, and the text of writing tablets. The intention here is to focus on inscriptions on stone, sometimes termed monumental inscriptions (for a useful summary of the distinctions between, and different types of, inscriptions, see Tomlin 2011: 135; Cooley 2012: 117–​126; RIB I and RIB III record inscriptions on stone). Arguably these monumental inscriptions were far from mundane, since in Roman Britain they seem to have been relatively unusual and thus were, and continue to be, special. Many of the surviving inscribed stone objects were prestige items or associated with substantial monuments, and the creation of these objects and buildings required considered decisions, as well as expense. To inscribe on stone involved careful evaluation of what to say and how to say it, and implied that the chosen words would have a claim on permanency. The inscriber or dedicator was thinking about the past, present, and future, since they were recording what was important in both their past and present and projecting it forwards in time. When setting up, or commissioning, the inscriptions given ‘at the head of this chapter’, Vettius Benignus, Sepronius Martinus, and the members of the Second Legion must have been conscious of saying the right things, communicating a message, and publicly honouring commitments and relationships (equal and unequal, human and divine). Inscriptions on stone were generally only a small part of a much bigger whole—​both a bigger sequence of actions, rituals, or customs and a bigger monument, building, or structure. To be understood in full, inscriptions cannot be isolated from these acts and monuments or from the people who made them and the audience that viewed them. In reality, reconstructing how inscriptions worked, were read, and reacted to, and how these things changed across time, is not always straightforward owing to frequent loss of context. Nevertheless we can still ask who commissioned inscriptions in Roman Britain, where they were set up and why, and what collectively and individually these inscriptions reveal about life in the province. This chapter looks briefly at where, and in what numbers, stone inscriptions survive from Roman Britain and the nature of the ‘epigraphic habit’ in the province, before considering how and what the inscriptions communicated and their role in identity construction.

Numbers and Distribution From Roman Britain there are, approximately, 2,800 surviving inscriptions on stone. Many of these inscriptions are incomplete, rendering reading and full understanding

Inscriptions and Identity    287 of the original text impossible. Where the purpose of the inscription can be established, most fall into the broad categories of religious dedications (mainly to named gods and goddesses on altars), epitaphs (mainly on stele, with a few from larger monuments), building dedications (mainly inscribed panels recording the construction or restoration of buildings), centurial stones (small panels indicating construction work completed by centuries of legions and other military units, particularly common on Hadrian’s Wall), and milestones (pillars recording the emperor’s titles and distances). There has been little research on Romano-​British stone inscriptions as a complete body of evidence and thus numerical (and other) comparisons between different types of inscriptions. Most research focuses on specific categories of inscription—​for example, epitaphs or building dedications. In addition, attaching figures to these broad categories is problematic owing to damage to stones and also the overlap between categories. Some very rough numbers for these categories would be:  votive inscriptions—​well in excess of 1,000; epitaphs—​in the region of 600; building dedications—​around 200; centurial stones—​ around 250; and milestones—​approximately 100; with the remaining fragmentary inscriptions being of uncertain use. In terms of find-​spots, the surviving inscriptions are not evenly distributed across the province. There are marked concentrations of inscriptions at military bases, especially in the north of the province (Biró 1975; Cepas 1989; Mattingly 2008). Most inscriptions originate from the legionary garrisons at Chester, Caerleon, and York, from along the length of Hadrian’s Wall, and from the provincial capital of London. The coloniae of legionary veterans (Colchester, Lincoln, and Gloucester) and towns that had strong military or official connections (Bath and Richborough) also produce a number of inscriptions. The native civitates, by contrast, did not extensively set up inscriptions, though small clusters are found at Silchester, Wroxter, and St Albans (Jones and Mattingly 1990: map 5.10). If inscriptions in other materials are included to create overall ‘epigraphic indexes’, these concentrations change little (Mattingly 2008; for mapping of the epigraphy (of all types) at Roman London, see Holder 2007; find-​spots for sculpture from across Roman Britain also suggest a similar distribution, with clear military and urban biases; see Stewart 2010). The chronological distribution of these inscriptions is also uneven, though it needs to be noted that the precise dating of inscriptions is a precarious business. Latin inscriptions rarely contained dates or other information (for example, references to emperors or consuls or specific events) that allow for the allocation of a certain date or date range. However, approximate dates (and periods) can be suggested for some inscriptions by comparison with examples that can be dated (e.g. RIB 331, 665, 667), reference to known troop movements (though these are often supported or inferred by inscriptions, r​ isking circular arguments), and certain epigraphic and stylistic conventions. The approximate chronological distribution of the inscriptions of Roman Britain indicates that production increased during the later first century, was at its height during the second and early third centuries, with a marked decline in numbers during the later third century; there may have been a small resurgence in Christian epitaphs during the late fourth century (Hope 1997; Handley 2001; Knight 2010).

288

288   Valerie M. Hope

The Epigraphic Habit Compared to many other provinces of the Roman Empire, the number of inscriptions that survive from Roman Britain is paltry. In terms of its epigraphy, Roman Britain would seem to be the poor relation to the rest of the empire, an oddity and an epigraphic backwater. Is this epigraphic difference between Britain and other provinces the misfortune of what has survived or does it reveal important things about Britain’s place in the Roman Empire? Are the number and distribution of inscriptions related to levels of literacy, to the uptake of Latin, to urbanization and the extent of ‘Romanization’? Big questions and issues such as these are often explored through epigraphy, and the act of inscribing on stone in Latin can be viewed as a gauge (however crude) of acculturation and the dialogue between Roman and indigenous cultures (Cepas 1989: 1; Cherry 1998; Woolf 1998). At the outset it needs to be acknowledged that chance and circumstance have played their role. The vagaries of survival may account for low numbers of inscriptions from some parts of Roman Britain, with many stones vanishing from view, owing to loss, reuse, or destruction. For example, if so many inscribed stones had not been reused in, and then recovered from, Chester’s late defensive walls, the state of Romano-​British epigraphy would be all the poorer (Richmond and Wright 1956; Clay 2004). However, this same factor of chance survival could hold for many other regions of the Roman Empire, where only a small percentage of what once existed has probably survived (for relative numbers, see Bodel 2001: 6–​10). No doubt many inscribed stones have been lost from Roman Britain, but what now exists may still be roughly indicative of relative rates of production. Another factor that may have impacted upon the creation of inscriptions is the poor supply of suitable stone, especially in the south of the province (Mann 1985; see also Stewart 2010 for stone used in sculptures). However, some areas did overcome supply issues, even importing stone from large distances. Inscribed imported marble slabs are to be found, for example, at Caerleon (RIB 330) and London (Hassall and Tomlin 1989: 326), while the early tombstone to the legionary centurion Marcus Favonius Facilis (RIB 200) at Colchester was made from a Gallic limestone also used for military tombstones at Mainz, Bonn, and Neuss (Hayward 2006; see also Isserlin 1998; Hayward 2009). The absence of good stone must have had an impact upon the ability to inscribe and monumentalize at some sites, but is not a sufficient explanation for the overall low number of Romano-​British inscriptions. Britain may have fewer inscriptions than many other provinces, yet it is not alone in the unevenness (geographical and chronological) of what survives, which is important for our understanding of the epigraphy of the province. In general, the act of inscribing in the Roman world was an urban phenomenon, and, in many regions, it was closely aligned with the military and centres of administrative power (Bodel 2001: 80). Large areas of Gaul—​such as Brittany and western Gaul, for example—​produced no or very few inscriptions (Woolf 1998:  maps 4.5 and 4.6); from the German provinces most

Inscriptions and Identity    289 inscriptions have been found at military forts and veteran colonies (for example, Mainz and Cologne). What survives from Britain, even in reduced numbers, follows this trend. The use of epigraphy was also not chronologically static. A brief article published by MacMullen (1982; building on earlier work by Mrozek 1973) has had a substantial impact in identifying ‘the epigraphic habit’ and its inconsistencies. MacMullen noted that the production of inscriptions during the Roman period rose and fell; inscribing was not a constant phenomenon. In particular, MacMullen mapped how inscribing increased in popularity during the first and second centuries ad and then rapidly fell during the third century ad. We should not overlook that the accurate dating of inscriptions is fraught with difficulty (see ‘Numbers and Distribution’, above), so such charting of rises and falls can only ever be approximate. Nevertheless, even allowing for differences between sites (and different types of epigraphy), MacMullen identified an empire-​wide trend. It is clear that the act of inscribing in Latin, and setting up certain types of monuments with which inscriptions were associated, went in and out of fashion. Explaining ‘the epigraphic habit’ has taxed a few minds, and explanations have noted regional variations and patterns and have highlighted factors such as status display, social mobility, geographic mobility, elite emulation, and patronage networks (Meyer 1990, 2011; Woolf 1996; Häussler 1998). It may well be that there is no simple all-​encompassing explanation for the rise and fall of the ‘epigraphic habit’ that suits all sites or parts of the empire equally; and it is certainly the case that finding factors (for example, spread of citizenship or status display) for the uptake of inscribing is often easier than accounting for its decline (Meyer 2011). For Roman Britain, with its comparatively low numbers of inscriptions, it is possible that the ‘epigraphic habit’ had insufficient time to catch on, before it went into decline among key groups (Hope 1997). This possibility becomes more marked when it is noted that elsewhere in the Latin-​inscribing west most inscriptions were epitaphs (Saller and Shaw 1984: 124; Bodel 2001: 30; Meyer 2011: 193). In the German provinces, from where many of Britain’s ‘Roman’ soldiers first arrived, the number of epitaphs, and the ‘epigraphic habit’ in general, peaked in the late first century. By the time the troops were securely established in Roman Britain, the relevance of inscribed tombstones may have already been on the wane, for soldiers at least. In Roman Britain epitaphs are outnumbered by votive offerings to deities, and these also eclipse epitaphs at some other military sites elsewhere in the empire during the second century ad, although never rivalling the earlier popularity of epitaphs (Biró 1975: 27, 42; Decker and Selzer 1976; Cepas 1989: 56; Forbis 1996; Derks 1998; Hope 2001: 93). In short, the Roman ‘epigraphic habit’ (at least in the west), which was mainly an urban and military phenomenon, incorporated different types of epigraphy, and that associated with commemorating the dead, which was often the most popular, did not substantially take hold in Roman Britain. In Roman Britain the low level of epigraphy cannot be viewed as a general indicator of poor literacy, low levels of urbanization, or some sort of rejection of ‘Roman’ display (especially funerary) culture. Instead, epigraphy remained largely military, largely religious, and not of interest or relevance to the majority of the population, and, in this,

290

290   Valerie M. Hope Roman Britain is not that different from some other parts of the empire. Elsewhere it was epitaphs that most urban populations ‘bought into’, and the significance of these epitaphs may have been declining (especially among the military that came to Britain) by the time Roman forts and towns were securely established. In general among the civil population, neither the poor nor the moderately well-​off nor the wealthy set up inscriptions. The absence of the most economically and politically successful from the epigraphic record is striking, and it seems likely that the indigenous British elite did not compete within either the monumental or the epigraphic sphere, since their position was embedded in native traditions not established or contested through urban architectural munificence (Biró 1975; Blagg 1990; Millett 1990: 105; Revell 2007; 2009: 183; Eckardt et al. 2009). There is evidence, for example, for a range of tomb types, including some substantial structures, with an eye to visibility and integration into the landscape, in both rural and peri-​urban zones, even if these were, for the most part, un-​inscribed (Esmonde Cleary 2000: 137; Strück 2000; Pearce 2011). The low level of epigraphy does not indicate that the civil population was actively rejecting epigraphy, but that it may simply have preferred other methods of display, such as mosaics (Mattingly 2008: 68), and display locations other than just the main urban centres. A range of factors was probably at play as to why Britons did not take up epigraphy in any great numbers, but this makes those inscriptions that do survive, and those who used them, all the more interesting.

Communicating to an Audience Inscriptions, by their very nature, suggest social interaction and communication. The commissioner decided that he or she had something—​a building, for example—​and also the very act of creating (or restoring) that building that were worth recording in verbal form. Initially the commissioner of the inscription entered into a dialogue with the stonemason, engaging him and dictating, or creating with him, a suitable text. Once set up, the inscription was poised to engage the passer-​by, to gain an audience, who, by looking and reading, would continue to note the building and the actions of the people named. It is possible that some inscriptions were concealed from view (inside a tomb or shrine, for example), visible to only a select few, but the vast majority of Romano-​British inscriptions on stone do appear to have been public statements, even if we cannot know the exact identity of the target audience, or the extent to which that audience paused, read, understood, and appreciated the information therein. Inscriptions were not cut in native dialects. Latin was the language of inscriptions (for the few exceptions see ‘Identity’). It was through Latin that the inscription communicated; through Latin that the commissioner of the inscription chose to speak. The use of inscribed Latin suggests a certain level in both the use and the understanding of Latin among those communities (military and civil) in which inscriptions were set up; however, it is difficult to generalize about how many in the inscribing military–​urban

Inscriptions and Identity    291 zones, or indeed those from non-​inscribing areas, spoke and read Latin. The presence or absence of epigraphy, or other forms of the written word (for example, writing tablets), is not a simple gauge of relative literacy (see, e.g., Raybould 1999; Pearce 2004; Adams 2007: 580–58​1; Parsons 2011: 117–​119). To set up a Latin inscription was, nevertheless, a considered choice, and presumes an audience that would be able either to understand the text or to grasp its symbolic significance or both. Inscriptions, and the monuments they were associated with, were put up both to be read and to be seen. How individuals responded to the texts depended on their level of literacy to be sure, but also on factors such as status, personal background, and previous experiences (Barrett 1993: 237). Many inscriptions may have been aimed at peers—​at a fellow Latin-​speaking audience of soldiers and officials—​but inscriptions were accessible and visible to others, who may have responded in varied fashions. It is worth noting, for example, that inscriptions were frequently riddled with abbreviations, and that we can create competing interpretations for the use and impact of these abbreviations. There were, no doubt, cost implications here. The stonemason charged by the letter, and the more letters cut the greater the expense. Abbreviations saved money, yet can also reinforce the impression that inscriptions were a rarefied means of communication, for the special few, for those in the ‘know’ who could break the code. On the other hand, it could be argued that abbreviations made inscriptions more accessible, since it gave them a pictorial or pictogram aspect that, due to repetitive use, combined with the general simplicity of content, made inscriptions easy to decipher and demanded only a low level of literacy (Woolf 1996: 28; Petronius, Satyricon 58.7). Set in context—​a cemetery, an altar, a large building—​inscriptions may have been readily understandable on a general level. You did not need to be able to read Latin or, even if you could, stop and take the time to read it, to get the meaning or construct some sort of message from what was seen, even if this, and the individual response, diverged (in some respects) from the message of the original commissioner. Names were a central part of all inscriptions, and again these were probably relatively easy to spot and ‘read’ in terms of the inscription content. Inscriptions recorded the acts of people, either individuals or groups, who set up buildings or honoured the emperor or dedicated altars to the gods or remembered a dead person. The inscription commemorated these acts and named names, in the process revealing and/​or constructing identities for those involved (see ‘Identity’). Some inscriptions were, in fact, just names. Centurial stones were inscribed with the centurion’s name or other military unit titles, but, by their placement, they affectively, if often with a subtle simplicity, communicated the contribution of that unit to the construction of the building involved. Hadrian’s Wall produces many such markers indicating the completion of stretches of the wall and its structure, such as that carried out by the century of Julius Rufus (RIB 1357), whose men worked on the eastern part of the wall. Epitaphs were similarly centred on names, that of the dead person, and also that of the person(s) who had set up the monument—​a clear indication that inscriptions were about commemorating actions as much as marking places (such as graves or altars). Epitaphs might also state the relationship between the deceased and the commemorator, add an age at death (and or length of service and unit for a soldier), and employ the

292

292   Valerie M. Hope opening phrase Dis Manibus (‘To the spirits of the departed’), often abbreviated to D.M. At Chester, for example, Lucius Sempronius Probianus commemorated his young son: D(is) M(anibus) /​L(ucio) F(e)STINIO PROBO F(i)L(io) /​VIX(it) AN(nos) II D(ies) XXVIIII /​L(ucius) SEM(pronius) PROBIANVS /​PATER F(aciendum) C(uravit) (RIB 537). The epitaph started with the abbreviation D.M., then recorded the boy’s name (Lucius Festinius Probus), before carefully noting his age of 2 years and 29 days, followed by the name of the commemorator (Lucius Sempronius Probianus), his relationship (pater) to the deceased and that he had made the monument (faciendum curavit). Votive altars often placed a similar emphasis on names and relationships, though in this case between the human and the divine. The god’s name was usually inscribed first, followed by that of the dedicator. A statement that the vow had been willingly and deservedly fulfilled (votum solvit libens merito—​often abbreviated to V.S.L.M.) might be included, and sometimes the motivation behind the dedication specified. In County Durham, for example, a cavalryman gave thanks to the god Silvanus for the capture of a fine wild boar (RIB 104), although most votive inscriptions did not reveal the explicit reason for the dedication (for an overview of the content of epitaphs and religious dedications, see Raybould 1999: 10–​33, 94–​118; for lists of inscriptions with brief descriptions, by type, including milestones, building dedications, and religious inscriptions, see de la Bédoyère 1999). Building dedications focused on the name of the dedicator(s), often linking this with deities and or the name of the emperor. A reference to the type of structure was, sometimes but certainly not always, included. In Caerleon, an unspecified building (perhaps the fort rampart) was marked by an imported inscribed panel dedicated by the Second Legion to the emperor Trajan (RIB 330, see above). In York, the fortress gates were similarly dedicated to Trajan, whose full name and honours were listed, by the Ninth Hispanic Legion (RIB I. 665). Milestones, generally in the shape of pillars, also placed emphasis on imperial titles, often giving only the name and titles of the emperor; the distances that these milestones marked were not always inscribed on them (perhaps they were added in paint), but they were affective territorial indicators and statements of imperial authority, especially during the third and early fourth centuries ad. Most inscriptions were then brief and formulaic, yet, because of their verbal dimension, they have been readily mined for insights into Roman society. Names can provide status indicators and suggest ethnic origins; recorded relationships highlight family formation; ages at death hint at demographic profiles; unit titles indicate troop movements; religious dedications suggest ritual practices; and so on. Because inscriptions often survive in large numbers (at least from elsewhere in the empire), they have formed the basis for statistical analysis on demography in particular (see, e.g., Saller and Shaw 1984; Saller 1994; Scheidel 2012). Such approaches are not without problems, since the population commemorated is a self-​selecting one or is skewed towards certain social groups (in particular, see Hopkins 1966). There have been some attempts to employ such approaches with Romano-​British inscriptions, with epitaphs in particular, although the relative paucity of their numbers makes their use and interpretation particularly problematic (Raybould 1999; and, with caution, Tobler and Adams

Inscriptions and Identity    293 2007). To note one example of the difficulties: Saller and Shaw (1984: 139–1​45) have argued that the lack of familial references in Romano-​British epitaphs suggests that there was little local recruitment into the army, contrasting this with provinces such as North Africa and Spain, where soldiers were frequently commemorated by family members, which, they believe, indicates local recruitment. Compared with many other provinces, however, Roman Britain has few relevant epitaphs, and those that do not mention military families mainly date to the first and early second centuries ad, with familial references becoming more frequent in epitaphs dating to the second and early third centuries—​the periods also best represented by the epitaphs from Africa and Spain. In other words, the chronology of the epitaphs is key to the interpretation of the available evidence, and the issue of local recruitment (or not) into the Romano-​British army cannot be resolved using this epigraphic data (Alföldy 1987; Hope 2001: 70–​7 1; 2007: 116–​117; Phang 2001: 245–​246; note also Dobson and Mann 1972; Raybould 1999: 95). The ‘epigraphic habit’, as we have seen, was not chronologically or geographically static, and it appealed to some groups more than to others. In terms of their content, inscriptions were often standardized and repetitive (hence their appeal for number crunching), but they were also only one part of a bigger whole. A building inscription, for example, was a relatively small aspect of the temple, arch, or baths to which it was attached. To catch the eye, inscriptions might be cut on large panels and the latter might be of expensive stone and/​or framed with architectural or sculptural details. Letter size could vary, with larger text size for key information; and some letters may have been painted. The building dedication found at Caerleon to the emperor Trajan by the Second Augustan Legion (RIB 330) was inscribed, and then the letters picked out in red paint, on a large imported marble slab, with the first line of the inscription—​the name of the emperor—​cut in the largest letters. The tomb of the procurator Classicianus, from London, was a substantial funerary altar, the front surface of which was dominated by the text of the epitaph, the opening line ‘DIS MANIBVS’ cut in letters large enough to be read from a distance of hundred paces (Grasby and Tomlin 2002: 71; RIB 12). Several distance slabs from the Antonine Wall used sculptural motifs, such as legionary emblems and deities, to create ornate panels often cleverly combining text and images. In one example a winged Victory holds up a wreath in the centre of which is inscribed the unit title of the Twentieth Legion (Keppie 1998: 87–​89; RIB 220). Similarly epitaphs were often accompanied by images, most often placed above the epigraphic panel; sculpted portraits, for example, drew the eye to the tombstone and worked with the text to build up a picture of the deceased. A few inscriptions also stood out by toying with the standardized content, deviating from the norm by adding a poem or extra personal details (for example, verse epitaphs: RIB 265, 292, 684, 758, 1253). In two examples the inscriptions literally spoke, giving the dead a voice and addressing the traveller. A soldier of the Fourteenth Legion in Wroxeter describes his life in the underworld (RIB 292) advising the reader to ‘live honestly while your star grants you time for life’; while a Greek epitaph (RIB 758) from Brough-​under-​Stainmore, encourages the traveller to call out the name of 16-​year-​old Hermes, for in doing so ‘you will do him a good service’.

294

294   Valerie M. Hope In many ways inscriptions played with the senses of the passers-​by, especially the experiences of sight, hearing, and touch. Inscriptions (and associated sculpture, décor, and buildings) were clearly visual; also we can imagine inscriptions being read aloud and some being touched, as fingers traced letter forms, or when buildings, altars, and graves were visited (especially in ritual contexts). Events (for example, feasting and sacrifices) associated with the places and structures that inscriptions marked may have drawn on the other senses of taste and smell. All this suggests that inscriptions could both be unusual (in Roman Britain in particular) and special, while simultaneously being an integrated and unremarkable aspect of the lived environment. In the urban, military, religious, and cemetery situations in which they were mainly found, inscriptions were part of the landscape, entwined with the structures and patterns of daily life, potentially long after those identified in them were forgotten. Setting up inscriptions may have marked those structures, and the people named, as significant (in those settings at least) and made a claim on posterity, yet the human capacity to remember can be a fickle thing. Inscriptions could have longevity, but some also had a limited life span (Barrett 1993). In Chester, it has been argued that tombstones of the Twentieth Legion may have been deliberately defaced and reused in the North Wall of the fort (Clay 2004); the distance slabs of the Antonine Wall may have been buried when the army drew back (Keppie 1998: 52); inscribed votive altars found buried at Maryport, once thought to have been concealed as part of religious ritual, probably made for convenient ballast for supports for a new building (Hill 1997; Hilts 2011). Individual inscriptions did not change with the times, since they were rarely re-​cut or updated. Some inscriptions stayed in place, continuing to speak when the inherent message ceased to have current relevancy, while many other inscriptions were forgotten, defaced, prematurely silenced, reused, or buried. It is difficult to reconstruct how people read and reacted to inscriptions, in their original setting and across their lifespan, but, whether feted or ignored, inscriptions were about communicating, and they now communicate to modern audiences. Inscriptions are, to modern commentators, highly valuable as sources of information on, for example, the use of Latin language, or demography, or religious practice. Isolating certain elements within inscriptions, at the expense of others, however, can be problematic and risks looking at words without the inscription, the inscription without any associated images, the images without the monument, and the monument without its wider physical and locational context. It is inevitable that we cannot access all these dimensions, but we can prioritize the knowledge that inscriptions were a means, often a multifaceted means, of telling a story, albeit a select and carefully edited story. In their original context, inscriptions spoke to social groups to whom the dedicator or inscriber was claiming membership, as well as to others (outsiders) who saw the monument and reacted in their own individual way. There were, sometimes, other communicative dimensions (even if apparently one-​sided) to these inscriptions, such as spiritual messages to the gods or the claim on a relationship with a higher authority, such as a patron or the emperor. Inscriptions were clearly connected to memory—​put simply, they were set up to commemorate, to record, and to preserve people’s names, and the lives and actions of those people. Inscribed monuments memorialized names and acts for the

Inscriptions and Identity    295 future, by employing a repertoire of established words, phrases, and images; they drew on past practice, they looked to the future, and reflected (or constructed) their present. Inscriptions, and the monuments with which they were associated, could give a sense of rootedness, of continuity; they placed individual acts in a wider communal and societal context (for communicative dimensions, see, in particular, Meyer 2011).

Identity Latin inscriptions are centred on people: most often, named individuals and sometimes groups; for Roman Britain, military units and subunits are often recorded as acting together. It is these names that bring inscriptions to life, especially the sense that inscriptions often record the acts of relatively ordinary individuals (or groups of people). Inscriptions do not state only the names of Roman emperors, governors, military commanders, and the elite of Roman Britain, but also rank-​and-​file soldiers, children, women, and even slaves. Latin names, as recorded in inscriptions, can be very revealing, since they indicate gender and can provide clues about legal status and place of origin. For example, Quintus Valerius Fronto, of the Second Legion, commemorated at Chester (Q(uintus) VALERI/​VS Q(uinti) F(ilius) CLA(udia tribu) /​FRONTO CELE/​A MILES LEG /​II AD(iutricis) P(iae) F(idelis) AN/​NORVM L /​STIPENDIORV/​M XXV … (RIB 479)), is revealed as a Roman citizen by his three names (tria nomina), as free-​born, through his father’s praenomen (‘son of Quintus’) and his voting tribe (‘Claudian’), and as originating from Celeia (in Noricum, now in Slovenia). The flipside of such a wealth of information are those names that say frustratingly little. The single name of Vacia, commemorated at Carlisle (DIS (Manibus) /​VACIA INF/​ANS AN(norum) III (RIB 961))], reveals only her gender, and not explicitly her legal status, place of origin, or parentage, but the accompanying portrait, and her description as ‘an infant of three years’, powerfully evokes the sentiments caused by her death. Because inscriptions are centred on people, and their names, they present and fashion images of the self. Any commissioner, as noted above, could not control the audience response to, and the ultimate fate of, an inscribed monument, but he or she did make considered choices about what the inscription should say and how individuals and groups should be represented. This active decision-​making is connected to why inscriptions had relevance as a means of communication to some individuals, and groups, more than to others. Romano-​British inscriptions do not include a random cross-​section of people, as tempting as it might be to wish for this. Time, geography, and personal experience all played their part in where, when, and by whom inscriptions were set up. Those who used inscriptions often had something particular to say, or present, and sometimes an axe to grind, or a difference to announce. There were reasons why some people chose to inscribe and equally why others did not. Romano-​British inscriptions were largely military, and thus male, in character. The army was a substantial presence, but still only ‘a minority group’ in terms of the overall

296

296   Valerie M. Hope population of Roman Britain (Mattingly 2004: 15). Soldiers were the main inscribers, or commissioners of inscriptions, and thus these inscriptions acted as a clear symbol of military identity (Mattingly 2008: 68) and also military community. Inscriptions, and the monuments associated with them, were aspects of the backdrop, routine, and conformity of military life, which helped forge a shared identity and commonality for soldiers who may have originated from across the empire (MacMullen 1984; James 1999; Kampen 2006: 132). Inscriptions were a significant feature of how soldiers communicated with each other, and also with others (or represented themselves to others) such as enemies, civilians, authority figures (for example, the emperor), and even the gods. We should not forget, however, that different types of inscription had different roles. In the military context, milestones, centurial stones, building dedications, and some votive offerings often had an official or semi-​official purpose, such as recording the completion of designated acts and allegiance to the military unit and/​or the emperor. The content of these inscriptions might vary little, but their impact, especially when taken together, was one of organization, power, and control—​a symbol of both the Roman presence and military authority. These inscriptions were part of how Rome was written onto the British landscape. For ordinary soldiers, however, these same inscriptions were perhaps less a symbol of power and more one of common acts, common allegiances, and a shared military identity. Inscriptions were not just unchanging signs of the Roman and military presence, since many, especially epitaphs and votive offerings, incorporated more personal dimensions, and, in addition, the role and impact of inscriptions also shifted with time. Inscriptions could celebrate military identity, military unity, differences between soldiers and civilians, and also differences between military units (see below), but the relevancy of some of these messages would have waxed and waned as the army marched on, or established permanent bases, or sent squadrons to serve elsewhere in Britain and the empire. In the early years of the province, for example, building dedications within a fort, and the epitaphs and tombstones that ringed that fort, may have symbolized military unity to the soldiers, their ‘Roman’ identity, and a territorial claim on the new province; while to the recently subjugated population they may have been all these things with a more sinister twist. As the years passed, however, and units relocated or became more permanently stationed, or were even locally recruited, the messages inherent in, and to some extent the relevance of, these inscriptions (both old and new) would have changed. Identity could be highly personal and individualized, and to suggest that inscriptions primarily promoted some sort of overarching military identity can overlook this. Inscriptions often suggest diversity and difference and underline the complex and interwoven aspects of individual identity in the Roman world (Hope 2000a; Mattingly 2004: 10–​11; 2006: 18). Military epitaphs and tombstones, for example, might suggest the unity and conformity of the army, but many of these memorials marked the graves of men of rank, or of men serving away from their home base (Hope 2003; see also Russell 2010). These people, and those who commemorated them, did not want the things that made them different in the environments where they were commemorated, to be forgotten (Hope 2003: 135). Similarly, the most striking Romano-​British tombstones, those

Inscriptions and Identity    297 decorated with portraiture, were promoting not just an un-​problematic Roman identity, but often a sense of difference from the majority culture—​a sense, among both those commemorating and commemorated, of being, in some respects, an outsider. These decorated tombstones were often set up to soldiers in newly conquered areas or to auxiliary troops who were not Roman citizens and to women who were not fully enfranchised members of the communities in which they lived (Hope 1997; Stewart 2009). The complexities of identity are well illustrated by auxiliary tombstones decorated with detailed cavalry scenes, which were popular mainly during the first century. A recently discovered tombstone from Lancaster commemorating Insus, son of Vodullus, a citizen of the Treveri of the ala Augusta, depicts him in a splendid helmet, brandishing, by the hair, a decapitated head, the corpse lying at the feet of his rearing mount (Hassall and Tomlin 2006: 468–​471). On this and similar monuments, horsemen ride down fallen barbarians, representing, it would seem, Roman victory and supremacy, yet the epitaphs reveal that those commemorated, including Insus, were members of non-​citizen auxiliary units, who might be paid less well than their legionary counterparts, might be regarded as more expendable in battle, and were often recruited from recently subjugated provinces (Hope 1997, 2000b; Stewart 2010). In addition to these funerary monuments, we can also note that, in the north of the province, votive altars, including those set up to Jupiter, were more often the product of auxiliary soldiers than of legionaries—​ which again has been related to their non-​citizen identity (Vanderspoel 2005: 46–​52). There is an irony here that, on the surface, many Latin inscriptions present and symbolize Roman (and military) identity, yet, if we dig a little deeper, those associated with these inscriptions were sometimes individuals for whom this identity could be contested. This is not to claim that all the inscriptions of Roman Britain were aberrations, generated only by those with chips on their shoulders, but that identity could be complex and multifaceted; in the Roman world one could both belong and not belong, and the act of inscribing (with associated monuments and sculpture) could be a method of presenting, editing, and negotiating that identity, as witnessed elsewhere in the Roman world for groups such as freed slaves and gladiators (see, e.g., Kleiner 1977; Kockel 1993; Hope 1998, 2001; Mouritsen 2005). To end, we can note a group of Romano-​British inscriptions that stand out because of a notable deviation to the norm. It was observed above that language united the inscriptions—​the Latin in which they were composed was integral to the identity that was projected. This is not, however, strictly true, since there is a handful of Romano-​British inscriptions cut in Greek (RIB 461, 758, 808, 1072, 1124, 1129, 3151). The choice to inscribe in Greek was a considered one, suggesting the significance of Greek culture and language to a few. Finding someone to inscribe in Greek may have been challenging and emphasizes a choice ‘to make a strong statement of ethnic and cultural identity’ (Noy 2010: 22). Others present in Roman Britain who originated from Greece chose Latin not Greek for their inscriptions (RIB 251, 864, 955). Where Greek is used, it seems to be loosely connected to medicine (Noy 2000: 173) and trade—​once more a choice to emphasize difference and individuality. The poem to young Hermes (as noted in ‘Communicating to an audience’ —​RIB 758) was

298

298   Valerie M. Hope found at Brough-​under-​Stainmore, which may have served as a trade depot (Birley 1979: 111); two votive altars from Chester (RIB 461, 3151) were both set up by doctors; a dedication at Maryport to Asclepius (god of healing) may also hint at medical connections for its dedicator Aulus Egnatius Pastor (RIB 808); and the same god was honoured at Lanchester in a bilingual inscription (RIB 1072) set up by a soldier. Perhaps the ultimate symbol of the complexities of identity, however, is the tombstone of Regina, which proclaims her connections to Britain, Rome, and Palmyra (RIB 1065). Regina may never have visited the latter two places, but had moved from south-​east England (she is described as a Catuvellaunian) to South Shields (where the tombstone was found), and the Latin of her epitaph, coupled with her Palymrene husband (Barates), links her to the wider empire. The accompanying portrait, perhaps carved by a Palmyrene hand (Phillips 1976), presents Regina, an ex-​slave, as a ‘Roman’ lady, although sporting apparently native British clothing (Noy 2010; Carroll 2012). To her husband went the last words, cut in Palmyrene Aramaic: ‘Regina, freedwoman of Barates, Alas!’ As with the Greek inscriptions, we are left to wonder who carved the Palmyrene text. Was it Barates himself? Would anyone else in South Shields have been able to understand it? Did that matter—​after all everyone would be able to tell that there was something different about the tombstone, and about Regina and Barates, without having to grasp all the details (Mullen 2012: 1–​5)? Regina’s short life, even if confined to the shores of Britain, had been a diverse journey and experience, and at the last that life was summed up and viewed through the prism of her commemorator’s equally diverse experiences (for a discussion of multiculturalism, see Nesbitt, this volume; see also Leach et al. 2010; Tolia-​Kelly 2010). These Greek inscriptions, and Regina’s bilingual inscription, toy with identity, both reflecting and creating it; they celebrate a sense of belonging and a sense of difference; the inscriptions capture cultural diversity and mixed identities that characterized at least some people’s experiences of living in the Roman Empire. Setting up inscriptions in Roman Britain may never have been popular, or the norm, but, for a few, it was a powerful medium for communicating, for recording important actions, and for telling stories. The inscriptions on stone that survive from Roman Britain are so low in numbers, and so focused on certain areas and certain social groups, that it can be hard to generalize from them; nevertheless they do suggest aspects of varied identities and how these were represented, negotiated, and constructed. The inscriptions suggest a sense of what it was to be ‘Roman’ (and often a member of the army) in Britain, and simultaneously the fluidity of identities in an outpost of empire.

Abbreviations RIB   R. G. Collingwood and R. P. Wright, The Roman Inscriptions of Britain: Inscriptions on Stone. Oxford: Clarendon Press, 1965–​to date; Oxford: Oxbow Books, 2009.

Inscriptions and Identity    299

References Primary Sources Petronius. The Satyricon, trans. J. P. Sullivan. London: Penguin Classics, 2011.

Secondary Sources Adams, J. N. (2007). The Regional Diversification of Latin 200 bc–​ad 600. Cambridge: Cambridge University Press. Alföldy, G. (1987). ‘Das Heer in der sozial Struktur des römischen Kaiserreiches’, in G. Alföldy (ed.), Römische Heersgeschichte: Beitrage 1962–​1985. Amsterdam: J. C. Gieben, 33–​34. Barrett, J. (1993). ‘Chronologies of Remembrance:  The Interpretation of Some Roman Inscriptions’, World Archaeology, 25/​2: 236–​247. Birley, A. (1979). The People of Roman Britian. London: B. T. Batsford. Biró, M. (1975). ‘The Inscriptions of Roman Britain’, Acta Archaeoloica Academiae Scientiarum Hungaricae, 27: 13–​58. Blagg, T. F.  C. (1990). ‘Architectural Munificence in Roman Britain:  The Evidence of the Inscriptions’, Britannia, 21: 13–​31. Bodel, J. (2001) (eds). Epigraphic Evidence:  Ancient History from Inscriptions. London and New York: Routledge. Carroll, M. (2012). ‘“The Insignia of Women”:  Dress, Gender and Identity on the Roman Funerary Monument of Regina from Arbeia’, Archaeological Journal, 169: 281–​311. Cepas, A. (1989). The North of Britannia and the North-​West of Hispania:  An Epigraphic Comparison. BAR International Series 470. Oxford: Archaeopress. Cherry, D. (1998). Frontier and Society in Roman North Africa. Oxford: Clarendon Press. Clay, C. (2004). ‘Iconoclasm in Roman Chester: The Significance of the Mutilated Tombstones from the North Wall’, Journal of the British Archaeological Association, 157: 1–​16. Cooley, A. (2012). The Cambridge Manual of Latin Epigraphy. Cambridge:  Cambridge University Press. Decker, K., and Selzer, W. (1976). ‘Mogontiacum: Mainz von Zeit des Augustus bis Ende der römischen Herrschaft’, Aufstieg und Niedergang der Römischen Welt II.5. Berlin: de Gruyter, 457–​559. de la Bédoyère, G. (1999). Companion to Roman Britain. Stroud: Tempus. Derks, T. (1998). Gods, Temples and Ritual Practices: The Transformation of Religious Ideas and Values in Roman Gaul. Amsterdam: Amsterdam University Press. Dobson, B., and Mann, J. C. (1972). ‘The Roman Army in Britain and Britons in the Roman Army’, Britannia, 4: 191–​205. Eckardt, H., Brewer, P., Hay, S., and Poppy, S. (2009). ‘Roman Barrows and their Landscape Context: A GIS Case Study at Bartlow, Cambridgeshire’, Britannia, 40: 65–​98. Esmonde Cleary, S. (2000). ‘Putting the Dead in their Place: Burial Location in Roman Britain’, in J. Pearce, M. Millet, and M. Strück (eds), Burial, Society and Context in the Roman World. Oxford: Oxbow Books, 127–​142. Forbis, E. (1996). Municipal Virtues in the Roman Empire: The Evidence of Roman Honorary Inscriptions. Stuttgart and Leipzig: Teubner. Grasby, R. D., and Tomlin R. S.  O. (2002). ‘The Sepulchral Monument of the Procurator C. Julius Classicianus’, Britannia, 33: 43–​76.

300

300   Valerie M. Hope Handley, M. (2001). ‘The Origins of Christian Commemoration in Late Antique Britain’, Early Medieval Europe, 10/​2: 177–​199. Hassall, M. W.  C., and Tomlin, R. S.  O. (1989). ‘Roman Britain in 1988, II:  Inscriptions’, Britannia, 20: 326–​345. Hassall, M. W.  C., and Tomlin, R. S.  O. (2006). ‘Roman Britain in 2005, III:  Inscriptions’, Britannia, 37: 467–​488. Häussler, R. (1998). ‘Resta, viator et lege: Thoughts on the Epigraphic Habit’, Papers from the Institute of Archaeology, 9: 31–​56. Hayward, K. M. J. (2006). ‘A Geological Link between the Facilis Monument at Colchester and the First Century Army Tombstones from the Rhineland Frontier’, Britannia, 37: 359–​363. Hayward, K. M.  J. (2009). Roman Quarrying and Stone Supply on the Periphery—​Southern England:  A  Geological Study of First-​ Century Funerary Monuments and Monumental Architecture. BAR British Series 500. Oxford: Archaeopress. Hill, P. R. (1997). ‘The Maryport Altars:  Some First Thoughts’, in R. J.  A. Wilson (ed.), Roman Maryport and its Setting: Essays in Memory of Michael G. Jarrett. Cumberland and Westmorland Antiquarian and Archaeological Society, 92–​104. Hilts, C. (2011). ‘Jupiter, Best and Greatest—​ Revisiting Maryport’s Ritual Pits’. Current Archaeology, 259 (accessed 24 September 2013). Holder, N. (2007). ‘Mapping the Roman Inscriptions of London’, Britannia, 38: 13–​34. Hope, V. M. (1997). ‘Words and Pictures: The Interpretation of Romano-​British Tombstones’, Britannia, 28: 245–​258. Hope, V. M. (1998). ‘Negotiating Identity and Status:  The Gladiators of Roman Nîmes’, in J. Berry and R. Laurence (eds), Cultural Identity in the Roman Empire. London and New York: Routledge, 179–​195. Hope, V. M. (2000a). ‘Status and Identity in the Roman World’, in J. Huskinson (ed.), Experiencing Rome:  Culture, Identity and Power in the Roman Empire. London and New York: Routledge, 125–​152. Hope, V. M. (2000b). ‘Inscription and Sculpture:  The Construction of Identity in the Military Tombstones of Roman Mainz’, in G. Oliver (ed.), The Epigraphy of Death. Liverpool: Liverpool University Press, 155–​186. Hope, V. M. (2001). Constructing Identity: The Roman Funerary Monuments of Aquileia, Mainz and Nimes. BAR International Series 960. Oxford: Archaeopress. Hope, V. M. (2003). ‘Remembering Rome: Memory, Funerary Monuments and the Roman Soldier’, in H. Williams (ed.), Archaeologies of Remembrance: Death and Memory in Past Societies. New York: Kluwer/​Plenum, 113–​140. Hope, V. M. (2007). ‘Age and the Roman Army: The Evidence of Tombstones’, in M. Harlow and R. Laurence (eds), Age and Ageing in the Roman Empire. JRA Supplementary Series 65. Portsmouth, RI: Journal of Roman Archaeology, 111–​130. Hopkins, K. (1966). ‘On the Probable Age Structure of the Roman Population’, Population Studies, 20: 245–​264. Isserlin, R. M. J. (1998). ‘A Spirit of Improvement: Marble and the Culture of Roman Britain’, in R. Laurence and J. Berry (eds), Cultural Identity in the Roman Empire. London and New York: Routledge, 125–​178. James, S. (1999). ‘The Community of Soldiers: A Major Identity and Centre of Power in the Roman Empire’, in P. Barker, C. Forcey, S. Jundi, and R. Witcher (eds), TRAC 98: Proceedings

Inscriptions and Identity    301 of the Theoretical Roman Archaeology Conference, Leicester 1998. Oxford:  Oxbow Books, 14–​25. Jones, B., and Mattingly D. (1990). An Atlas of Roman Britain. Oxford: Blackwells. Kampen, N. (2006). ‘The Art of Soldiers on a Roman Frontier: Style and the Antonine Wall’, in E. D’Ambra and G. Métraux (eds), The Art of Citizens, Soldiers and Freedmen in the Roman World. BAR International Series 1526. Oxford: Archaeopress, 125–​134. Keppie, L. (1998). Roman Inscribed and Sculptured Stones in the Hunterian Museum, University of Glasgow. Britannia Monograph Series 13. London: Britannia. Kleiner, D. E E. (1977). Roman Group Portraiture: The Funerary Reliefs of the Late Republic and Early Empire. New York: Garland. Knight, J. K. (2010). ‘An Inscription from Bavai and the Fifth Century Christian Epigraphy of Britain’, Britannia, 41: 283–​292. Kockel, V. (1993). Porträtreliefs Stadtrömischer Grabbauten. Mainz: Zabern. Leach, S., Eckardt, H., Chenery, C., Müldner, G., and Lewis, M. (2010). ‘A Lady of York: Migration, Ethnicity and Identity in Roman Britain’, Antiquity, 84: 131–​145. MacMullen, R. (1982). ‘The Epigraphic Habit’, American Journal of Philology, 103: 234–​246. MacMullen, R. (1984). ‘The Legion as a Society’, Historia, 33: 440–​456. Mann, J. (1985). ‘Epigraphic Consciousness’, Journal of Roman Studies, 75: 204–​206 Mattingly, D. (2004). ‘Being Roman: Expressing Identity in a Provincial Setting’, Journal of Roman Archaeology, 17: 5–​25. Mattingly, D. (2006). Imperialism, Power and Identity:  Experiencing the Roman Empire. Princeton: Princeton University Press. Mattingly, D. (2008). ‘Urbanism, Epigraphy and Identity in the Towns of Britain under Roman Rule’, in V. E. Hirschmann, A. Krieckhaus, and H. M. Schellenberg (eds), A Roman Miscellenay:  Essays in Honour of R.  Birley on his Seventieth Birthday. Gdańsk: Akanthina, 53–​7 1. Meyer, E. (1990). ‘Explaining the Epigraphic Habit in the Roman Empire: The Evidence of Epitaphs’, Journal of Roman Studies, 80: 74–​96. Meyer, E. (2011). ‘Epigraphy and Communication’, in M. Peachin (ed.) The Oxford Handbook of Social Relations in the Roman World. Oxford: Oxford University Press, 191–​226. Millett, M. (1990). The Romanization of Britain:  An Essay in Archaeological Interpretation. Cambridge: Cambridge University Press. Mouritsen, H. (2005). ‘Freedmen and Decurions:  Epitaphs and Social History in Imperial Italy’, Journal of Roman Studies, 95: 38–​63. Mrozek, S. (1973). ‘A propos de la repartition chronologique des inscriptions Latines dans le haut-​Empire’, Epigraphica, 35: 113–​118. Mullen, A. (2012). ‘ Introduction: Multiple Languages, Multiple Identities’, in A. Mullen and P. James (eds), Multilingualism in the Graeco-​ Roman Worlds. Cambridge:  Cambridge University Press, 1–​35. Noy, D. (2000). Foreigners at Rome: Citizens and Strangers. London: Duckworth. Noy, D. (2010). ‘Epigraphic Evidence for Immigrants at Rome and in Roman Britain’, in H. Eckardt (ed.), Roman Diasporas:  Archaeological Approaches to Mobility and Diversity in the Roman Empire. JRA Supplementary Series 78. Portsmouth, RI:  Journal of Roman Archaeology, 13–​26. Parsons, D. N. (2011). ‘Sabrina in the Thorns: Place Names as Evidence for British and Latin in Roman Britain’, Transactions of the Philological Society, 109/​2: 113–​137. Pearce, J. (2004). ‘Archaeology, Writing Tablets and Literacy in Roman Britain’, Gallia, 61: 43–​51.

302

302   Valerie M. Hope Pearce, J. (2011). ‘Marking the Dead: Tombs and Topography in the Roman Provinces’, in M. Carroll and J. Rempel (eds), Living through the Dead: Burial and Commemoration in the Classical World. Oxford: Oxbow Books, 134–​158. Phang, S. (2001). The Marriage of Roman Soldiers (13 bc–​ad 235). Leiden: Brill. Phillips, E. J. (1976). ‘A Workshop of Roman Sculptors at Carlisle’, Britannia, 7: 101–​108. Raybould, M. (1999). A Study of the Inscribed Material from Roman Britain. BAR British Series 281. Oxford: Archaeopress. Revell, L. (2007). ‘Architecture, Power and Politics: The Forum-​Basilica in Roman Britain’, in J. Sofaer (ed.), Material Identities. Oxford: Blackwell, 127–​151. Revell, L. (2009). Roman Imperialism and Local Identities. Cambridge:  Cambridge University Press. Richmond, I., and Wright, R. (1956). Catalogue of Roman Inscribed and Sculptured Stones in the Grosvenor Museum Chester. Chester: Chester and North Wales Archaeological Society. Russell, B. (2010). ‘Sarcophagi in Roman Britain’, Bollettino di Archeologia on line. Volume Speciale:  XVII International Congress of Classical Archaeology, Roma 22–​26 September 2008  (accessed 26 June 2012). Saller, R. (1994). Patriarchy, Property and Death in the Roman Family. Cambridge: Cambridge University Press. Saller, R., and Shaw, B. (1984). ‘Tombstones and Roman Family Relations in the Principate: Civilians, Soldiers and Slaves’, Journal of Roman Studies, 74: 124–​156. Scheidel, W. (2012). ‘Epigraphy and Demography:  Birth, Marriage, Family and Death’, in J. Davies and J. Wilkes (eds), Epigraphy and the Historical Sciences. Proceedings of the British Academy, 177. Oxford: Oxford University Press, 101–​29. Stewart, P. (2009). ‘Totenmahl Reliefs in the Northern Provinces: A Case-​Study in Imperial Sculpture’, Journal of Roman Archaeology, 22: 253–​274. Stewart, P. (2010). ‘Geographies of Provincialism in Roman Sculpture’, Journal of the International Association of Research Institutes in the History of Art 0005 (27/​7/​2010)  (accessed 26 June 2012). Strück, M. (2000). ‘High Status Burials in Roman Britain (First–​Third Centuries ad)—​ Potential of Interpretation’, in J. Pearce, M. Millet, and M. Strück (eds), Burial, Society and Context in the Roman World. Oxford: Oxbow Books, 85–​96. Tobler, G. W., and Adams R. (2007). Romano-​British Tombstones between the First and Third Centuries:  Epigraphy, Gender and Familial Relations. BAR British Series 437. Oxford: Archaeopress. Tolia-​Kelly, D. P. (2010). ‘Narrating the Postcolonial Landscape:  Archaeologies of Race at Hadrian’s Wall’, Transactions of the Institute of British Geographers, 36: 71–​88. Tomlin, R. S. O. (2011). ‘Writing and Communication’, in L. Allason-​Jones (ed.), Artefacts in Roman Britain: Their Purpose and Use. Cambridge: Cambridge University Press, 133–​152. Vanderspoel, J. (2005). ‘Male Deities in Northern Roman Britain: Patterns and Perspectives’, Religious Studies and Theology, 24/​1: 33–​59. Woolf, G. (1996). ‘Monumental Writing and the Expansion of Roman Society in the Early Roman Empire’, Journal of Roman Studies, 86: 22–​39. Woolf, G. (1998). Becoming Roman:  The Origins of Provincial Civilization in Gaul. Cambridge: Cambridge University Press.

Chapter 15

Ideas of Ch i l dh o od in Roman Bri ta i n The Bioarchaeological and Material Evidence Rebecca Gowland

Introduction Once overlooked, ancient children are now subjects of overviews. But—​and here we are less fortunate—​it is some 2,000 years too late to learn very much about them. (Golden 2011: 262)

Childhood in the past is a burgeoning field of study and has received a great deal of attention from scholars of the ancient world, including Golden (2011: 262) quoted at the head of this chapter. As a result of a focus on childhood within the discipline of history (e.g. Ariès 1962; Stone 1977; Pollock 1983), it is now widely recognized that perceptions of childhood do not subscribe to a universal reality, but instead are contingent upon historical and cultural context. The seminal book by Ariès, Centuries of Childhood (1962), led to a shift in the concept of the family from that of a fixed biological entity to a historically contingent construct (James et al. 1998). Within archaeology, feminist and gender discourse, together with the influence of studies of childhood within history, have acted as a stimulus to debates concerning children in narratives of the past. Initial studies of childhood within archaeology were concerned with children’s ‘invisibility’ in archaeological discourse. The neglect of children, it was argued, stemmed from the same anthrocentric biases that had previously served to marginalize women (Lillehammer 1989; Baker 1997): children have been both absent/​invisible from the archaeological record, and invisible, unknowable at the conceptual level. Contemporary culturally constructed social

304

304   Rebecca Gowland knowledge, embedded as it is in masculist ideologies, fits ‘children’ in the interpretative framework as incomplete humans, that is, not male/​masculine. (Baker 1997: 187)

Since this important initial research, studies of childhood have developed towards a consideration of the culturally specific constructions of childhood within different time periods and places (e.g. Sofaer Derevenski 1994; Moore and Scott 1997; Gowland 2001; Halcrow and Tayles 2008). A useful way of conceptualizing childhood was described within the discipline as follows: Childhood provides an interpretive frame for contextualising the early years of human life. Childhood, as distinct from biological immaturity, is neither a natural or universal feature of human groups but appears a specific structural and cultural component of many societies. (Prout and James 1997: 8)

Furthermore, studies of the past started to recognize that children are active agents within society, rather than passive beings imprinted by the socializing forces of adults (Prout and James 1997). The importance of children as a focus of study in their own right is now acknowledged by numerous scholars and is reflected by the development of the journal Childhood in the Past. Studies have focused on childhood not as a separate or distinct entity, but a construct that must be analysed and understood within the chronology of the life course as a whole. Age is now conceptualized as a central aspect of social identity and a highly significant structuring element within society. A life-​course perspective has been adopted by most social scientists in their study of age identity. The life-​course approach differs from previous approaches to age identity in that it considers the fluidity of identity throughout the life of an individual from birth to death (see A. Moore, this volume). This chapter provides an overview of these debates in relation to the archaeological evidence for ideas of childhood in Roman Britain. In particular, this chapter will focus on the bioarchaeological and funerary evidence. Before we discuss the various strands of archaeological data, it is worth first outlining what classical sources had to say about childhood in the Roman Empire in order to provide some context.

Childhood in the Roman World: Classical Perspectives Since the 1990s there has been a considerable amount of research on childhood and the family in the Roman world, culminating in numerous books and edited volumes (e.g. Dixon 1992; Rawson and Weaver 1997; Harlow and Laurence 2002; Rawson 2003, 2011; Dasen and Späth 2010). This research is almost exclusively derived from a historical or classical perspective, drawing upon literary sources, medical texts (especially Soranus’ Gynaecology), as well as epigraphic and monumental evidence (e.g. Laurence

Ideas of Childhood in Roman Britain    305 2000; Harlow and Laurence 2002). This work has been important for understanding Roman perceptions of children. The aforementioned studies have demonstrated that childhood was recognized as a distinctive stage of the life course in the Roman world, and historical texts from this period allude to the charm of childish characteristics (Dixon 1992). Iconographic evidence likewise depicts children learning and at play—​ activities that we would also regard as particularly childlike. The demarcation of this developmental period into a series of distinct stages is also evident from historical documentation. The term infantia, which literally means ‘not speaking’, was ascribed to children until the age of 7 years. From then, terminology differentiated between the sexes; puer or puella were employed, and males and females experienced divergent life-​course trajectories in terms of social age transitions. Virginity is, for the first time, attributed as a characteristic after the age of 7 years (Fraschetti 1997). Roman males underwent a significant rite of passage at approximately 14–​16 years of age, during a ceremony that took place in both public and private; they replaced their toga praetexta with the toga virilis and removed their bulla (Eyben 1993: 6; Fraschetti 1997: 64). This event signified a new social age for males, who were then considered ‘more responsible’ individuals as adolescens, until approximately 25–​30 years of age (Weidemann 1989: 116). Females have no similar rite of passage; the onset of menarche does not appear to have been socially significant, and only upon marriage did they experience a change in status (Fraschetti 1997:  63). Epigraphic and documentary evidence indicates that high-​status females in Rome may have married as early as 12 years of age—​puberty apparently not being a prerequisite for marriage (Hopkins 1965). However, in a re-​ examination of the epigraphic evidence from Rome and the wider empire, Shaw (1987) convincingly demonstrated that, for the majority of Roman women, marriage tended to occur from the late teens to early twenties and for males in the mid-​twenties. Shaw (1987: 33) argues that those very few historical sources that discuss the age of marriage for women tend to refer only to the ‘narrowest of elites’ and are thus largely irrelevant for most of the population. The great majority of historical sources pertaining to Roman childhood are biased towards Rome and Italy. As a society’s life course is culturally constructed, it is subject to temporal and spatial differences, which have been observed in studies of funerary data from the Roman Empire (e.g. Gowland 2001; Revell 2005). Therefore, the relevance of historical sources from Rome for Roman Britain may be regarded as questionable. In the absence of such a rich corpus of historical data from Britain, we must turn to the archaeological evidence. As highlighted above, one particularly fruitful source of information is the funerary context. This context is unique, because it provides a crucial link between individuals and material culture (Gowland and Knüsel 2006) and thus one can infer aspects of age-​related social identity through patterns of deposition and ritual treatment. When interpreting funerary evidence, one must be cautious not to be overly simplistic; we are after all dealing with the treatment of the dead, not the living, and glimpsing past identities through the distorting lens of ritual practice. Nevertheless, the funerary context provides a key reservoir of data for understanding perceptions of

306

306   Rebecca Gowland childhood in Roman Britain. It is not only the mode of burial that is of significance, but also the skeletal remains. The bioarchaeological evidence has been an under-​exploited resource for examining perceptions of childhood in Roman Britain, though research over the last ten years has been highlighting its pivotal importance.

The Bioarchaeology of Roman Childhood The bioarchaeological analysis of children provides crucial direct evidence for their well-​being in relation to their social and physical milieu. The potential for this biological evidence to yield significant social information about the Roman family is immense and yet it is under-​utilized. This is a consequence of sub-​and interdisciplinary boundaries and a lack of communication across these—​a situation that is thankfully now changing (see Redfern and Gowland 2012 for a discussion). The conceptualization of childhood as a social construction may also have contributed to the marginalization of their physical remains. Within such a schema, the body as a physical entity tends to be perceived as irrelevant (Shilling 1993). However, while societies may construct their own perceptions of the life course, one cannot overlook the corporeal aspects of childhood: the physical and emotional changes that accompany growth and the attainment of bodily maturation. An infant, for example, is helplessly dependent on adults for care and has basic needs that must be met if he or she is to survive. As Franks (1991) writes, our bodies are ‘an obdurate fact’ and as such should be included in narratives of childhood in the past. James and colleagues (1998: 51) also discuss the fact that ‘childhood is united by the universal biology of human physical development and cognitive potential but, in the same moment, radically differentiated by the varied social contexts in which this growth can be culturally enacted in the life course’. The interaction between the physical and the social worlds in the forging of identities will be discussed later in the chapter. below.

Age and Sex One of the key limitations of the skeletal analysis of non-​adults is the inability of osteological techniques to provide a reliable determination of biological sex. While numerous studies of non-​adult remains have attempted to provide methodological sexing criteria, unfortunately these have not performed sufficiently well during independent tests. In the future it is possible that ancient DNA analysis will become more affordable and accessible, in which case this problem may be circumvented. At present, however, osteoarchaeologists rarely attempt to estimate the sex of non-​adults (10 >20

N 0

100

200 km

Figure 30.1  Distribution of inscribed votive altars from Roman Britain. Source: after Millett (1995a: 110). © Amy Zoll.

624

624   Amy Zoll inscribed altars are poorly attested at most southern urban sites. Access to resources may be partly to blame, although adequate stone was available to these same civilian centres for large-​scale building programmes, statuary, and other types of inscription. Nor are lower rates of literacy or a weakly internalized epigraphic habit sufficient explanation, as some inhabitants of southern Britain could, and did, employ the written word in their interactions with the gods in other materials. Differences in choice of offering may instead relate to dedicator identity and the roles religious institutions played in the communities of the north and south. Votive altars are not the only remnants of religious practice irregularly distributed across Roman Britain. Mattingly (2006:  520–​522) compares the relative frequencies of several categories of material culture related to ritual and finds some artefacts and features to be more common among urban and rural settlements of the civilian-​ administered south, while being scarce in military areas. Among text-​bearing objects, votive leaves and curse tablets have been unearthed in greater numbers in civilian contexts. Votive altars also show little geographic overlap with known temple sites (Figure 30.2), particularly those of the Romano-​Celtic type, a form thought to pre-​ date the Claudian conquest (King 2007). Some notable exceptions occur in the south-​ west, where the large temple complex at Bath has produced altars, curse tablets, and inscribed offerings dedicated to the resident goddess, Sulis Minerva, and other deities (Cunliffe 1988). Improved recording of stray finds has shed new light on the frequency and distribution of non-​monumental religious objects, including the regional dispersal of numerous rings bearing the cryptic inscription ‘tot’. The discovery of a specimen with an expanded legend (deo tota) confirms these items of personal adornment refer to a deity, possibly the god Toutatis (Daubney 2010). Dated stylistically to the second and third centuries ad, the rings were produced mainly in silver, with a few in gold or copper alloy. Their find-​spots conform to the civitas of the Corieltauvi in the east Midlands, centring on the territorium of the Roman colony at Lincoln, although most come from rural contexts. This region is poorly served by evidence for either temples or altars, but the scattering of tot rings illustrates an alternate means of expressing religious knowledge and a shared sense of cult embodied by the wearer, rather than rooted in a location (although these items may themselves signify pilgrimage to a sacred site). Inscribed objects pertaining to religion and ritual are notoriously difficult to date, most having been recovered from secondary contexts. Like the tot rings, the majority of British defixiones are either unstratified or redeposited finds, while votive altars—​ composed of solid and handily dressed blocks of stone—​were frequently incorporated into later structures. Despite their sacred function, these monuments were not immune to reuse even in antiquity, as evidenced by the series of altars to Jupiter Optimus Maximus (Best and Greatest), chief god of Rome, found buried in pits near the fort at Maryport, now interpreted as the repurposed footings for a massive timber building, possibly dating to the late fourth century ad (Haynes and Wilmott 2012). The coarse chronologies for religious inscriptions hinder the ability to trace the development of the practices they represent. The one major text known to come from the

Altars Temples tot Rings Hadrian’s Wall

N 0

100

200 km

Figure 30.2  Main concentrations of votive altars. Source: after Millett (1995a: 110); temples after Millett (1995a: 112); tot rings after Daubney (2010: 112). © Amy Zoll.

626

626   Amy Zoll first century ad is the famed Purbeck marble plaque from Chichester (RIB 91), which records the dedication of a temple to Neptune and Minerva, authorized by a British client king for the welfare of the emperor’s domus divina (divine house). This example is extraordinary not only for its early date, but also for its depiction of a leading member of the native aristocracy involved in the promotion of the gods of Rome and the Imperial Cult. Local elites in the provinces, as at Rome, are thought to have patronized civic and private cults and vied for priestly offices as a means of affirming and enhancing their prestige within the imperial power structure (Gordon 1990; Woolf 1997; Derks 1998). In Britain, however, epigraphic evidence for the participation of high-​ranking Britons and civilian authorities in religious institutions is sparse, but this may be less an indication of an ambivalent attitude towards socio-​political competitiveness than a reflection of the types of inscriptions being produced, the rituals they commemorate, and the individuals and groups responsible for their creation.

Vows and Curses The majority of inscribed offerings from Britain were votive in the true sense: presented to the gods not to placate or elicit future favours, but as payment for services requested and rendered. The vow was a central tenet of Roman religion, articulating the reciprocal relationship between mortals and immortals through ritualized exchange (Scheid 2003: 101–​105). At Rome, public vows were formally pronounced by officials and priests to ensure the well-​being of the emperor and the Roman people at regular intervals and at times of national crisis or insecurity. Sacrifices, games, temples, statues, and valuable gifts were pledged in return for divine protection. Private individuals undertook similar vows on a much reduced scale when confronted with an uncertain fate owing to a grave illness, perilous journey, risky venture, or other anxiety. While vows and curses differed in intent—​one to solicit aid, the other to inflict harm—​ both represent specific and finite transactions between gods and mortals that followed similar ritual structures whereby the obligations and expectations of both parties were made explicit. In the case of personal vows, little survives of the initial declaration of terms (nuncupatio), which may have been recorded on sealed wooden tablets stored in the cult temple (Derks 1998: 224–​231). The sacrificial rites and offerings promised were deliverable only upon a satisfactory result, to be presented at the ceremonial closure of the vow (solutio). The legends on votive altars and other inscribed gifts clearly state to which god they were owed and by whom, often concluding with the phrase ‘votum solvit libens merito’ (willingly and deservedly fulfilled a vow), indicating the gratitude of the dedicator at the completion of the compact. These transitory performances were reproduced through the monuments and inscriptions left to adorn the shrines, advising visitors of the correct names and rites by which to approach the deity and serving as lasting testaments to the power of the god and success of the vow.

Names of Gods   627 Defixiones represent a different kind of pact, in which the conditions have been preserved, but the outcome remains a mystery. Continental examples of engraved curses sought to exercise supernatural influence over a variety of situations. By contrast, British curse tablets are preoccupied with theft (Tomlin 1988); their authors’ appeals likened to a form of ‘judicial prayer’ (Versnel 1991; Adams 2006). The creation of a tablet took on the role of nuncupatio, dedicating the stolen property (or the thief him or herself) in absentia to the god, who was exhorted to torment the perpetrator—​known or unknown—​ until the missing items were returned to the shrine. For this service, the deity would receive a portion of the recovered goods in reward (solutio). The details of these agreements were concealed in tightly rolled strips of lead alloy, affixed to the sanctuary walls or base of the cult statue, or cast into sacred waters. No doubt, the prospect of divine justice held a certain psychological appeal, particularly among those who may have felt disenfranchised from the terrestrial legal system (Kiernan 2004). The longevity of the practice attests to its perceived efficacy. Although the physical and textual remains of vows and curses are readily identifiable, these rites were responses to specific circumstances, rather than prescribed parts of routine religious observance. They were by no means the only ways to communicate with the gods, or indeed the most common. The majority of vows were fulfilled with consumable sacrifices and perishable gifts, rather than permanent monuments such as altars. Many more unengraved offerings from countless anonymous donors are found at sacred sites (see also Smith, this volume), but these tend to be overshadowed by text-​bearing objects in scholarly discussion. At least half of the over two dozen stone altars from Coventina’s Well, at Carrawburgh, were uninscribed. Whether unfinished or originally painted, these anepigraphic monuments were less carefully documented than their counterparts with legible dedications to the goddess (Allason-​Jones and McKay 1985: 18). From this same deposit, large quantities of pottery, bone, and more than 13,000 coins demonstrate that Coventina’s worshippers expressed their devotion materially in a variety of ways, not just through votives fashioned and inscribed for the purpose.

Votaries Whittaker (1997) regards religion and politics as inseparable forces for the imposition and encouragement of the Roman world view. Public priests and civilian magistrates were charged with the maintenance of social and spiritual order through the conduct of sanctioned rites on behalf of the citizenry, in recognition of a communal obligation to the divine (religio) (Beard 1991; Scheid 2003: 102). This was Roman religion in its strictest sense: a unifying set of precepts and practices, integral to law and government, set out in municipal charters and official calendars (Ando 2007). All else tends to be relegated to the catchall of ‘private’ cult, and it is from this category that the bulk of epigraphic evidence for religion in Roman Britain derives.

628

628   Amy Zoll Identification of both supplicant and recipient deity seems to have been a requisite component for most votive and defixio texts. The inscribed invocation was a symbolic act, rooting the god’s presence in space and time. The inclusion of the worshipper’s name served a corresponding function:  situating the individual within the wider religious tradition of the community and reinforcing personal and collective relations with the gods (Beard 1991: 46–​48). Unlike epitaphs, however, religious inscriptions were meant to commemorate an event rather than celebrate a life and tend to be short on biographical detail. Even so, some commonalities of participant identity emerge. The names of petitioners on curse tablets differ from those of altar dedicators, including at Bath, where both forms of object are present, possibly denoting the activities of distinct social groups within a shared sacred space or a shift in clientele over time (Tomlin 2002: 171–​172). Authors of defixiones exhibit a higher percentage of Celtic personal names, with few employing the nomenclature of the Roman citizen more commonly found on altars. The small sums and basic possessions listed among the items stolen—​such as articles of clothing, livestock, and agricultural implements—​suggest rural practitioners of modest means. At Uley, the values of missing goods are somewhat greater, indicating that even affluent farmers at times felt the need for divine redress (Adams 2006: 10). Variations in the material and quality of tot rings may also imply bearers of varying status who expressed a common membership through their visible commitment to the same deity. Other small objects, such as inscribed plaques (tabulae ansatae) and votive leaves, found mainly in the south, further demonstrate how the civilian populace adopted and adapted writing in their dealings with the gods. But it was the Roman army that made the most substantial contribution to Britain’s corpus of religious epigraphy (Cepas 1989; Raybould 1999). Soldiers routinely included their unit and rank in inscriptions, attesting to the importance of military affiliation in self-​representation within this group. The army is the most epigraphically recognizable of any profession or other subset of the population in Roman Britain, but even without the abundance of textual clues, the marked distribution of stone inscriptions shows that soldiers and their associates made disproportionate use of this mode of expression. Mattingly (2006: 166) estimates Britain’s ‘community of soldiers’, including auxiliaries and dependent families, merchants, slaves, and freedmen, constituted less than 3 per cent of the total population. Yet, members of the Roman army, acting individually or as units, were responsible for well over half of all extant votive inscriptions on stone from the island. The inordinate number of military dedicators in British religious inscriptions has been noted elsewhere (Zoll 1995a: 136–​137). Research in progress by this author suggests this was not an isolated phenomenon, but part of a wider pattern in which votive inscriptions were one of the most prevalent forms of monumental epigraphy along the empire’s northern limes, in provinces with permanent garrisons and areas of intensive long-​term military activity (Zoll, forthcoming). Soldiers observed their own festival calendars and military rites, but also displayed considerable diversity in belief and practice throughout the empire (see also Haynes, this volume). The profusion of altars set up by units and their commanders may represent an alternative to elite-​administered and urban-​focused civic worship, with its

Names of Gods   629 official priesthoods, euergetism, and monumental temple architecture. Rather than symbolizing the export of Roman religion, unchanged, to the edges of the empire, the presence of inscribed altars in large quantities may instead signify its adaptation in response to regional differences in social structure and administration.

Gods of Empire, Elsewhere, People, and Place Convenient labels—​military or civilian, official or unofficial, public or private—​applied to the evidence for religion have long exerted an undue influence on its interpretation (Millett 1995b; Goldberg 2009a). Simple classificatory schemes fail to capture the wide spectrum of ritual and belief, but are especially pervasive in the characterization of provincial pantheons, where gods are categorized as ‘Roman’ or ‘native’, or some measurable combination of the two. Eastern cults are typically segregated from discussion owing to the perceived ‘otherness’ of the rites, philosophies, and participant experience. The majority of practitioners were unlikely to have made such clear distinctions, however. Treatises on religio by ancient authors and repeated attempts by lawmakers to codify its correct observance show these definitions were not fixed but required periodic action and adjustment when confronted with different gods and practices within the empire’s borders (Ando 2007; 2008: 1–​18). Even within the army, ‘official’ customs varied. Given the high percentage of military dedicators represented in Britannia’s epigraphic assemblage, it is unsurprising that Jupiter Optimus Maximus, supreme protector of Rome and the empire, is by far the most frequently named deity on votive altars, with well over 100 surviving examples. According to Whittaker (1997: 151–​152), Jupiter had become synonymous with the city, imperial law, and the political order by the second century ad. His monuments occur in multiples at forts along Hadrian’s Wall and related northern defences, set up primarily by auxiliary cohorts and their commanding officers. In most instances, Jupiter Optimus Maximus (often abbreviated IOM) is addressed alone or with the divine authority of the emperor (numen Augusti). These altars are usually ascribed to a military version of the annual public vota ceremony performed at Rome for the welfare of the emperor and Roman people, which also served as an occasion for the renewal of service oaths by the army (Stoll 2007: 462). Few votive altars have been found associated with Britain’s three permanent legionary bases at York, Chester, and Caerleon, but comparable series of IOM monuments from other provinces were erected on different dates to mark the anniversary of the formation of units under their standards (e.g. CIL II. 2552–​2553) and the tenures of governors’ adjutants (beneficiarii consulares) at military outposts (Mirković 1989). Despite Jupiter Optimus Maximus’ close association with the Roman state, his identity is not always so clear cut. His sole appearance on a British defixio from

630

630   Amy Zoll Ratcliffe-​on-​Soar, Nottinghamshire (Turner 1963), seems to confirm a hypothesis by Versnel (1991: 90) that authors of curses tended to avoid the ‘great gods’, opting instead to make requests of deities with strong local ties who might be more sympathetic to their plight. However, Jupiter Optimus Maximus is also the most common recipient of votive altars in parts of Gaul and north-​west Hispania, where he is seen by some to disguise an indigenous sky or mountain god (Green 1989: 74–​130; Richert 2005: 3–​6). Jupiter’s conflation with a local sky god from Doliche, in southern Turkey, gained a following across the Roman world, especially among soldiers. Inscriptions and cult images of Jupiter Optimus Maximus Dolichenus have been found at several sites at Hadrian’s Wall, including the unexpected discovery of altars and a shrine to this ‘non-​traditional’ deity within the walls of Vindolanda fort (Birley and Birley 2010: 25)—​an honour once thought reserved for sanctioned divinities of the empire and the army. Just as the definition of ‘Roman’ blurs upon closer examination, so too does that of ‘native’. The Matres were frequent subjects of votive texts and depictions in stone. Similar trios of mother goddesses are well known from the continent, but there were emphatically linked to a particular place or group by use of localized epithets (Derks 1998: 119–​130; Woolf 2001). When the type of Matres is specified in Britain, however, they are usually of a more inclusive variety and often allude to imported origins for both dedicators and deities, such as the Matres Domesticae (of the household or homeland (RIB 652, 2025, 3210)), Tra(ns)marinae (from overseas (RIB 919, 1224, 1318)), or Ollototae (a Celtic descriptor translated as ‘from other peoples’ or ‘foreign-​born’ (RIB 574, 1030; AE 2009, 738)). Two small votive monuments dedicated to the assorted Matres of Italy, Gaul, Germania, Africa, and Britannia by military personnel also reflect the mixed make-​up of the communities in which they were erected (RIB 88, 653). Most gods were approached in locations—​both built and natural—​where their power was seen to be most manifest. The placement of altars, niches, and small shrines (aediculae) in secular and domestic settings shows that temples were not the only venues for encounters with the divine. Soldiers frequently called upon the protection of the goddess Fortuna in the baths (RIB 730, 1212, 2146), while her more judgemental counterpart, Nemesis, was sought in amphitheatres (RIB 3149). Sacred springs were home to the Nymphs (RIB 460, 3316) and sometimes presided over by named goddesses, such as Sulis Minerva at Bath or Coventina at Carrawburgh, whose influence does not appear to have extended much beyond their watery sanctuaries. When the name of the resident divinity was not forthcoming, the spirit of the place (genius loci) might be invoked (e.g. RIB 450, 1984, 3195). A few gods seem to have held a wider authority, with distributions tying them to the social and physical geography of a region (Figure 30.3). Belatucadrus is attested from some twenty-​nine altars confined to the western end of Hadrian’s Wall in northern Cumbria, an area associated with the civitas of the Carvetii. The god Cocidius is known from twenty-​three altars, mainly from the mid-​section of the wall and a few forward installations. He is also schematically depicted in panoply on two silver votive plaques (RIB 986, 987) from Bewcastle, believed to be the site of a shrine to the god, from which

Names of Gods   631 Cocidius 1 2 5 " Mars Cocidius Silvanus Cocidius # Vernostonus Cocidius Hadrian’s Wall

" "

"

"

#

Belatucadrus 1 2–3 7 Mars Belatucadrus Hadrian’s Wall

Vitiris/Veteres 1 2–3 4–6 >10 Moguns Vitiris Hadrian’s Wall

N

0

25

50 km

Figure 30.3  Distribution of votive inscriptions dedicated to the gods Cocidius, Belatucadrus, and Vitiris/​Veteres, including double-​named variants. Source: after Zoll (1995a). © Amy Zoll.

it took the name Fanum Cocidi. To the east, more than sixty altars were set up to the enigmatic deity Vitiris, who is portrayed in a third of examples as the plural Veteres, and in two instances from Chester-​le-​Street, as female (RIB 1047, 1048).

632

632   Amy Zoll Vitiris is second to Jupiter in number of extant dedications in Britain, with Mars his only close rival. The distribution of Vitiris relates to the main eastern watershed of northern England, and he is most likely a local divinity, despite some seemingly Germanic renditions of his name (Goldberg 2009b). Falling to opposite sides of the Pennines, altars to Vitiris and Belatucadrus are consistently among the smallest in Britain and most are crudely carved. Both suffer from rampant spelling inconsistencies, suggesting the oral transmission of names and a lack of central authority over their worship. Devotees of Belatucadrus and Vitiris generally go by only one name, and few mention a military association, but many stones are either too small or incomplete to include much dedicator information. In contrast to these followers of low or no rank, Cocidius attracted a higher-​status clientele—​mostly officers and legionary units—​and his monuments tend to be larger and more skilfully crafted. It is difficult to ascertain just how the attentions of soldiers contributed to the recognition and perpetuation of non-​classical deities. Many local theonyms survive in stone as a result of epigraphic customs among the military, but this was a selective process, subject to the concerns of the inscribing community (Zoll 1995a). Among the classical gods recorded in votive texts from Britain, Juno appears at most twice (RIB 3460, 3505), despite her elevated position as a member of Rome’s Capitoline Triad alongside Jupiter Optimus Maximus. Venus was a popular subject in Romano-​British decorative arts, but no written dedications to the goddess are known from the province. The many mass-​produced pipe-​clay votive figurines modelled in her likeness are seen to represent an alternate, popular, and indigenous set of beliefs that operated outside classical convention (J. Webster 1997: 332–​334), but there is no assurance that practitioners saw any inherent contradiction in their actions. Uninscribed offerings reflect different trajectories of worship by segments of Romano-​British society for whom epigraphic commemoration may have been inappropriate or unavailable. Women are poorly represented in Britannia’s votive epigraphy on stone, comprising just 3 per cent of known altar dedicators. In other provinces, such as Hispania Citerior, they are better attested, particularly in areas without a significant military presence (Zoll, forthcoming). Women participated in the enacting of curses at Bath and Uley, although not to the same frequency as men. Three women are, however, among the eight devotees identified on gold and silver votive leaves and other objects donated to Senuna at her shrine near Ashwell, Hertfordshire (Jackson and Burleigh 2007). While monumental epigraphy does not appear to have been the custom at this site, worshippers still engaged the local goddess through vows and discharged their obligations in the accepted fashion of that community.

By Any Other Name … Much scholarly attention has been lavished upon gods with non-​classical names, whether in an attempt to decode earlier beliefs or as an indicator of regional ethnicity.

Names of Gods   633 Deities with standard Roman theonyms often recede into the background, assumed to be wholesale introductions, uniformly transmitted and received. That Roman and non-​ Roman divinities were invoked side by side on the same monument and even as the same entity reveals the innovative ways in which local pantheons developed in response to practitioner need and identity. Appeals were made to those higher powers deemed most likely to respond, based on prior experience and evidenced by the votive gifts amassed at sacred sites. How these divinities came to be worshipped in this manner, and by whom, have been central to the debate over religious change in the Roman west. Many arguments hinge upon the epigraphic phenomenon of double naming, in which a Roman theonym was linked with that of a local god and addressed as a single being. Such dual deities are not unique to Britannia and comprise only a small percentage of god names from the island (Zoll 1995b). They have, nonetheless, been imbued with special significance as intermediaries on the path to assimilation: Tacitus’ interpretatio Romana personified (Germania 43). The equation of autochthonal deities with Roman counterparts is seen by Henig (1984: 55–​67) as the predictable result of incomers wishing to venerate local divinities in familiar forms, terms, and rites, setting the example for leading members of native society to follow. For Webster (1995), these pairings were products of an asymmetric colonial discourse: deliberate superimpositions, privileging one set of beliefs over another and inciting subject peoples to react to the forcible conversion of their gods with overt rejection or subtler forms of resistance. Derks (1998: 91–​118), however, disputes the characterization of such hybridized entities as instruments of cultural imperialism, instead viewing the selectivity of deity pairings in Gaul as a reflection of a native understanding and application of the Roman pantheon to further their own aims. Rather than being transitional and temporary, many twinned deities show remarkable resilience and portability. At Bath, engraved entreaties to Sulis Minerva, arguably the most famous doubly named god, occur on stone from the mid-​second century ad and continue in curse texts into the fourth century, appearing alongside similar examples addressed solely to Sulis, but never to Minerva alone. Spoons from the late fourth century ad Thetford hoard bear the names of a series of little-​known divinities; in some instances coupled with the god Faunus and in others singly, but Faunus himself does not appear unaccompanied (Johns 1986). Non-​classical bynames were more than mere descriptive epithets; they were important for proper identification and observance. Many deities so named retained their distinct appellations and character far from their home territories. Well-​travelled gods such as Hercules Magusanus (RIB 2140), Mars Camulus (RIB 2166, 3014), Loucetius Mars (RIB 140), and Mars Lenus (RIB 309) were probably already paired when they were brought to Britain from the European mainland, where they had established followings. Even Mars Toutatis may have been imported in this form; the two plaques bearing his name (RIB 219 and AE 2001, 1298) were found in very disparate locations, seemingly unrelated to the well-​defined distribution of tot rings. For theonyms exclusive to Britain, Mars occurs in the majority of pairings, followed (to a much lesser extent) by Apollo and Silvanus. Other classical deities commonly

634

634   Amy Zoll involved in similar relationships on the continent, such as Mercury and Hercules, are barely represented. Mars’ appeal among the military communities of northern Britain is understandable given the god’s martial qualities, but Henig (1984: 50–​55) attributes his popularity in civilian contexts to the polyvalent and composite nature of most classical divinities. Derks (1998: 94–​107) makes a cogent argument for why Roman deities such as Mars and Hercules were affiliated with local gods along the Lower Rhine, there cast as guardians of field or flock based on regional ecology and subsistence practices. Webster (1995: 159–​160) posits that some segments of the population, such as native Britons and low-​ranking soldiers, did not engage in name-​twinning to the degree of higher-​status devotees. This seems to be borne out in the case of Apollo Maponus (RIB 1120–​1122, 2063), but in other examples where dedications to the same paired and unpaired deity survive, as with the numerous inscriptions to Sulis, Cocidius, and Belatucadrus, little distinguishes votaries employing one or the other naming variant. Nor do the votive gifts differ markedly in terms of object type, size, or distribution. Webster (2003: 49) also contends that some deities were more resistant than others to epigraphic twinning. None of the many votive dedications to Vitiris/​Veteres conflate the god with a Roman deity, but he is once paired with Mogons, an obscure non-​Roman figure, some distance from Vitiris’ main territory (RIB 971). Coventina, too, seems to have possessed no named counterpart, but is twice referred to as ‘nymph’ (RIB 1526, 1527) and depicted as one in relief (RIB 1534), allying her with those widely worshipped water goddesses. Visual cues can present a very different understanding of a deity from the textual. Minerva might have been mistaken for the beneficiary of the votive leaves from Ashwell on which her image appears, if not for the embossed legends identifying the hitherto unattested goddess Senuna as the intended recipient (Jackson and Burleigh 2007). Certain classical divinities also seem less receptive to twinning than others. Although often invoked in his Dolichene aspect (RIB 1022, 1131, 1725), Jupiter Optimus Maximus is only once paired with an apparent western god in Britain. Called ‘Tanarus’ on an altar erected by a legionary officer at Chester (RIB 452), this incarnation is sometimes construed as a variant of the Gallic Taranis mentioned by Lucan (Pharsalia i. 446) and popularly, if disputably, associated with continental figures of a bearded deity with wheel and thunderbolt (Green 1986). These same symbols appear on the sides of altars set up to IOM by the Second Cohort of Tungrians at Castlesteads (RIB 1981–​1983), but lack any textual indication of a hybrid nature and otherwise resemble the many other altars along Hadrian’s Wall honouring IOM as per military custom. Jupiter’s own resistance to twinning and his rare occurrence on defixiones suggest it may have been inappropriate to conceive of this god in localized form in Britain. Double-​named deities illustrate how votaries grappled with the problem of identification when a single theonym failed to convey the totality of a divinity or situate it adequately within the larger cosmological scheme. Variations in name pairings suggest these associations were contextually contingent rather than centrally dictated. Cocidius is equated with Mars at the west end of his domain, once with Silvanus in the east (RIB 1578), and again with an otherwise unknown deity, Vernostonus, on an altar set up by a German at Ebchester, south of Hadrian’s Wall and outside Cocidius’ primary range (RIB

Names of Gods   635 1102). The deity presiding over the shrine at Uley is twice addressed in curse tablets as Mars rather than the usual Mercury (Tomlin 1993). Another invokes ‘Mars Mercury,’ while a defixio to ‘Mars Silvanus’ was subsequently corrected to Mercury. Tomlin (1993: 115) notes that a toponym in several of the Uley inscriptions is applied to Mars as well as Mercury, implying that all of these names may refer to the same entity. The persistence of double-​named deities alongside other variants hints at greater complexity and nuance in the development of regional pantheons, reflecting diverse constituencies and concerns, rather than the supplanting of one set of divinities by another. Deities who escaped twinning cannot be assumed to have been any less changed by the processes that led their names to be set down in stone and metal. The preservation of British theonyms is inextricably linked with specific rituals by which religious knowledge and notions of deity were affirmed and propagated across the western Roman provinces. Local gods must have undergone some degree of conceptual translation in order to be epigraphically observable (if indeed they possessed pre-​Roman antecedents). Although Derks (1998: 234–​239) argues for a prior votive tradition in northern Gaul, the rites, language, and covenants recorded in inscriptions are all hallmarks of Roman practice. Devotees engaged these entities through Latin text and the Roman-​ styled rituals of vows, curses, and formal sacrifice, seeking divine intervention close to home while referencing a greater shared system of practice and belief.

Conclusions Roman religious knowledge is often described as grounded in orthopraxy: the proper conduct of prescribed rites (Scheid 2003; Ando 2008). The enacting of vows and curses helped create and validate this knowledge. Dedicated objects served as media for communication with the divine and as proof of the efficacy of these modes of address. But just what constituted ‘correct action’ was subject to setting and circumstance. Woolf (2003: 149–​150) notes that, for most provincial inhabitants, conceptions of deity were formed and experienced primarily at local shrines. Variations in the allocation of sacred space within temple complexes in Britannia and Hispania suggest to Revell (2009) that different aspects of religious performance were emphasized by their attendant populations. Selectivity in ritual and its physical expression is also evident in the epigraphy from Britain. The perception and reception of the new, Roman-​inspired, symbolic and ceremonial repertoire depended upon the identity of the practitioner, the customs of the community, and their individual and collective relationships within the larger imperial system. The evidence for religious activity in Roman Britain is a study in contrasts. Efforts to construct a coherent narrative of belief from disparate remains tend to obscure social and regional diversity while favouring certain subgroups and sources of information. Continuity of belief from the pre-​Roman Iron Age is difficult to trace and literary accounts of provincial practice are lacking, leading to a reliance on epigraphy and

636

636   Amy Zoll introduced rituals and votive forms. The territory is, however, comparatively inscription-​poor, except in areas in the north and west long occupied by the Roman military, where votive altars prevail. Non-​monumental inscriptions—​such as curse tablets deposited in western sanctuaries and engraved rings from the east Midlands—​offer another perspective on the gods of Britannia and their followers. The written sentiments on votive monuments and personal offerings have clear origins in Roman practice, although differences have been detected in their authors, applications, and distributions. Vows and curses define the expectations and obligations of gods and mortals through formalized exchange, but neither was observed by all inhabitants equally. In Britain, inscriptions preserved on stone overwhelmingly represent the religious activities of soldiers and their dependants, while other subsets of the populace etched their pleas for divine justice on lead tablets or declared their devotion through personal adornment. These objects offer an intimate view of how individuals and groups articulated their social and spiritual identity within their respective communities. The participants and trappings of a top-​down, centrally imposed, and officially encouraged state religion may have influenced the form and execution of these modest monuments, engraved exhortations, and other mementoes of votive practice, but did not dictate them. The gods themselves were not static or absolute, but took on names and aspects that accorded with the views and aspirations of their constituents. As such, these deities have proved difficult to pigeonhole, with past structural approaches resulting in false dichotomies that mask their complexity. Nowhere is this more evident than in the study of the handful of instances in which divine beings were addressed by two names. These were often seen to represent a snapshot of syncretism in action rather than one of many possible outcomes of the multifaceted and ongoing processes of cultural negotiation. All gods of the provinces, regardless of their origins, were subject to these forces and are best understood within their social, temporal, and geographic context. So, too, the items that bear their appellations must be regarded as contextualized objects, with forms, functions, authors, and audiences, rather than simply disembodied texts. In this way, the kaleidoscopic variety of divinities that comprise the known pantheon of Roman Britain can be seen to reflect the needs of the faithful and their efforts to be heard by the gods they called by name.

Abbreviations AE 

L’Année Epigraphique.

Paris: Ernest Leroux/​Presses Universitaires de France, 1888–​to date.

CIL  Corpus Inscriptorum Latinorum.   Consilio et Ductoritate Academie Litterarum Regiae Borussical Edition. Berlin: Academieder Wissenschaften, 1862–​to date. RIB   R. G. Collingwood, and R. P. Wright, The Roman Inscriptions of Britain. Stroud: Alan Sutton, 1992–​date.

Names of Gods   637

References Primary Sources Lucanus [M. Annaeus], Pharsalie Libri X, ed. C. H Weise. Leipzig: G. Bassus, 1835. Tacitus, De Germania, ed. H. Furneaux. Oxford: Clarendon Press, 1894.

Secondary Sources Adams, G. (2006). ‘The Social and Cultural Implications of Curse Tablets [defixiones] in Britain and on the Continent’, Studia Humaniora Tartuensia, 7: 1–​15. Aldhouse Green, M., and Raybould, M. E. (1999). ‘Deities with Gallo-​British Names Recorded in Inscriptions from Roman Britain’, Studia Celtica, 33: 91–​135. Allason-​Jones, L., and McKay, B. (1985). Coventina’s Well:  A  Shrine on Hadrian’s Wall. Hexham: Trustees of the Clayton Collection, Chesters Museum. Ando, C. (2007). ‘Exporting Roman Religion’, in J. Rüpke (ed.), A Companion to Roman Religion. Oxford: Blackwell, 429–​445. Ando, C. (2008). The Matter of the Gods. Berkeley and Los Angeles:  University of California Press. Beard, M. (1991). ‘Writing and Religion:  Ancient Literacy and the Function of the Written Word in Roman Religion’, in J. H. Humphrey (ed.), Literacy in the Roman World. Ann Arbor, MI: Journal of Roman Archaeology Supplementary Series 3, 35–​58. Birley, A. R., and Birley, A. (2010). ‘A Dolichenum at Vindolanda’, Archaeologia Aeliana, 39: 25–​51. Biró, M. (1975). ‘The Inscriptions of Roman Britain’, Acta Archaeologica Academiae Scientiarum Hungaricae, 27: 13–​58. Cepas, A. (1989). The North of Britannia and the North-​West of Hispania:  An Epigraphic Comparison. BAR International Series 470. Oxford: Archaeopress. Creighton, J. (2006). Britannia: The Creation of a Roman Province. London: Routledge. Cunliffe, B. (1988) (ed.). The Temple of Sulis Minerva at Bath Volume II:  The Finds from the Sacred Spring. Oxford University Committee for Archaeology Monograph 16. Oxford: Oxford University Committee for Archaeology. Daubney, A. (2010). ‘The Cult of Totatis:  Evidence for Tribal Identity in Mid Roman Britain’, in S. Worrell, G. Egan, J. Naylor, K. Leahy, and M. Lewis (eds). A Decade of Discovery: Proceedings of the Portable Antiquities Scheme Conference 2007. BAR British Series 520. Oxford: Archaeopress, 109–​120. Derks, T. (1998). Gods, Temples and Ritual Practices: The Transformation of Religious Ideas and Values in Roman Gaul. Amsterdam: Amsterdam University Press. Goldberg, D. M. (2009a). ‘The Dichotomy in Romano-​Celtic Syncretism: Some Preliminary Thoughts on Vernacular Religion’, in M. Driessen, S. Heeren, J. Hendriks, F. Kemmers, and R. Visser (eds), TRAC 2008: Proceedings of the Eighteenth Annual Theoretical Roman Archaeology Conference, Amsterdam 2008. Oxford: Oxbow Books, 187–​202. Goldberg, D. M. (2009b). ‘Divinities and Ritual Sites of Rivers in Northern England and Southern Scotland’. Unpublished Ph.D. thesis, University of Glasgow. Gordon, R. (1990). ‘Religion in the Roman Empire: The Civic Compromise and its Limits’, in M. Beard and J. A. North (eds), Pagan Priests: Religion and Power in the Ancient World. Ithaca, NY: Cornell University Press, 235–​255.

638

638   Amy Zoll Green, M. J. (1986). ‘Jupiter, Taranis and the Solar Wheel’, in M. Henig and A. King (eds), Pagan Gods and Shrines of the Roman Empire. Oxford: Oxford University Committee for Archaeology, 65–​75. Green, M. J. (1989). Symbol and Image in Celtic Religious Art. London: Routledge. Green, M. J. (1998). ‘God in Man’s Image: Thoughts on the Genesis and Affiliations of Some Romano-​British Cult-​Imagery’, Britannia, 29: 17–​30. Harris, W. V. (1989). Ancient Literacy. Cambridge, MA: Harvard University Press. Haynes, I. P., and Wilmott, T. (2012). ‘The Maryport Altars: An Archaeological Myth Dispelled’, Studia Universitatis Babes-​Bolyai, Historia, 51/​1: 25–​37. Henig, M. (1984). Religion in Roman Britain. New York, NY: St Martin’s Press. Henig, M. (1985). ‘Graeco-​Roman Art and Romano-​British Imagination’, Journal of the British Archaeological Association, 138/​1: 1–​22. Jackson, R., and Burleigh, G. (2007). ‘The Senua Treasure and Shrine at Ashwell (Herts)’, in R. Haeussler, A. C. King, and P. Andrews (eds), Continuity and Innovation in Religion in the Roman West. Portsmouth, RI: Journal of Roman Archaeology Supplementary Series 67, i. 37–​54. Johns, C. (1986). ‘Faunus at Thetford: An Early Latian Deity in Late Roman Britain’, in M. Henig and A. King (eds), Pagan Gods and Shrines of the Roman Empire. Oxford: Oxford University Committee for Archaeology, 93–​103. Johns, C. (2003). ‘Art, Romanisation and Competence’, in S. Scott and J. Webster (eds), Roman Imperialism and Provincial Art. Cambridge: Cambridge University Press, 9–​23. Jones, B., and Mattingly, D. (1990). An Atlas of Roman Britain. Oxford: Blackwell. Kiernan, P. (2004). ‘Did Curse Tablets Work?’, in B. Croxford, H. Eckhardt, J. Meade, and J. Weekes (eds), TRAC 2003:  Proceedings of the Thirteenth Annual Theoretical Roman Archaeology Conference, Leicester 2003. Oxford: Oxbow Books, 123–​134. King, A. C. (2007). ‘Romano-​Celtic Temples in Britain:  Gallic Influence or Indigenous Development?’, in R. Haeussler, A. C. King, and P. Andrews (eds), Continuity and Innovation in Religion in the Roman West. Portsmouth, RI: Journal of Roman Archaeology Supplementary Series 67, i. 13–​18. MacMullen, R. (1982). ‘The Epigraphic Habit in the Roman Empire’, American Journal of Philology, 103/​3: 233–​246. Mann, J. C. (1985). ‘Epigraphic Consciousness’, Journal of Roman Studies, 75: 204–​206. Mattingly, D. (2006). An Imperial Possession:  Britain in the Roman Empire, 54 bc–​ad 409. London: Penguin. Millett, M. (1995a). Roman Britain. London: Batsford/​English Heritage. Millett, M. (1995b). ‘Re-​Thinking Religion in Romanization’, in J. Metzler, M. Millett, N. Roymans, and J. Slofstra (eds), Integration in the Early Roman West: The Role of Culture and Ideology: Papers arising from the International Conference at the Titelberg (Luxembourg) 12–​13 November 1993. Luxembourg:  Musée National d’Histoire et d’Art, Luxembourg, 93–​100. Mirković, M. (1989). ‘Beneficiarii consularis and the New Outpost in Sirmium’, in V. A. Maxfield and M. J. Dobson (eds), Roman Frontier Studies (1989): Proceedings of the XVth International Congress of Roman Frontier Studies. Exeter: University of Exeter Press, 252–​256. Raybould, M. E. (1999). A Study of Inscribed Material from Roman Britain: An Inquiry into Some Aspects of Literacy in Romano-​British Society. BAR British Series 281. Oxford: Archaeopress. Revell, L. (2009). Roman Imperialism and Local Identities. Cambridge:  Cambridge University Press.

Names of Gods   639 Richert, E. A. (2005). Native Religion under Roman Domination: Deities, Springs and Mountains in the North-​West of the Iberian Peninsula. Oxford: Archaeopress. Ross, A. (1967). Pagan Celtic Britain: Studies in Iconography and Tradition. London: Routledge and Kegan Paul. Scheid, J. (2003). An Introduction to Roman Religion. Bloomington, IN:  Indiana University Press. Stewart, P. (2010). ‘Geographies of Provincialism in Roman Sculpture’, Journal of the International Association of Research Institutes in the History of Art, 5: 1–​59. Stoll, O. (2007). ‘The Religions of the Armies’, in P. Erdkamp (ed.), A Companion to the Roman Army. Oxford: Blackwell, 451–​476. Tomlin, R. S. O. (1988). ‘The Curse Tablets’, in R. S. O. Tomlin (ed.), Tabellae Sulis: Roman Inscribed Tablets of Tin and Lead from the Sacred Spring at Bath. Oxford: Oxford University Committee for Archaeology, 59–​105. Tomlin, R. S.  O. (1993). ‘The Inscribed Lead Tablets:  An Interim Report’, in A. Woodward and P. Leach (eds). The Uley Shrines: Excavation of a Ritual Complex on West Hill, Uley, Gloucestershire:  1977–​9. English Heritage Archaeological Report 17. London:  English Heritage, 113–​130. Tomlin, R. S. O. (2002). ‘Writing to the Gods in Britain’, in A. Cooley (ed.), Becoming Roman, Writing Latin? Literacy and Epigraphy in the Roman West. Portsmouth, RI: Journal of Roman Archaeology Supplementary Series 48, 165–​179. Turner, E. G. (1963). ‘A Curse Tablet from Nottinghamshire’, Journal of Roman Studies, 53: 122–​124. Versnel, H. S. (1991). ‘Beyond Cursing: The Appeal to Justice in Judicial Prayers’, in C. A. Faraone and D. Obbink (eds), Magika Hiera:  Ancient Greek Magic and Religion. Oxford:  Oxford University Press, 60–​106. Webster, G. (1986). The British Celts and their Gods under Rome. London: Batsford. Webster, J. (1995). ‘“Interpretatio”:  Roman Word Power and the Celtic Gods’, Britannia, 26: 153–​161. Webster, J. (1997). ‘Necessary Comparisons: A Post-​Colonial Approach to Religious Syncretism in the Roman Provinces’, World Archaeology, 28/​3: 324–​338. Webster, J. (2003). ‘Art as Resistance and Negotiation’, in S. Scott and J. Webster (eds), Roman Imperialism and Provincial Art. Cambridge: Cambridge University Press, 24–​51. Whittaker, C. R. (1997). ‘Imperialism and Culture: The Roman Initiative’, in D. J. Mattingly and S. E. Alcock (eds), Dialogues in Roman Imperialism: Power, Discourse, and Discrepant Experience in the Roman Empire. Portsmouth, RI:  Journal of Roman Archaeology Supplementary Series 23, 143–​163. Woodward, A. (1992). Shrines and Sacrifice. London: English Heritage/​Batsford. Woolf, G. (1996). ‘Monumental Writing and the Expansion of Roman Society in the Early Empire’, Journal of Roman Studies, 86: 22–​39. Woolf, G. (1997). ‘Polis-​Religion and its Alternatives in the Roman Provinces’, in H. Cancik and J. Rüpke (eds), Römische Reichsreligion und Provinzialreligion. Tübingen:  Mohr Siebeck, 71–​84. Woolf, G. (1998). Becoming Roman:  The Origins of Provincial Civilization in Gaul. Cambridge: Cambridge University Press. Woolf, G. (2001). ‘Local Cult in Imperial Context:  The Matronae Revisited’, in P. Noelke (ed.), Romanisation und Resistenz in Plastik, Architektur und Inschriften der Provinzen des Imperium Romanum: neue Funde und Forschungen. Mainz: Zabern, 131–​138.

640

640   Amy Zoll Woolf, G. (2003). ‘Seeing Apollo in Roman Gaul and Germany’, in S. Scott and J. Webster (eds), Roman Imperialism and Provincial Art. Cambridge: Cambridge University Press, 139–​152. Zoll, A. (1995a). ‘A View through Inscriptions:  The Epigraphic Evidence for Religion at Hadrian’s Wall’, in J. Metzler, M. Millett, N. Roymans, and J. Slofstra (eds), Integration in the Early Roman West: The Role of Culture and Ideology: Papers Arising from the International Conference at the Titelberg (Luxembourg) 12–​ 13 November 1993. Luxembourg:  Musée National d’Histoire et d’Art, 129–​137. Zoll, A. (1995b). ‘Patterns of Worship in Roman Britain: Double-​Named Deities in Context’, in S. Cottam, D. Dungworth, S. A. Scott, and J. Taylor (eds), TRAC 94: Proceedings of the Fourth Annual Theoretical Roman Archaeology Conference. Oxford: Oxbow Books, 32–​44. Zoll, A. (forthcoming). ‘A Vow Fulfilled: Inscribed Votive Altars from Britannia and Hispania Citerior’, Ph.D. dissertation, University of Pennsylvania, Philadelphia.

Chapter 31

Ritual Dep o si t i on Alex Smith

Introduction It is increasingly apparent that the nature of sacred space and forms of ritual expression varied tremendously within the overall geographical and chronological parameters that define Roman Britain. Even in terms of what may be regarded as ‘formal’ cult sites—​ that is, those that are spatially separate from domestic spheres and often associated with prescribed architectural principles—​there is still a significant array of variants, which frequently leads to interpretational issues. In fact, it is often only on the basis of the frequency, type, and patterning of artefactual and ecofactual deposits within such contexts that any consensus about a religious interpretation can be reached. The concept of votive deposition within sacred sites was well established within Roman religious traditions across the empire (Rives 2007: 24–​27) and indeed a number of classical authors (notably Caesar (6.13; 6.17), Strabo (4.1.13; 12.5.2), and Diodorus Siculus (5.27)) attested to the display and deposition of objects within sacred loci in pre-​conquest Gaul (see A. T. Smith 2001: 14, table 1.3). These ritual deposits have long been recognized archaeologically on late Iron Age and Roman temple sites in France and Britain (Wait 1985; Brunaux 1988; Woodward 1992; A. T. Smith 2001; King 2007b; Bagnall-​Smith 2008), whether in the form of ‘structured’ sub-​surface deposits, concentrations of material associated with boundaries, or just in the general differentiation of find types across the site. Nevertheless, interpretation of such deposits even within these undeniably religious contexts is not always straightforward, and the likelihood is that they represent the vestiges of a variety of methods of divine propitiation, from formal cult rituals performed by temple priests to personal devotions, perhaps given in anticipation or recompense of answered prayers (Ghey 2007: 25). Many of these offerings and the rituals surrounding their deposition were probably part of a broader understanding of ritual expression throughout the Roman Empire, which, although far from homogenous, had certain underlying shared characteristics quite possibly linked to the creation of wider cultural identities (Revell 2007: 212). In this respect,

642

642   Alex Smith although the many different types of votive deposit seen in Romano-​British temple sites could be viewed as an important element in the construction of local community identities (see Mattingly 2006: 520–52​2), they may also reflect—​in part at least—​something of the wider understanding of what it meant to ‘act’ as Roman. Furthermore, this may be expressed not only in the type and nature of votive offering, but also in the architectural setting of the temple itself. The very concept of specialized constructed sacred space was, with a small number of notable exceptions, one that only really developed in Britain after the Roman conquest (A. T. Smith 2001: 67), and so, despite the extensive physical variety in site types, many may be seen as local interpretations of a Roman ‘identity’. In terms of temple offerings, it has been argued that many may have been deposited as an element in the ritual of the vow, an established and formulaic method of communication with the deity within the Roman world, for which evidence is well attested in certain parts of Gaul and Britain (Derks 1995; Bagnall-​Smith 2008: 153–​154). The two parts of this votum, a temporary pact made with the god, comprised the nuncupatio, the promise to pay a gift in return for divine service, and the solutio, whereby the vow is paid by the votary. Specific evidence for the nuncupatio part of the vow is generally lacking in Britain, aside from the inscribed lead curse tablets found in quantity at Bath and Uley (Bagnall-​Smith 2008), while the solutio part is attested by numerous inscriptions of VSLM (votum solvit libens laetus merito), indicating the vow was paid, found in parts of Britain, though mainly in military zones (Zoll 1995). The extent to which this ritual of the vow was explicitly practised in other temples in Britain remains uncertain (though inscriptions suggesting this have been found at five temples (A. T. Smith 2001: 156)), but the often considerable quantity of objects recovered from such sites that may be regarded as votive offerings would suggest that at least the basic principle of ‘paying off a debt’ to the god was acknowledged to a degree. Of course, votive offering and deposition within a formal temple setting is only one aspect of ritual expression within Roman Britain, albeit one that until comparatively recently formed a significant focus of academic study in terms of Romano-​British religion. The other major body of evidence concerns deposition of metalwork and other objects within natural watery settings, in particular rivers, lakes, and bogs. This has been demonstrated to have had a long tradition stretching back into early prehistory (Bradley 2000: 47–​63) and certainly continued into the later Iron Age and Roman periods, with examples of weapons, ceramic and pewter vessels, jewellery, and other items being recorded as distinct deposits—​for instance, within the rivers Thames (Booth et al. 2007: 217–​220) and Witham (Field and Parker Pearson 2003: 171–​178). Aside from such ritual activity associated with what may be regarded as sacred sites, whether constructed shrines or natural watery features, and following on from the work of Hill (1995) and Cunliffe (1995) on ritual deposition within Iron Age settlements, Fulford (2001) recognized that ‘structured’ or ‘special’ deposits were also widespread within Roman domestic contexts, and not just within rural settlements but also within urban centres alongside all of the overt trappings of Romanitas. Further work by Clarke (2001; forthcoming) and Hingley (2006) has sought to go further and identify patterns within such deposition in terms of physical location (for example, ‘boundary deposits’),

Ritual Deposition   643 in the types and materiality of objects and the ways in which they were deposited, and in combinations of ‘object and place’, which, it is argued, may have ultimately determined the power of the votive action (Hingley 2006: 239). The ritual significance within domestic contexts of different categories of finds (for example, ornamentation: Puttock 2002) and especially animal species (for example, dogs: K. Smith 2006; corvids: Sergeantson and Morris 2011; see also Black 2008 and Morris 2011) has also been increasingly discussed in recent years, and this has all led to a much greater recognition of such unusual deposits within the archaeological record, as will be shown below. In the main part of this chapter it is the intention to provide an overview of the practice of ritual deposition within both formal sacred sites (excluding the body of evidence from natural watery sites) and the ‘structured deposits’ in ostensibly secular settlements within Roman Britain. In both cases, there are some similarities in the types of objects and the ways in which they were deposited, whether this is in terms of articulated animal remains or deposition of whole pots, and yet there are also clearly many differences, which may be culturally significant. Furthermore, there is not always a clear differentiation between the different types of ritual expression, with, for example, purpose-​built ritual pits and shafts being located within and around secular settlements and classic ‘structured deposits’ within disused wells and pits associated with temple sites. As ever, when it comes to personal religious experience in the Roman world, the boundaries between secular and divine were consistently blurred.

Ritual Deposition Within Temple Contexts As stated above, it has long been recognized that, compared to the post-​conquest period, evidence for constructed cult sites in late pre-​Roman Iron Age Britain was quite minimal, with a number of notable and in many ways unusual exceptions (for example, Hayling Island, Uley, etc.: A. T. Smith 2001: 67). While there continues to be new discoveries, such as at Snows Farm (Evans and Hodder 2006) and Marcham/​Frilford (Kamash et al. 2010), that provide further potential evidence for the existence of pre-​ Roman religious foci (see below), the rarity of Iron Age shrines—​or at least their lack of recognition in the archaeological record—​remains undeniable. Yet within fifty years of the conquest, constructed temples were a regular part of the landscape, at least in many parts of southern and eastern Britain, as ubiquitous as roads, towns, and villas. Indeed, in looking at the finer chronological details, such sacred sites were often among the first substantial pieces of architecture to be constructed in these regions during the post-​ conquest period (A. T. Smith 2001: 144). There are two fundamental factors in the interpretation of formal cult sites in Roman Britain: recognition of standardized forms of religious architecture (for example, the Romano-​ Celtic temple) and/​ or architectural elements conforming to underlying

644

644   Alex Smith religious spatial principles (for example, emphasis on enclosure and entranceway, pathways, and so on), and perhaps most importantly in the type, volume, and patterns of finds associated with the site. The earliest systematic studies of Roman temples in Britain (e.g. Lewis 1966; Wilson 1975) tended to concentrate on matters of architectural detail and classification, and it was not really until the publication of Temples, Churches and Religion in Roman Britain (Rodwell 1980) that the nature of finds within these sacred sites really began to be explored at any level. Since then there have been various wider studies of votive deposition within Romano-​British temple sites (e.g. Woodward 1992; A. T. Smith 2001; King 2007b; Bagnall-​Smith 2008), in addition to specific analyses of particular categories of deposit within such formal religious contexts (for example, coins: Haselgrove 2005; King 2007a; Wythe 2007; animal remains: King 2005; K. Smith 2006). Furthermore, there have been a number of important recent publications on excavations at sacred sites, often with significant details of ritual deposition, such as at Snows Farm (Evans and Hodder 2006), Hallaton (Score 2011), Higham Ferrers (Lawrence and Smith 2008), Springhead (Andrews et al. 2011), and Great Chesterford (Medlycott 2011). Recent excavations at Marcham/​Frilford in Oxfordshire have revealed extensive parts of a huge rural religious complex mostly outside of the temple and temenos (religious precinct) area, which have demonstrated the presence of a range of different ritual foci; this is helping us to understand something of the wider functions and intricacies of such sanctuaries (Kamash et al. 2010). The following section will first look briefly at the general character of votive offerings most commonly found at cult sites, before exploring some of these newly excavated and/​or published sites in terms of their contribution to our understanding of votive deposition. In addition, the results of some previous analyses of sites such as Uley will be summarized, as they remain fundamental to studies of ritual practice within Romano-​British temples.

Votive Offerings Within Cult Sites In a previous study by the current author of seventy-​five temple sites across southern Britain, the major categories of votive artefacts were quantified and discussed (see Table 31.1). The first important observation to make is that there was a significant variety in the numbers and types of votive objects between temple sites, undoubtedly reflecting the nature of the individual cult as well as the personal wealth and aspirations of the supplicant. There is some evidence for wider regional patterns (for example, relative lack of coins within Sussex temples), but even here there are many discrepancies, and overall it seems clear that there were no widespread regulations concerning the types of object that could be offered to the deities. Nevertheless, there are certain artefact classes that would seem to have been regarded as more appropriate votive material, and the most ubiquitous find type within temple contexts is coinage, often found deposited in large quantities (for example, Uley, Harlow, Wanborough). There have been various recent

Ritual Deposition   645 Table 31.1 Occurrence of selective votive objects in southern Romano-​British cult sites Votive Item

Occurrence (no. of sites out of 75)

Coins

60 (29)

Personal ornamentation and toiletries

46 (20)

Martial items Items associated with healing

25 (4) 8 (1)

Miniature items

20 (3)

Votive/​miniature pots

14 (3)

Votive plaques/​leaves

13 (3)

Iconography Inscriptions (including curse tablets)

44 (12) 23 (5)

Note: Bold figures in parentheses denote sites where items were found in significant comparative quantity. Source: After Smith (2001: 155, figure 5.13).

studies on the deposition of both Iron Age (Haselgrove 2005) and Roman (King 2007a; Wythe 2007) coins in temple contexts, where it was noted that Iron Age coins in particular were particularly prevalent deposits at certain sacred sites around the time of the Claudian conquest, and that, spatially, the coins from these sites were concentrated in front and to the left of the cult focus (Haselgrove 2005: 417). In terms of Roman coins, their deposition occurred right up until the end of the fourth century ad, and indeed it was observed that coins continued to be deposited at temples long after they had structurally fallen to decay, a testament to the sacred memory of such sites (King 2007a: 37). Other important categories of ‘everyday’ objects used as votive deposits at temple sites include personal objects such as brooches, finger rings, and hairpins. In a study concerned with the ritual significance of jewellery in Roman Britain, Puttock (2002: 115) suggested that such items deposited in ritual situations carried the hopes and prayers of the people, and it was not only the object type itself that was significant, but also the shapes, colours, motifs, and materials of the ornaments. At its most obvious, this is probably seen with the prevalence of horse and rider brooches at a number of religious sites such as Lamyatt Beacon and Woodeaton (Leech 1986; Bagnall-​Smith 1995). It is further argued that the symbolism behind the jewellery was often connected with fertility, healing, or protection from evil, with certain categories (for example, beads, hairpins, and bracelets) being particularly associated with women and linked with aspects such as childbirth and fertility (Puttock 2002: 115). Of course, the ritual significance of objects such as coins and personal items is only readily apparent when they are found in such unequivocal religious contexts as temple

646

646   Alex Smith sites, except perhaps when they form part of recognized structured deposits within settlements. There are, however, many types of object that seem to have been fabricated purely with religious associations in mind, and these form the very wide category of ‘religious objects’, ranging from inscribed altars and curse tablets to figurines of deities, leaf/​feather plaques, and objects of priestly regalia (see Bagnall-​Smith 2008 for a useful summary of these types). The type and occurrence of such items varies enormously within temple sites across Britain, although there are several characteristics that are more widespread, most notably that of miniaturization. Miniature objects can take a variety of forms, with tools and weapons the most common, especially axes that are found on temples and shrines across Britain. Kiernan (2009: 211–​213) has suggested that, far from having some universal significance implied by their miniaturization, these models carried a variety of meanings, with some being direct substitutes for real objects, others being indirect substitutes for ritual activities or human activities such as sacrifice (axes) or travel (shoes, anchors), and finally others being purely symbolic and taken from the iconography of a divinity (for example, wheels). An act of ritual expression that seems to occur across a range of intrinsically religious and ‘normal’ votive objects is that of destruction. This is a widespread phenomenon both in Iron Age and Roman contexts and is generally believed to be a sacrificial act, removing the object from earthly existence and transferring it to the divine realm (Green 1995: 470–​471). Classic examples of this ritual have recently been found at the shrine at Higham Ferrers in Northamptonshire, including a bone pin whose stem was broken and partially split and a brooch whose pin had been detached and deliberately wrapped around the bow (Lawrence and Smith 2008: 331). Perhaps the most significant ritual act within many sacred sites would appear to be the sacrifice of living animals. Animal bone usually forms one of the largest single categories of find within temple sites, with over 50,000 recorded from Uley, Gloucestershire (Woodward and Woodward 2004) and over 32,000 from the relatively small shrine at Snows Hill, Cambridgeshire (Evans and Hodder 2006). In a recent analysis of animal remains from temples in Roman Britain, King (2005) noted a number of different characteristics, with some temples having large numbers of bones of a distinctive species with evidence for regular periods of sacrifice (presumably at festivals), others with smaller quantities of bone, often in structured deposits (see below), and some with very few bones, often those temples associated with healing such as at Bath. The distinction between those articulated skeletons within structured sub-​surface deposits and the mass of unarticulated bone (presumably the remains of ritual feasting) is of great importance, and is explored further later in this chapter.

Offering Zones There are a number of inherent problems in identifying patterns of votive deposition within cult sites, including a lack of excavation within the temenos, selective (and

Ritual Deposition   647 non-​spatial) recording of those finds encountered, and more broader issues of redeposition. Nevertheless, there are sufficient excavated sites with enough spatially specific information to gain some understanding of the general patterns of finds distribution (A. T. Smith 2001; King 2007b: 191–​193). In the first instance, there is a broad division between those finds that were intended to be deliberate deposits in a sub-​surface environment (for example, pits, ditches, and so on), and those associated with the occupation surface, which are by their nature often far more disturbed and difficult to interpret unless there is an excellent state of preservation. Within one of the Springhead temples in Kent, for example, there were a number of votive finds recorded as still apparently lying in situ upon the temple cella floor, including a miniature axe and a Venus figurine (Penn 1962: 116). If this was indeed the case, there is good reason to argue that the finds were originally displayed here and remained in their primary context, whereas it is likely that in most cases there was significant redeposition of votive objects that may have been on display before being removed and disposed of elsewhere, either within the temple precinct or beyond. When looking at the overall finds distribution patterns within temple complexes, it is usually possible to obtain only quite a generalized impression of potential ‘offering zones’ (that is, areas for display and/​or deposition of votive offerings) that undoubtedly mask a complex array of rituals enacted at the cult site. Nevertheless, it remains that in the majority of sites studied, it is—​perhaps unsurprisingly—​the temple building itself, and especially the central cella, that formed the main focus for the display and/​or final deposition of votive objects (A. T. Smith 2001: 154). Coins were particularly prevalent finds in this location, perhaps being displayed before the main cult image (see Wythe 2007: 47–​50 for a detailed analysis of coin deposits in temple buildings). This is seen, for example, within the Romano-​Celtic temple at Bourton Grounds in Buckinghamshire, where 90 per cent of the 315 coins came from the southern half of the cella, interpreted by the excavators as the remains of offerings given to the deity through openings in the cella walls (Green 1966: 359). Even within less elaborate sacred sites, such as the simple circular shrine at Claydon Pike, Gloucestershire, which lay close to a modest late Roman villa, there was a distinct concentration of objects within the interior of the shrine building, including the largest single collection of coins from the site (Miles et al. 2007: 183). However, variations to this do of course exist, as King has recently pointed out in his study of coins from temples in Britain, where he noted concentrations of coin deposition in zones to the south-​east of the temple in at least three instances, at Hayling Island, Bath, and Wanborough (King 2007a: 26–​28). A recently excavated shrine on one side of the Nene valley within a roadside settlement at Higham Ferrers in Northamptonshire comprised a walled enclosure (though open towards the river), monumental entrance, and two conjoined precincts, though without any evidence for an actual temple structure (Lawrence and Smith 2008). The inner precinct in particular contained a substantial volume of votive objects including coins, personal items, and metal leaves, in addition to animal remains, all lying relatively undisturbed and in situ on top of the Roman occupation surface. The finds were concentrated in the south-​east of the precinct (which is perhaps significant, given the observations of

648

648   Alex Smith King already noted), particularly around a 3-​metre-square area largely devoid of objects, which was suggested as space reserved for a cult focus of some kind (ibid., 332). In this respect the open inner precinct probably acted as a temple cella much the same as in a traditional Romano-​Celtic temple. The quantity of iron nails within the inner precinct suggested the presence of wooden structures of sorts, to which offerings would perhaps have been attached, and the large number of hobnails within the precinct, which show a similar distribution, may have been used for making such attachments. In addition to the main cult focus within the heart of the sacred site, other notable offering zones within sanctuaries, where there were often distinct concentrations of finds, included entrances to both the temple and the outer precinct, the area in front of the temple, sometimes on either side of a defined pathway, and in the proximity of the temenos boundary (A. T. Smith 2001: 154). In many ways this reflects the fundamentally important architectural principles of sacred sites in different cultures, where there is often a distinct emphasis on the passage from the first transitional point of the outer boundary through to the prime religious focus of the site (Barrie 1996: 148). Following metal-​detector discoveries of large numbers of Iron Age coins on a hilltop location at Hallaton, Leicestershire, excavations revealed what appears to have been an open-​air shrine dating from the mid to late first century bc to the early post-​conquest period, with distinct concentrations of finds and animal bone around the east-​facing entranceway within an outer ditch (Score 2011: 156–1​60). As with Higham Ferrers, there appears to be no evidence for a temple building, but the excavators have suggested a processional way, taking into account prehistoric features in the landscape, leading through the entrance in the ditch, which has been interpreted as a potential enclosure surrounding the hilltop. The finds included 4,943 Iron Age coins (and 149 Roman coins dated pre-​ad 43), buried in a series of separate deposits on one side of the boundary ditch near the entranceway, along with a remarkable Roman cavalry parade helmet, a silver bowl, two ingots, and a substantial animal bone assemblage (6,929 identified fragments), 97 per cent of which was young pig. This was clearly a site of unusual significance, and the wealth of finds along with the timeframe of deposition may well be associated, in part, with the stresses of the Roman conquest, but nevertheless the clear concentration of votive activity along the boundary and entranceway conforms to well-​established patterns in terms of offering zones. A similar emphasis of ritual deposition has been found during recent excavations at the vast religious complex at Marcham/​Frilford in Oxfordshire, dating from the late first to late fourth century ad, but with possible evidence for an earlier ritual focus of Iron Age date (Kamash et al. 2010: 98–​102). Previous excavations of the temple had shown that robbing and plough damage had destroyed most of the interior levels, but concentrations of coins were noted on the temple pathway (Bradford and Goodchild 1939: 32), and this was again encountered during the recent excavations, in addition to large quantities of coins being focused on the monumentalized entrance into the temenos (Kamash et al. 2010: 104–​105). Preliminary analysis of wider finds distribution at this site has produced interesting results, with concentrations of coins, fineware pottery, and vessel glass immediately outside the temenos, the latter two categories suggesting

Ritual Deposition   649 feasting took place in this zone (Kamash et al. 2010: 106). Furthermore, there were also notable concentrations of personal objects around a small subsidiary shrine and an area of boggy ground to the south of what is described as a semi-​amphitheatre, lying c. 90 metres east of the temenos entrance, and it is clear that full analysis of the distribution of objects from this site will be of paramount importance in furthering our understanding of Romano-​British ritual practices within sacred sites beyond the confines of the temple and precinct. A ritual site with similar evidence for longevity, from the late Iron Age to the fifth century ad, was the exceptionally well-​recorded temple at Uley in Gloucestershire. This site contained a huge volume of finds that had all been spatially recorded, and in which a number of offering zones were discerned, despite the inherent problems of disturbed stratigraphy and significant redeposition (Woodward and Leach 1993; A. T.  Smith 2001: 106). The temple cella itself was clearly a focus for deposition, and the spreads of votive material (which included miniature pots, lead curse tablets, and votive leaves along with many coins and personal items) deposited along the demolished buildings representing the inner boundary to the cult site are argued to have come from a clearing-​ out of the temple when it was substantially modified in its final phase. Similar acts of redeposition are suggested for other Roman sanctuaries, including at Marcham/​Frilford, where a large pit and midden in the south-​east corner of the temenos contained deposits rich in oyster shell, animal bone, and pottery, thought to have originally been deposited elsewhere within the sacred precinct (Kamash et al. 2010: 103). At the Henley Wood temple in Somerset, where substantial quantities of votive objects were also recovered and spatially recorded, several distinct patterns were revealed, with coins again being concentrated in the temple building, as well as adjacent to a possible entranceway into the site and on either side of the main axis leading from here towards the temple (Watts and Leach 1996; A. T. Smith 2001: 92). Items of personal adornment (brooches, bracelets, and so on) had a slightly different depositional pattern, with most being found within the outer ditch (taken as the temenos boundary), often in distinct concentrations, and interpreted by the excavators as representing episodes of redeposition of objects that had originally been displayed in the temple (Watts and Leach 1996: MF 515). Whether or not this was the case, the spatial distributions of different object types suggests that these were not random dumps of unwanted material, but carefully structured deposits, of a type that are found in sub-​surface features (for example, ditches, pits, and shafts) within many temple sites.

Structured Deposits Within Temple Contexts Determining the location of offering zones within a cult site is reliant upon analysis of the distribution of votive objects. As already discussed there is a distinction between

650

650   Alex Smith those objects associated with occupation surface/​spreads, which may be either casual losses, the remains of items left ‘on display’, or else the product of post-​depositional processes (alluviation, colluviation, ploughing, and so on), and those objects deliberately deposited in pits, ditches, and so on, probably involving prescribed sets of rituals. In many ways, the structured deposits within sacred sites are easier to interpret, although may still represent the final stages of a myriad of different rituals associated with a variety of different beliefs. The items within such deposits, whether the remains of animals or artefacts, were probably regarded as appropriate messages to the gods, and placing them within specific sub-​surface contexts may have been part of their transformation on the journey from the earthly world to that of the divine (Aldhouse-​Green 2001: 24). In the context of the discussion of the animal remains at the shrine at Snows Farm, Haddenham, Cambridgeshire, the excavators suggested that the different types of deposit on site might reflect different ritual acts, with those animals deposited in pits relating to rites of intercession (that is, asking the gods for favour or paying off a divine debt) and the slaughter of animals alone (without burial) only intended for divination (Evans and Hodder 2006: 358). The animal remains at this shrine, which dated from the second to fourth century ad (and with a potential Iron Age precursor) were particularly prevalent (32,933 fragments), with an emphasis on sheep and evidence for the selection of particular body parts in places (Evans and Hodder 2006). Two zones were selected for structured sub-​surface deposits, with the floor area of the shrine containing a single articulated sheep (interpreted as a foundation deposit) and a number of ‘head and hoof ’ burials with coins placed in the jaws, while the north-​west corner of the precinct contained pits with articulated sheep skeletons associated with complete pots (Evans and Hodder 2006: 336). These deposits represented different types of rituals undoubtedly with different purposes, and all again were different from the mass spread of bones and other objects in the zone around the front of the shrine and perimeter of the enclosure, which are suggested as the remains of animals consumed in ritual feasting, which had been carefully redeposited from a previous midden at a time when the shrine underwent significant structural changes (Evans and Hodder 2006: 413). Large numbers of animal remains were also recovered from the temple precinct at Great Chesterford, Essex, which lay around 1 kilometre east of the walled Roman town (Medlycott 2011). The overwhelming majority of bones were those of lambs, which displayed a kill-​off pattern indicative of regular periods of sacrifice, probably at festivals, similar to the situation at other temples such as Harlow (Medlycott 2011: 84). The greatest concentration of bones came from a group of pits in the south-​west corner of the temenos, where it was estimated that there were the remains of over 1,000 sheep. How far these may be viewed as structured deposits is uncertain, as there was no evidence of articulation or patterning, and, as they seemed to have been dumped with oyster shell and pottery (and few other finds), they are more likely to be the remains of ritual feasting, as was suggested for the dump of material in pits in the south-​east corner of the temenos at Marcham/​Frilford, as already mentioned. They were certainly very different in nature to the groups of finds lying apparently just beneath the temple pathway, which comprised many personal objects and specifically votive items such as bronze

Ritual Deposition   651 feathers and a silver mask plaque, which had been deliberately folded before deposition (Medlycott 2011: 85). A recently excavated and published temple complex with extensive evidence for structured deposition is that at Springhead in Kent, on the line of Roman Watling Street (Andrews et al. 2011). The principal walled temple complex, comprising two Romano-​ Celtic temples and at least six other structures, was excavated during the 1950s and 1960s, while the recent excavations revealed substantial additional areas of the site, which may best be described as a major religious complex incorporating domestic, commercial, and possibly administrative functions (Andrews et al. 2011: 212). A late Iron Age religious focus is hinted by the deposition of Iron Age coins in the spring and structured deposits of pottery vessels and a horse burial in pits below a ‘processional’ way on the hill slope above the spring. During the second century ad, the area around the spring was transformed into what has been termed the sanctuary complex, with a shrine, a portico structure, free standing posts, and a number of pits or shafts containing structured deposits. These included a 4.5-​metre-​deep shaft by the entrance to the sanctuary, which contained at least twenty dogs, several buried with their chains, a number of near-​complete pots, a human skull, a group of animal skulls, and a cow placed in the bottom (Andrews et al. 2011: 80–​82, figure 2.55). The ritual deposition of dogs in pits and shafts is now a well-​known phenomenon in Iron Age and Roman Britain, occurring in both what may be considered secular and religious contexts (K. Smith 2006). Perhaps the most striking examples from what may be regarded as a sacred site is at Cambridge, where a series of at least thirteen ritual shafts (2–​3 metres deep) dating to the third–​fourth centuries were located in the centre of the small town, adjacent to an unusual second-​century shrine, which had burnt down (Alexander and Pullinger 2000). Each shaft contained the articulated remains of a mature small dog in the southern corner and a rush matt and wicker basket containing an infant burial laid on large pot sherds. In most shafts two burials occurred one above the other and in five shafts were the remains of small shoes that would have fitted an older child. Although these shafts were filled in immediately after the infant burials, it is likely that they were not elaborate funerary shafts, and yet their deposition along with the dogs would probably have still been regarded as very powerful ‘magic’ or messages to the gods.

Ritual Deposition Within Settlement Contexts It has already been mentioned that the boundaries between secular and divine were constantly blurred within Romano-​British society, and that structured deposits of an assumed ritual nature were not just confined to designated sacred sites. Since Fulford’s paper in 2001 in particular, the recognition of ritual deposition (often termed ‘special

652

652   Alex Smith deposits’) in Romano-​British settlement contexts has increased dramatically, to the point where care must be taken not to over-​interpret every single articulated animal or complete pot as ritual in nature, to the detriment of potentially more prosaic explanations (see Morris 2011: 182–18​5). Fulford (2001: 201) defined the evidence for ritual as having both a repetitive nature and ‘irrational’ characteristics, and in this sense there certainly are considerable numbers of archaeological examples throughout Britain, though, as Clarke (forthcoming) has recently asserted, the overall banner of ‘ritual’ undoubtedly masks a diverse range of actors and motivations. Broad patterns are certainly noted in many aspects of these deposits, but, as Hingley (2006: 239) states, ‘each discovery represents a unique act of deposition and should be studied and interpreted accordingly’. Examples of what would appear to have been ritual deposition have been demonstrated extensively in urban, rural, and military contexts, with Fulford (2001) looking primarily at such deposits with Romano-British towns, drawing on data from Silchester, London, Neatham, Baldock, and Verulamium. In the latest Silchester publications on the mid and late Roman occupation of the town, it has been demonstrated that ‘special deposits’ in pits and wells occurred throughout the excavated insula, though there were chronological differences in spatial patterning and assemblage composition, and the difficulties in defining them as ritual in nature were duly acknowledged (Fulford et al. 2006; Fulford and Clarke 2011). One third-​century pit was particularly noteworthy, containing one complete and five partial dog skeletons, in addition to partial raven skeletons and a folding knife or razor with an ivory handle in the shape of two coupling dogs (Fulford and Clarke 2011: 311). Other possible ritual deposits from Silchester seem associated with the construction of buildings, including complete pots (one of considerable antiquity) in the make-​ups of two adjacent corners of a structure and an iron mason’s trowel incorporated into one wall. Significantly, some of these pots appear to have been deliberately pierced, thus perhaps undergoing the rite of sacrifice, as noted above for many votive deposits in shrines. The motivations behind such deposits would surely have been different from those associated with the pits, perhaps as suggested by the excavators ‘to ensure the life, safety and security of the buildings’ (Fulford and Clarke 2011: 333). In rural contexts, ‘special deposits’ have been interpreted at a growing number of settlements, as seen, for example, in Upper Thames Valley sites such as Horcott Quarry and Cotswold Community, which have examples of articulated animal remains in pits and ditches that could be regarded as being deposited for spiritual reasons (Powell et al. 2010: 135; Hayden et al., forthcoming;). Elsewhere, Rudling (2008: 121–​129) has reviewed the evidence for religious practice across Sussex, including a number of ‘special deposits’ within rural settlements, such as a large number of dog remains within a well at Chilgrove villa and burials of dogs and complete pots in pits at the recently excavated villa at Barcombe. One factor to consider throughout all of these examples of ‘special deposits’ within non-​religious contexts is the extent to which the pits or shafts were dug for the specific purpose of ritual activity. In the case of the Silchester examples, it would appear that

Ritual Deposition   653 most were dug originally as wells or latrine pits, and subsequently used for ritual purposes, while, at the other extreme, it is likely the shafts from the Roman small town at Cambridge noted above were dug specifically to house their unusual deposits before being rapidly filled in. This would seem in many ways to cross the blurry divide between sacred site and secular settlement. Ritual shafts, especially in southern Britain, have long been studied (notably Ross 1968; M. J. Green 1986), and in a sense have come to be regarded as a sub-​class of sacred site, especially when apparently isolated, such as the 80-​ metre-​deep shaft at Findon in the Sussex Downs, which contained the remains of many dogs (Rudling 2008: 118). The extent to which other specifically dug ritual pits/​shafts can be regarded as shrines in themselves is uncertain, but perhaps we should not be so worried about contemporary interpretational classifications, when it is unlikely that such divisions were always made in the past. A recently excavated site at Bretton Way, Peterborough, is illustrative of such blurred divisions, as a substantial late-​third-​century aisled building lay close by to a contemporary pit lined with reused monumental stone blocks, which contained a large assemblage of pottery (including complete vessels) and animal remains (including a dog skull and antler), along with coins, personal items, and leather shoes (Pickstone 2011). The pit seemed to serve no other obvious purpose than to receive these deposits, and the use of the huge stone blocks in its construction implies a considerable undertaking, as would perhaps befit a special offering place to the gods.

The Character of ‘Special Deposits’ Within Settlements The character of votive deposits within sacred sites has already been discussed, with a noted concentration of coins and personal items, along with objects of an intrinsic religious nature. Although the first two categories at least do certainly appear as part of ‘special deposits’ within domestic contexts, they are generally much less common, and instead there is a far greater concentration of items such as pottery and in particular animal remains, often articulated. Of course, animal remains are also present in considerable quantities within many sacred sites, but their prevalence in domestic ‘special deposits’, and more importantly the (apparently) comparative lack of other typical temple offerings may suggest a different emphasis of ritual expression, perhaps one more rooted in indigenous practices. The widespread presence of dog skeletons in ‘special deposits’ is particularly noteworthy. In addition to the dog remains at Silchester, others in unequivocally ‘non-​ religious’ urban contexts include a fourth-​century pit within a suburb of Roman Winchester with a minimum of eight dogs along with a raven skeleton and ten complete or near complete colour-​coated beakers (Maltby 2010a). Also, a remarkable collection of over eighty mostly adult dogs were found within pits in ‘backyard’ enclosures in the Roman town at Dorchester (Dorset), interpreted as urban foundation deposits

654

654   Alex Smith (Woodward and Woodward 2004; though see Maltby 2010b, for a critique of this interpretation). In a rural context, the villa at Keston in Kent was associated with eight pits or shafts, many of which contained complete or partial dog skeletons, along with other animal parts and often complete pots (Philp 1999). Finally, in the context of military sites, the deposits within 107 deep pits and wells at the late-​first–​second century fort at Newstead in the Scottish borders have been studied extensively, and include quite a number of dog skulls, in addition to a huge range of other material, likely to have been deposited as part of a range of different ritual acts, no doubt for different purposes (Clarke, forthcoming). The interpretations of articulated animal remains within ‘secular’ contexts has recently been discussed by Morris (2011: 159–​162), who noted that the Romano-​British period had by far the largest number of dog ABGs (Associated Bone Groups) compared to other periods. Although functional interpretations for such deposits such as culling or simply dying of old age are used occasionally (see Maltby 2010b), Morris (2011: 162) asserts that most are explained in terms of ritual activity, although whether or not this implies deliberate sacrifice is often unstated. K. Smith (2006: 24) pointed out, however, that there does not necessarily have to be strict division between practical solutions such as disposing of culled or naturally dead animals and religiously motivated acts. If we assume that most of the dog ABGs found in pits at places such at Cambridge, Silchester, Winchester, Dorchester, Keston, and Newstead have been deposited as the result of ritual activity of some kind (notwithstanding the reservations of Morris (2011) and Maltby (2010b)), then the question arises as to why such a species was specifically selected, and why they were particularly associated with pits, wells, and shafts. In the Greco-​Roman world, dogs have known associations with both the underworld and healing, and a number of probable healing temples in Britain contained some association (for example, bones or imagery) with these animals (for example, Lydney Park, Pagans Hill, and Springhead). Black (2008: 2) has recently suggested that many of the Roman ritual shafts containing dog skeletons were dedicated to the Gallic deity Sucellos (or a local version thereof), patron of crafts and the underworld, though this is more difficult to substantiate across a wider spectrum of sites across Britain. There are, of course, many other animal species noted in ‘special deposits’, including all the main domesticates. However, other more unusual animals, not often recorded in temple assemblages, comprise wild birds and in particular corvids (ravens, crows, and so on). In fact, the combination of articulated canine and corvid remains noted above at Silchester and Winchester is one that is seen increasingly in other structured deposits at Iron Age and Roman sites across Britain (Sergeantson and Morris 2011). It is suggested that the unique character of these birds, including their tolerance of humans, scavenging habits, and their distinct voice, is what made them particularly suitable as messengers to the gods (Sergeantson and Morris 2011: 99). A similar link with humans and scavenging is, of course, also found with dogs, which may go some way to explain their association with corvids in structured deposits.

Ritual Deposition   655

Conclusions It has been the intention of this chapter to provide an overview of the practice of ritual deposition in Roman Britain, and it is immediately apparent that our understanding of some aspects has increased significantly over the past 15–​20 years, both because there have been a number of new excavations and/​or publications and also because a more contextual view has been taken of the evidence. Religious sites of all types, from extensive temple complexes such as Marcham/​Frilford to simple rural shrines like Snows Farm, and sites like Hallaton and Higham Ferrers where no actual temple buildings existed, have demonstrated the huge architectural variety in cult expression, albeit with a few more consistent underlying conventions. But, perhaps more importantly, detailed analyses of their finds composition and deposition is enabling us to start asking new questions of the types of rituals practised at these sites, and even something of the possible beliefs behind them. Studies on the ritual significance of specific object types and animal remains have demonstrated their potential to go beyond the label of mere ‘votive deposit’ to gain some insight into the possible motivations of the supplicant and perhaps even their use in the creation of cultural identities. In particular, it has been recognized that any understanding of the motives behind such rituals relies upon detailed analysis of both object and place. The most significant advance in our understanding of Romano-​British ritual expression since 2000 has been concerned with looking outside of the sacred site, with the recognition of ‘special deposits’ within settlement contexts. In some ways, given that it has often previously been acknowledged that ‘religion was everywhere’ in both the Iron Age and Roman worlds, it is surprising that the ubiquity of these deposits had not been widely commented upon before, yet such deposits are now interpreted, or at least speculated upon, in a great many recent excavation reports on Roman settlements. As with the ritual deposition observed within sacred sites, it is only through detailed contextual analysis of object and place, as, for example, has been attempted at Silchester and Newstead, that further understanding may be gained of the motivations behind such acts. With the increasing amount of data concerning such ‘special deposits’ within Romano-​British settlements, it is surely the time to conduct a more in-​depth regional or even national review of the evidence. Furthermore, comparative analysis in order to understand the relationship between such ‘special deposits’ within settlement contexts and those within formal temple sites has not yet been undertaken on a widespread systematic basis, but would surely be of great benefit in our comprehension of local and regional religious and cultural expression.

Acknowledgements I am very grateful to Professor Michael Fulford and Dr Hella Eckardt for reading and commenting upon earlier drafts of this chapter.

656

656   Alex Smith

References Primary Sources Diodorus Siculus, The Library of History, trans. H. Oldfather. Loeb Classical Library. Cambridge, MA: Harvard University Press, 1947. Gaius Julius Caesar, The Conquest of Gaul, trans. S. A. Handford. London:  Penguin Classics, 1951. Strabo, Geography, trans. H. L. Jones. Loeb Classical Library. Cambridge, MA:  Harvard University Press, 1917.

Secondary Sources Aldhouse-​Green, M. (2001). Dying for the Gods:  Human Sacrifice in Iron Age and Roman Europe. Stroud: Tempus. Alexander, J., and Pullinger, J. (2000). ‘Roman Cambridge: Excavations on Castle Hill 1956–​ 1988’, Proceedings of the Cambridge Antiquarian Society for 1999, Volume 88. Andrews, P., Biddulph, E., Hardy, A., and Brown, R. (2011). Settling the Ebbsfleet Valley: High Speed 1 Excavations at Springhead and Northfleet, Kent, Volume 1. Oxford and Salisbury: Oxford/​Wessex Archaeology. Bagnall-​Smith, J. (1995). ‘Interim Report on the Votive Material from the Romano-​Celtic Temple Sites in Oxfordshire’, Oxonensia, 60: 172–​204. Bagnall-​Smith, J. (2008). ‘Aspects of Votive Offerings in South-​East Britain’, in D Rudling (ed.), Ritual Landscapes of South-​East Britain. Oxford: Heritage Marketing and Publications Ltd, 153–​170. Barrie, T. (1996). Spiritual Path, Sacred Place:  Myth, Ritual and Meaning in Architecture. London: Shambhala. Black, E. (2008). ‘Pagan Religion in South-​East Britain: Contexts, Deities and Beliefs’, in D. Rudling (ed.), Ritual Landscapes of South-​East Britain. Oxford: Heritage Marketing and Publications Ltd, 1–​26. Booth, P., Dodd, A., Robinson, M., and Smith, A. (2007). Thames through Time: The Archaeology of the Gravel Terraces of the Upper and Middle Thames. The Early Historical Period ad 1–​1000. Oxford: Oxford Archaeology Thames Valley Landscapes. Bradford, J. P.  S., and Goodchild, R. G. (1939). ‘Excavations at Frilford, Berks 1937–​8’, Oxoniensia, 4: 1–​70. Bradley, R. (2000). An Archaeology of Natural Places. London: Routledge. Brunaux, J.-​L. (1988). The Celtic Gauls: Gods, Rites and Sanctuaries. London: Seaby. Clarke, S. (2001). ‘Probably Ritual:  Assemblage Interpretation at the Roman Military Complex—​Towards a More Holistic Approach’, in A. T. Smith and A. Brooks (eds), Holy Ground: Theoretical Issues Relating to the Landscape and Material Culture of Ritual Space. British Archaeological Report International Series 956. Oxford: Archaeopress, 73–​83. Clarke, S. (forthcoming). ‘Identifying the Actors and Motivations behind Pit Deposition at Trimontium Military Complex (Newstead)’, in J Morris (ed.), Ritual in Context: Explaining Ritual Complexity in Archaeology. Oxford . [Accessed 2012] Cunliffe, B. (1995). Danebury:  An Iron Age Hillfort in Hampshire. Volume 6:  A  Hillfort Community in Perspective. CBA Research Report 102. York: Council for British Archaeology.

Ritual Deposition   657 Derks, T. (1995). ‘The Ritual of the Vow in Gallo-​Roman Religion’, in J. Metzler et al. (eds), Integration in the Early Roman West: The Role of Culture and Ideology. Luxembourg: Dossier d’archéologie du Musée National d’Histoire et d’Art 4, 111–​127. Evans, C,. and Hodder, I. (2006). Marshland Communities and Cultural Landscape:  The Haddenham Project. Cambridge: McDonald Institute for Archaeological Research. Field, N., and Parker Pearson, M. (2003). Fiskerton: An Iron Age Timber Causeway with Iron Age and Roman Votive Offerings: The 1981 Excavations. Oxford: Oxbow Books. Fulford, M. (2001). ‘Pervasive “Ritual” Behaviour in Roman Britain’, Britannia, 32: 199–​218. Fulford, M., and Clarke, A. (2011). Silchester: City in Transition. The Mid-​Roman Occupation of Insula IX c. ad 125–​250/​300. A Report on the Excavations since 1997. Britannia Monograph Series 25. London: Society for the Promotion of Roman Studies. Fulford, M., Clarke, A., and Eckardt, H. (2006). Life and Labour in Late Roman Silchester: Excavations in Insula IX from 1997. Britannia Monograph Series 22. London: Society for the Promotion of Roman Studies. Ghey, E. (2007). ‘Empty Spaces or Meaningful Places? A Broader Perspective on Continuity’, in R. Haeussler and A. King (eds), Continuity and Innovation in Religion in the Roman West: Volume 1. JRA Supplementary Series 67. Portsmouth, RI: Journal of Roman Archaeology, 19–​30. Green, C. W. (1966). ‘A Romano-​Celtic Temple at Bourton Grounds, Buckingham’, Record of Bucks, 17/​5: 356–​366. Green, M. J. (1986). The Gods of the Celts. Gloucester: Sutton Publishing. Green, M. J. (1995). ‘The Gods and the Supernatural’, in M. J. Green (ed.), The Celtic World. London: Routledge, 465–​489. Haselgrove, C. (2005). ‘A Trio of Temples: A Reassessment of Iron Age Coin Deposition at Hayling Island, Harlow and Wanborough’, in C. Haselgrove and D. Wigg-​Wolf (eds), Iron Age Coinage and Ritual Practices. Mainz am Rhein: Akademie der Wissenschaften und der Literatur, 381–​418. Hayden, C., Booth, P., Dodd, A., Smith, A., Laws, G., and Welsh, K. (forthcoming). Horcott Quarry: Prehistoric, Roman and Anglo-​Saxon Settlement and Burial. Oxford: Oxford Archaeology. Hill, J. D. (1995). Ritual and Rubbish in the Iron Age of Wessex. BAR British Series 242. Oxford: Archaeopress. Hingley, R. (2006). ‘The Deposition of Iron Objects in Britain during the Later Prehistoric and Roman Periods: Contextual Analysis and the Significance of Iron’, Britannia, 37: 213–​257. Kamash, Z., Gosden, C., and Lock, G. (2010). ‘Continuity and Religious Practices in Roman Britain:  The Case of the Rural Religious Complex at Marcham/​Frilford, Oxfordshire’, Britannia, 41: 92–​126. Kiernan, P. (2009). Miniature Votive Offerings in the Roman North-​West. Mainz: Harrassowitz. King, A. (2005). ‘Animal Remains from Temples in Roman Britain’, Britannia, 36: 329–​369. King, A. (2007a). ‘Coins and Coin Hoards from Romano-​Celtic Temples in Britain’, in R. Haeussler and A. King (eds), Continuity and Innovation in Religion in the Roman West: Volume 2. JRA Supplementary Series 67. Portsmouth, RI: Journal of Roman Archaeology, 25–​42. King, A. (2007b). ‘Characterising Assemblages of Votive Offerings at Romano-​Celtic Temples in Britain’, in M. Hainzmann (ed.), Auf den Spuren keltisher Gotterverehrung, Vienna: Verlag der Osterreich. Akademie der Wissenschaften, 183–​196. Lawrence, S., and Smith, A. (2008). Between Villa and Town:  Excavations of a Roman Roadside Settlement and Shrine at Higham Ferrers, Northamptonshire. Oxford Archaeology Monograph 7. Oxford: Oxford Archaeology.

658

658   Alex Smith Leech, R. (1986). ‘The Excavation of a Romano-​Celtic Temple and a Later Cemetery on Lamyatt Beacon, Somerset’, Britannia, 17: 259–​328. Lewis, M. J. T. (1966). Temples in Roman Britain. Cambridge: Cambridge Classical Studies. Maltby, M. (2010a). Feeding a Roman Town:  Environmental Evidence from Excavations in Winchester, 1972–​1985. Winchester: Winchester Museums. Maltby, M. (2010b). Zooarchaeology and the Interpretation of Depositions in Shafts’, in J. Morris and M. Maltby (eds), Integrating Social and Environmental Archaeologies:  Reconsidering Deposition. Oxford: BAR British Series 2077. Oxford: Archaeopress, 24–​32. Mattingly, D. (2006). An Imperial Possession:  Britain in the Roman Empire, 54 bc–​ad 409. London: Penguin. Medlycott, M. (2011). The Roman Town of Great Chesterford. Chelmsford:  East Anglian Archaeology 137. Miles, D., Palmer, S., Smith, A., and Perpetua Jones, G. (2007). Iron Age and Roman Settlement in the Upper Thames Valley: Excavations at Claydon Pike and other Sites in the Cotswold Water Park. Oxford Archaeology Thames Valley Landscapes Monograph 26. Oxford: Oxford Archaeology. Morris, J. (2011). Investigating Animal Burials: Ritual, Mundane and Beyond. BAR British Series 535. Oxford: Archaeopress. Penn, W. S. (1962). ‘Springhead: Temples II and V’, Archaeolgia Cantiana, 77: 110–​132. Philp, B. (1999). The Roman Villa Site at Keston, Kent:  Second Report. Canterbury:  Kent Archaeological Rescue Unit. Pickstone, A. (2011). Iron Age and Roman Remains at Bretton Way, Peterborough:  Post-​ Excavation Assessment and Updated Project Design. Oxford Archaeology East Unpublished Report. Powell, K., Smith, A., and Laws, G. (2010). Evolution of a Farming Community in the Upper Thames Valley. Excavation of a Prehistoric, Roman and Post-​Roman Landscape at Cotswold Community, Gloucestershire and Wiltshire. Oxford Archaeology Thames Valley Landscapes Monograph 31. Oxford: Oxford Archaeology. Puttock, S. (2002). Ritual Significance of Personal Ornament in Roman Britain. BAR British Series 327. Oxford: Archaeopress. Revell, L. (2007). ‘Religion and Ritual in the Western Provinces’, Greece and Rome, 54: 210–​228. Rodwell, W. (1980) (ed.). Temples, Churches and Religion in Roman Britain. BAR British Series 77. Oxford: Archaeopress. Rives, J. B. (2007). Religion in the Roman Empire. Oxford: Blackwells. Ross, A. (1968). ‘Shafts, Pits, Wells—​Sanctuaries of the Belgic Britons?’, in J. M. Coles and D. D. A. Simpson (eds), Studies in Ancient Europe. Leicester: Leicester University Press, 255–​285. Rudling, D. (2008). ‘Roman-​Period Temples, Shrines and Religion in Sussex’, in D. Rudling (ed.), Ritual Landscapes of South-​East Britain. Oxford: Heritage Marketing and Publications Ltd, 95–​138. Score, V. (2011). Hoards, Hounds and Helmets:  A  Conquest-​Period Ritual Site at Hallaton, Leicestershire. Leicester Archaeology Monograph 21. Leceister: Leicester Archaeology. Sergeantson, D., and Morris, J. (2011). ‘Ravens and Crows in Iron Age and Roman Britain’, Oxford Journal of Archaeology, 30/​1: 85–​107. Smith, A. T. (2001). The Differential Use of Constructed Sacred Space in Southern Britain from the Late Iron Age to the 4th Century ad. BAR British Series 318. Oxford: Archaeopress. Smith, K. (2006). Guides, Guards and Gifts to the Gods:  Domesticated Dogs in the Art and Archaeology of Iron Age and Roman Britain. BAR British Series 422. Oxford: Archaeopress.

Ritual Deposition   659 Wait, G. A. (1985). Ritual and Religion in Iron Age Britain. BAR British Series 149 (i and ii). Oxford: Archaeopress. Watts, L., and Leach, P. (1996). Henley Wood: Temples and Cemetery Excavations 1962–​69. CBA Report 99. York: Council for British Archaeology. Wilson, D. R. (1975). ‘Romano-​Celtic Temple Architecture’, Journal of Roman Archaeological Association, 38 [3rd Series], 3–​27. Woodward, A. (1992). Shrines and Sacrifice. London: Batsford. Woodward, A., and Leach, P. (1993). The Uley Shrines:  Excavation of a Ritual Complex on West Hill, Uley, Gloucestershire:  1977–​ 9. English Heritage Archaeological Report 17. London: English Heritage. Woodward, P., and Woodward, A. (2004). ‘Dedicating the Town: Urban Foundation Deposits in Roman Britain’, World Archaeology, 36: 68–​86. Wythe, D. (2007). ‘An Analysis of Coin Finds from 75 Roman Temple Sites in Britain’, in R. Haeussler and A. King (eds), Continuity and Innovation in Religion in the Roman West: Volume 2. JRA Supplementary Series 67. Portsmouth, RI: Journal of Roman Archaeology, 43–​66. Zoll, A. (1995). ‘A View through Inscriptions: The Epigraphic Evidence for Religion at Hadrian’s Wall’, in J. Metzler et al. (eds), Integration in the Early Roman West: The Role of Culture and Ideology. Luxembourg:  Dossier d’archéologie du Musée National d’Histoire et d’Art 4, 129–​150.

660

Chapter 32

Christia ni t y i n Rom an B ri ta i n David Petts

Introduction The study of early Christianity has tended to assume a rather peripheral position with regard to the wider study of ritual, religion, and cult behaviour in Roman Britain. It is often dismissed as a relative latecomer to the religious landscape of the province, with its true importance developing only in the early Middle Ages. The wider debate about its importance is often couched in the terms of ‘success’ or ‘failure’, judgemental terms that are rarely used in discussion of other Roman religious traditions, such as Mithraism. It also often implied that the spread of the church in Britain was far less extensive than in other part of the empire, although often these comparisons are with very different parts of the empire, and often with different periods (for example, comparing fourth-​century Britain with fifth-​or sixth-​century southern Gaul). While a proper understanding of Roman Christianity certainly presents some specific interpretational challenges, the often oversimplistic comparisons between Christianity and other religious traditions in Roman Britain often ignore some of the wider conceptual issues presented by the archaeological study of religious conversion. This chapter attempts to provide a brief overview of the extant evidence for the early church, but also hopes better to position Roman Christianity in its temporal and social landscape in fourth-​and fifth-​century Britain.

Early Christianity In ad 313, with the Edict of Milan, Christianity finally achieved full recognition as a licit religion within the Roman Empire. Although it was not until ad 380 that Theodosius

Christianity in Roman Britain    661 issued the Edict of Thessalonica making Christianity the state religion of the empire, its adoption by the House of Constantine meant that from that point onwards it became the de facto official cult of the Roman emperors (excluding a brief return to paganism under Julian). During the intermittent persecutions suffered by the church in previous centuries there was a pressure on the church to maintain a relatively low profile within Roman society. It is likely that only in Rome and some of the larger towns of the eastern Mediterranean were there congregations of significant size. This had the consequence that the church was slow to develop a distinctive set of architectural forms, and there was relatively little Christian public monumentality (Snyder 2003). This is not to suggest that there was no investment in Christian material culture, but it tended to be artistic endeavours limited to spatially controlled spheres, such as the burial catacombs in Rome or the third-​century house church at Dura Europos (Baur and Hopkins 1934; Bowes 2008). In both cases, the Christian symbolism is largely confined to wall-​paintings, a cautionary reminder of the extent to which evidence for pre-​fourth-​century Christian worship may have been lost. The practical impact of the Edict of Milan was the emergence of the church from the shadows, as the newly confident institution had both the legal right and, increasingly, the economic resources to develop a high profile within Roman society. For the first time it could invest extensively in large-​scale public expressions of faith, while, at an individual level, it was only then that it became safe to identify oneself as a Christian through personal items of material culture. This new freedom is reflected in both the increase in textual sources for Christianity in the fourth century, and also a massive rise in the quantity and distinctiveness of Christian art and architecture. It is not surprising, given this major change of circumstances, that the church groped for new ways of expressing its identity, and that, owing to the relative lack of existing Christian modes of representation, it drew more widely on forms and symbols used by other religious and social groups. For example, although the basilica became one of the dominant forms of church plan, it was an architectural blueprint that was widely used by existing secular and religious groups. This relationship between Christian modes of worship and pagan religious life is one that will be returned to at the end of this chapter.

Documentary Evidence Contemporary documentary evidence for pre-​Constantinian Christianity in Roman Britain is limited. Although Christian communities in Britain were mentioned by early church writers, such as Tertullian and Origen (Thomas 1981:  43–​44), these occur within very general passages expressing the geographical extent of the church. As such, these mentions need to be accepted with caution as prima facie evidence for Christianity in early Roman Britain. There are, though, hints from later textual sources that there may have been some Christians in second-​and third-​century Britain. The

662

662   David Petts early sixth-​century British writer Gildas refers to the martyrdom of Aaron and Julius at Caerleon, the site of a Roman legionary fortress (De Excidio 10.2), although it is not clear from where he was drawing his information (Knight 2001). However, it is clear that he derived his information on the martyrdom of St Alban from an earlier Passio Albani, possibly dating to between ad 430 and 480 (Sharpe 2001). There has been considerable debate about the dates of the martyrdoms of Aaron, Julius, and Alban; given the paucity of evidence, it is not possible to reach firm conclusions beyond a broad third-​or early fourth-​century date. The records of the Council of Arles, called by Constantine in ad 314 to settle the Donatist dispute, is the earliest clear evidence for a formally constituted Christian church in Britain (Rivet and Smith 1979: 49–​50). The Acta Concilii Arelatensis lists the British delegates: Eborius episcopus de civitate Eboracensi provincia Brittania Restitutus episcopus de civitate Londiensi provincia suprascripta Adelphius episcopus de civitate Colonia Londiniensium Exinde Sacerdos presbyter Arminius diaconus. [Eborius, Bishop of the city of York of the province of Britain. Resitutus, Bishop of the the city of London of the above named province. Adelphius, bishop of the city of Colonia Londinensium, Furthermore, Sacerdos the priest and Arminius the deacon.] (Munier 1963: 15, ll. 54–​58)

The first two names record the presence of bishops in York and London. The identification of Colonia Londiniensium is less certain; Lincoln and Colchester have been suggested, with the former the more likely. If the bishops listed represent the church leadership in three of the four provinces of Roman Britain (York—​Flavia Caesariensis; London—​Maxima Caesariensis; Lincoln—​Britannia Secunda), it is possible that Sacerdos and Arminius may have been delegates from the fourth province, Britannia Prima, which perhaps lacked a bishop at this point. British bishops were clearly well integrated into the wider Christian church in the fourth century, and they are recorded at church councils at Serdica (ad 343) and Ariminum (ad 359) (Thomas 1981: 121). There were also external interventions into the running of the British church. Victricius, bishop of Rouen, crossed the channel at some point in the ad 390s to deal with an unspecified dispute (De Laude Sanctorum 443–​444). Beyond this, there is very little explicit textual evidence for fourth-​century Christianity in Britain, although the careers of some administrators clearly indicate that they were Christian—​the late fourth-​or early fifth-​century vicar of the diocese, Chrysanthus, became a Bishop of Constantinople, a post that had earlier been held by his father (Salway 1981: 407–​408). This gives us a picture of a Romano-​British church with a developed ecclesiastical leadership who were active in church affairs beyond the confines of its own diocese. However, we know little about the extent to which this leadership changed and

Christianity in Roman Britain    663 developed over the fourth century. While the evidence from the Council of Arles indicates that there was a hierarchy at provincial level, it is not clear whether the named individuals were metropolitans with subsidiary bishops, possibly at civitas level. The names of bishops have been tentatively identified as inscribed on a number of artefacts, including a silver dish (the so-​called Risley Park lanx) that mentions a bishop named Exuperius; a lead salt-​pan from Shavington, which records a possible bishop by the name of Viventius, and, less certainly, an unnamed bishop on a pewter bowl from the Isle of Ely. In none of these cases is it possible to localize their spheres of influence (Clarke 1931; Johns 1981; Johns and Painter 1991; Penney and Shotter 1996).

Christianity and Material Culture There is evidence for the use of Christian imagery on a wide range of objects and artefacts from Roman Britain. However, it is not always easy to move from the simple identification of a symbol with Christian associations to an identification of it as an object connected to Christian belief or practice. This knotty problem can be seen most clearly in the use of the chi-​rho symbol. This basic symbol, bringing together the first two letters of the Greek word Christos, had clear Christian associations from at least the early fourth century, and was regularly associated with Constantine’s vision on the eve of the Battle of the Milvian Bridge. The image appears in two forms in Britain, the ‘Constantinian chi-​rho’ and the simpler rho-​cross (lacking the chi, but with a cross against the vertical stroke of the rho). The motif is found alone, but also regularly flanked by the Greek alpha and omega (an allusion to Revelations 1: 8). The ‘Constantinian’ chi-​rho is more common in a Romano-​British context, with the rho-​cross beginning to appear only in the later fourth century (Thomas 1981: 86–​91; Pearce 2008: 197–​201). The chi-​rho can regularly be found on objects of indisputable religious or ritual function, such as the votive leaves from the Water Newton treasure (Painter 1999). However, following the Edict of Milan, the chi-​rho quickly became associated with emperors in general and the House of Constantine in particular. It is regularly found being used in seemingly official contexts during the fourth century. The most widespread example of this was on coinage, particularly bronze nummi minted in Amiens by Magnentius in the mid-​fourth century (Moorhead 2000). It also appears stamped on pewter ingots (examples have been found in London) and a silver ingot found in Balline, County Limerick, Ireland, but of Roman origin (Ó Ríordáin 1947: 43–​53; Mawer 1995:  96–​98). In these cases we seem to find the chi-​rho being used as an indicator that the ingots came from mines under some form of imperial control. A small lead seal with a chi-​rho from the forum Silchester may have been used in a similar manner (Joyce 1881: 363). The symbol is also found on a number of personal items, including a group of finger rings (Fifehead Neville, Dorset; Thruxton, Hampshire; Richborough, Kent; Bagshot, Surrey) (Middleton 1881–​3: 68; Engleheart 1922: 215; Cunliffe 1968: 98–​ 99; Graham 2002). It is not possible to identify their owners, although the items could

664

664   David Petts clearly have been used by private individuals, state functionaries, and members of the ecclesiastical establishment. Christian imagery may also be found on other items, such as a group of buckles, belt fittings, and strap-​ends of a type likely to have been used by state officials, including members of both the army and the imperial civil service. The chi-​rho does not appear on these items—​except for a strap-​end from Sandy, Bedfordshire (Mawer 1995: 65); instead they carry images of peacocks, which had associations with Christianity in late antiquity: examples include the belt buckles from Pen-​y-​Corddyn, Clwyd; Tripontium, Warwickshire; Stanwick, North Yorkshire; and East Challow, Oxfordshire (Sanderson 1993: 2; Hawkes 1973: 145–​159, ­figure 3.1; Collingwood Bruce 1880: 90, fi ­ gure 8; Henig and Brown 2003). While this general form of belt buckles is widely found in the western empire, the use of the opposed peacock motif does appear to be a feature found almost exclusively in Britain, with only one other example known, from Westerwanna, Germany. These items were all made from copper alloy; there are few examples of the use of more precious metals in items with a possible Christian identity. The exceptions are the large silver buckle from the Traprain Law hoard and an unusual bow brooch of doubtful authenticity from Shepton Mallet (Johns 2001). However, a series of far more extravagant groups of objects with a probable Christian association has been recovered from fourth-​century contexts as elements in high-​value silver hoards. The assemblage with the most extensive range of Christian symbolism is the Water Newton hoard (Painter 1977). This consisted of twenty-​seven silver objects and a gold disc, including sixteen votive leafs and a range of cups, jugs, and other vessels. One cup carried the inscription innocentia et viventia … runt (Innocential and Viventia dedicated/​offered?) and a chi-​rho flanked by an alpha and omega, while another bore a short text accompanied by a chi-​rho that has resonances with elements of early liturgy (Painter 1999). The hoard probably dates from the second half of the fourth century, and it has been argued that this may have been a collection of ecclesiastical plate and ritual items, although the reason for its deposition is unclear. The hoard from Traprain Law also had a significant component of artefacts with Christian imagery, including the buckle already mentioned, a strainer with a chi-​rho design, a silver flask with biblical imagery, and a silver flask with a chi-​rho (Curle 1923; Painter 1999; 2010). Although the circumstances for its assemblage and deposition were doubtless very different from the Water Newton hoard, both groups attest to the potential wealth that could be amassed by Christian communities in Britain. Christian imagery was also found, on a smaller scale, in other major plate hoards from Britain, including ten spoons adorned with the chi-​rho from the early fifth century Hoxne hoard and two spoons decorated with a chi-​rho from a hoard found in Canterbury (Johns and Potter 1985; Johns 2010). These large silver hoards with Christian components are part of a wider fourth-​century tradition in Britain of silver hoards, with others, such as the Thetford treasure, containing only pagan imagery (Johns and Potter 1983). However, they are reminders that both pagan and Christian communities in late Roman Britain were able to amass substantial wealth in the form of silver plate on a scale largely unparalleled elsewhere in the empire at this period.

Christianity in Roman Britain    665

Figure 32.1  Silver ring with early Christian symbol (anchor and fish) from the Roman fort at Binchester, County Durham. Source: © Durham University.

In addition to the use of the chi-​rho, a range of other early Christian images are known from Romano-​British objects, such as the combination of fish and anchor found on the intaglio of a silver ring discovered in 2014 excavated at the Roman fort at Binchester (County Durham) (Figure 32.1).

Churches Despite clear documentary evidence of an established Christian church in fourth-​century Britain, the archaeological evidence for this is variable. The most problematic area

666

666   David Petts is in the identification of church structures. There is not a single building of Roman date that can unequivocally be identified as a Christian church, although there are around a dozen structures that have a greater or lesser degree of circumstantial evidence. One of the major challenges is that, as noted above, the most common ground plan for churches in the western empire in the fourth century was the basilica, a form that was also widely used in other religious and secular contexts. It is not easy, in the absence of additional evidence, to identify a structure as a church on morphological grounds alone. In the absence of an extensive late Roman tradition of epigraphy, and the generally poor preservation of most of these structures, good supporting evidence is largely lacking. One of the strongest candidates for a Roman church is the earliest of a sequence of structures on the site of the medieval parish church of St Paul-​in-​the-​Bail, Lincoln (Steane 2006:  129–​211). This was located above the central area of the forum of the Roman town. Beneath the medieval church were three earlier structures. The earliest was a rectangular building aligned west–​east with a potential rectangular apse at the eastern end. Above this was a larger timber structure with a substantial eastern apse. Although the western end of neither of these buildings was found, it is possible that they abutted the western portico of the forum, which may have acted as a simple narthex. Above this was a smaller rectangular building with a central burial of probable middle Saxon date. The dating evidence for this sequence is not precise, with the first two possible churches belonging anywhere between the very late Roman and the middle Saxon period on stratigraphic grounds (Steane 2006: 192–​194). The second building was also surrounded and cut by a series of burials that have a broad seventh/​eighth-​century ad date, which matches well with the date of the burial within the third structure. One of the most frequently adduced cases is a small (13 m × 9 m) apsidally-​ended basilica in Insula IV at Silchester, Hampshire (Frere 1976; King 1983; Ford 1994; Cosh 2004) (Figure 32.2). This building lies to the south-​east of the town’s main forum and basilica complex. Internally, there are two aisles and an eastern narthex with the apse lying at the western end of the structure. Most aspects of this building find parallels with churches elsewhere in the western empire; for example, the western apse echoes the plan of the early fourth-​century church of St Severin in Cologne (Krämer 1958; Frere 1976: 292). However, the major stumbling block for a certain Christian identity for this building is its date. It almost certainly post-​dates a nearby wooden structure that probably fell out of use in the late third century. If the probable pre-​Constantinian date for this structure militates against identification of the building as a church, then possible alternatives include use as a schola. In an urban context, another reasonable candidate for a church is the large-​aisled building of mid-​fourth-​century date found at Colchester House near Tower Hill, London (Sankey 1998). Although preservation was poor, excavation revealed the north-​ eastern corner of a basilica structure with double aisles and a short wall marking off the eastern end of the building. This was a significant structure; substantial in size (c. 50 m × 80–​100 m) with a stone and tile floor. Fragments of marble and window glass were also found, although in secondary contexts. The reconstructed plan of this building shows some parallels with the fourth-​century cathedral of St Tecla in Milan and the church

Christianity in Roman Britain    667

0

Forum

10m

street timber building

Well Church?

Figure  32.2 Potential Roman church from Silchester, Hampshire. Based on Frere (1975: ­figure 1). Source: © David Petts.

of St Giovanni Evangelista in Ravenna (Krautheimer 1986: 84, 184–​185). Another possible function for this building is as a granary for the collection of the annona, although the evidence for the use of marble in the interior may mitigate against this suggestion. Intriguingly, a recent analysis of the coin assemblage from the structure suggests that the coin loss profile has its best parallels with late Romano-​British religious and ritual sites (Gerrard 2011). Within the context of Romano-​British towns, these are the strongest candidates for churches, although other suggestions have been made, including the apsidal building at Flaxengate, Lincoln; the Roman building beneath the medieval church of St Mary-​de-​Lode, Gloucester, and the apsidal structure in the late Roman cemetery at Butt Road, Colchester. All, however, have only limited and circumstantial evidence for their religious use (Colyer and Jones 1979; Bryant 1980; Crummy et al. 1993: 164–​190; Heighway 2010: 42–​43). There are also a small number of possible Roman churches from rural contexts. The most likely example is the small rectangular structure associated with a group of over forty burials and a possible font from Icklingham, Suffolk (West and Plouviez 1976). The association of the structure with the burials is suggestive, since Roman pagan temples were rarely located adjacent to graves. More important, however, is the fact that three lead tanks carrying overt Christian motifs have been found in the immediate area. While a strong case can be made for the building being a small church or oratory,

668

668   David Petts this identification is again made only on the basis of its contextual associations rather than anything inherent in the form or plan of the building itself. A number of other small rectangular structures found adjacent to Roman rural temples have also been put forward as potential churches, possibly replacing an earlier, pagan, focus of worship, including examples from Lamyatt Beacon, Somerset, Brean Down, Somerset, and Uley, Gloucestershire (Ap Simon 1964–​5; Leech 1986; Woodward and Leach 1993). While they all have some features in common with early churches, the strongest parallels are with early medieval rather than Roman churches, and, in the absence of any internal dating evidence for most of these structures, an identification as a Roman church is hard to endorse. Finally, there are a small group of structures on Roman military sites that may be churches, although, like other examples, they suffer from problems in interpretation. A good case in point is the rectangular structure at the major fort at Richborough, Kent (Brown 1971). Located in the north-​east corner of the fort, this part of the site was excavated using relatively unsophisticated techniques in the 1920s. This work revealed two lines of possible stone post-​pads, which appeared to form the west and north walls of a structure. The structure is poorly dated, so there would be no suggestion that this was a church if it was not for its proximity to a probable masonry stone font that lay between the structure and the northern wall of the fort. On the northern frontier, there are a number of potential churches. A simple apsidally ended building was constructed within the courtyard of the praetorium at Vindolanda. Dating evidence is sparse, although it was clearly constructed over a build-​up of activity that overlay the flagstones of the courtyard; a date of early fifth century or later is most likely (Birley 2009). There is also another small apsidal building at nearby Housesteads (Crow 1995: 95–​97). As at Richborough, it lay within the north-​eastern corner of the fort, although the apse was at the west rather than the east end. It was discovered during Bosanquet’s excavations in the late nineteenth century, and its precise date is unclear, although it appears to be broadly late Roman. Slightly different in form is the putative church at the major fort and supply base to the south of the River Tyne at South Shields, Tyne and Wear (Bidwell and Speak 1994: 103–​ 104). This structure lay within the principia of the fort and consists of a possible altar surrounded on three sides by a stone structure, which might best be understood as elements of a rectangular apse. This building was quite substantial in size, measuring over 9 metres in width, and there is the potential that the main body of this structure could have been linked to part of the colonnade surrounding the courtyard. A final form of church that needs to be considered is the ‘house church’. The dedication of elements of private houses on a temporary or permanent basis to the celebration of the Eucharist has its origin in the earliest period of Christian practice (Bowes 2008). However, it is apparent that this practice continued into the post-​Constantinian era and that some villas may have had what were, in practice, private estate chapels. The best example from this in Britain is Lullingstone, Kent (Meates 1979, 1987). Here, exceptional preservation allowed a series of plaster wall-​paintings from an upper room in building to be reconstructed. The anti-​room or narthex was decorated with chi-​rhos flanked by

Christianity in Roman Britain    669

Figure 32.3  Painted wall plaster from Lullingstone Roman villa (Kent) showing chi-​rho symbol flanked by an alpha and omega. Source: © The Trustees of the British Museum.

an alpha and omega (Figure 32.3). Similar symbols appeared within the main room, as did a sequence of figures in the orans position (hands raised up), the position associated with prayer in the early church. At the same time as the paintings were created, it appears that the access routes to the rooms were changed and access, which had previously been via the main house, became possible only via an exterior door (Meates 1987). It is salutary to note that, without the survival of the wall-​paintings, there is nothing that would have suggested a Christian use or identity to these rooms. Lullingstone is not the only Roman villa that has produced decorative schemes with a clearly Christian identity. A number of villas from Britannia Prima (south-​west England) have produced mosaics that incorporate possible Christian symbolism. The best known of these is the central roundel from the large pavement from Hinton St Mary, Dorset, which shows a male head with a chi-​rho symbol behind it (Pearce 2008). A contiguous mosaic showed hunting scenes and Bellerophon fighting the Chimaera. The male bust has often been identified as an early (if not the earliest) figurative depiction of Christ (Toynbee 1964; Painter 1967, 1972; Thomas 1981). However, there have also been strong arguments made that the image is in fact a piece of imperial portraiture, intended to represent a member of the House of Constantine (Pearce 2008). The chi-​rho

670

670   David Petts also appears on a mosaic from Frampton, Dorset. As at Hinton St Mary, this was associated with a depiction of Bellerophon battling the Chimaera with the chi-​rho decorating the adjacent apse (Cosh and Neal 2006: 130–​140). From the evidence for urban, rural, and military churches, it is possible to make a number of general observations. There was the apparent change in the way in which late Roman formal space was being utilized by the end of the fourth century ad. Whether or not one accepts the Christian attribution to the buildings at St Paul-​in-​the-​Bail, Lincoln, Vindolanda, and South Shields, they were all located within the centre of what had formally been open courtyards. There was also a clear move away from architectural expressions of power drawn from Mediterranean architecture, which had an emphasis on courtyards and open spaces. It is also notable how little good dating evidence there is for these structures, which were often excavated with unsophisticated techniques, so that detailed stratigraphic insight is frequently lacking. Even when excavations were to modern standards, the lack of good diagnostic material culture of a fifth-​century date makes it extremely difficult to distinguish between buildings that were built broadly in the fourth century and those that were first constructed in the fifth century ad. It is also important to consider the British structures in their wider context. In the western Roman Empire very few churches can be securely dated to the fourth rather than the fifth century, and thus the evidence from Britain is, in fact, broadly comparable with developments in neighbouring provinces.

Baptism Churches are not the only built structures associated with early Christianity in Britain. A number of masonry structures that may be fonts or font bases are known. While the evidence for the church at Richborough may not be strong, the hexagonal masonry structure to its north is almost certainly a font (Brown 1971). This stone-​built basin has strong parallels with other certain examples from Boppard and Cologne, Germany (Brown 1971: 227–​228, plate XXXI). An octagonal example, built of tile rather than stone with a gravel drain or soakaway, has been excavated at Chimneys, Witham, Essex; this was found on the site of a Romano-​Celtic style temple built in the third century ad (Turner 1999: 51–​ 55). However, it appears to post-​date the phase of activity at the temple and may represent the Christianization of a pagan focus of worship. Unlike the Richborough example, there are hints of an external superstructure surrounding the font, with a series of post-​holes placed at each corner of the tank; this was succeeded by a later reworking to convert it into a four-​post structure. Another masonry cistern-​type font comes from Icklingham, Suffolk, where a small tile-​built tank was found close to the possible small church already discussed (West and Plouviez 1976: 71). A final possible example is the small stone-​lined tank that stood to the west of the possible church at Housesteads (Crow 1995: 96–​97). Although it is rectangular in form, similar simple forms are known from the continent, such as at Zurzach, Switzerland (Brown 1971: 228, plate XXXII).

Christianity in Roman Britain    671 In addition to complete font cisterns, there are also a number of potential font bases, the footings or platforms on which smaller tanks could have stood. Tile platforms, on an axial alignment with putative church structures, stood to the east of the small basilica at Silchester already mentioned and to the west of the apsidal structure in the cemetery at Butt Road, Colchester (Frere 1976: 290–​291; Crummy et al. 1993: 176–​177). Obviously, the case for accepting these as font bases is closely connected to the acceptance of the related buildings as churches. It is possible that some ornamental pools or cisterns from Roman villas may also have had a baptismal or at least quasi-​ritual function. At Chedworth, Gloucestershire, a small apsidal pool to the north-​west of the main building was fed by natural springs (Goodburn 2000: 24). Although constructed in the pre-​Constantinian period, it may have had a later Christian use, as a number of chi-​rho symbols were found carved onto stone slabs that probably originally lined the pool. It has been suggested that a group of octagonal pools from villas in south-​west England (Dewlish, Dorset; Lufton, Somerset; Holcombe, Devon) may also have had a similar ritual function (Perring 2003; Todd 2005). They were all located within existing bath-​suites but also had separate external access. However, although there is evidence for Christianity on mosaics and other material culture from villas in the general area, this functional identification of the pools must remain speculative (Henig 2006). An important class of objects is a group of large circular lead tanks some of which carry overt Christian imagery, with over twenty examples known. Circular in shape and constructed from sheets of cast lead, they vary in size from 0.46 metres to 0.97 metres in diameter. A number of these have been decorated with chi-​rho symbols (for example, Ashton, Northants; Pulborough, West Sussex; Icklingham, Suffolk), in some cases associated with the Greek letters alpha and omega. Two also carry figural images. A tank from Flawborough, Nottinghamshire, has four figures in the orans posture, while the most complex is a figurative frieze on a tank from Walesby, Lincolnshire. Within an architectural frame it shows three groups of figures; the central group consists of two clothed women flanking a naked woman, while the other two groups depict clothed males in cloaks and tunics. This has often been interpreted as the depiction of a baptism, with the naked catechumen being assisted to the font by two supporters (Thomas 1981: 221–​225; though see now Crerar 2012). These are usually associated with baptism, despite the more widespread tradition of baptism by immersion, and seen either as fonts or vessels associated with other elements of the baptismal liturgy, such as ritual footwashing (pedilavium) (Watts 1991: 171–​173). However, solid evidence for their precise function is far from clear. Although a small number of broadly similar lead tanks have been found outside Britain, these have not been found carrying Christian imagery, and they do seem to be a genuinely insular phenomenon. An important dimension to the distribution of the lead tanks is their specific depositional context. Many appear to have been placed in ‘watery’ contexts, including ditches (Flawborough), wells (Caversham; Ashton), rivers (Pulborough; Oxborough; Huntingdon), and water holes (Heathrow). This appears to parallel a long-​term northern European tradition of utilizing such contexts for acts of ritualized deposition. In

672

672   David Petts Roman Britain, particularly, there is a clear tradition of placing hoards of pewter vessels and tableware in similar locations, a rite that becomes increasingly popular from the later third century ad (Petts 2002). There is also evidence for regional focus to this rite, with the deposition of both lead tanks and other suites of items being most common in the east Midlands and East Anglia. This can even be seen scaled down to the local level, such as in the area around Icklingham, which has produced evidence for four lead tanks, as well as a cluster of other hoards, including coins and deposits of ironwork, with evidence that this local tradition had its origins in the early Roman period (Petts 2002).

Death and Burial As with other aspects of fourth-​century Christian belief, identifying a diagnostic Christian burial rite in the late Roman period is challenging. There are undoubtedly major changes in the late Romano-​British burial rite from the third century onwards (Philpott 1991; Quensel-​von Kalbern 1999; Petts 2004). Broadly speaking, there is a transition from cremation to inhumation as the predominant rite, although within this general pattern there is certainly considerable regional variation. The inhumation rite is itself far from homogenous, but it is possible to recognize two general traditions. One rite continued to see the deposition of a range of grave goods, including vessels and items of personal ornamentation. These burials were often aligned north–​south, and there was considerable variation in the positioning of the body, with some unusual treatments of the corpse, including prone burial and post-​mortem decapitation. The spatial arrangement of these burials within a cemetery often showed evidence for clustering in possible family groups or around focal graves. This rite contrasted with a tradition in which grave goods were largely absent. These graves were usually aligned approximately west–​east, and the graves within cemeteries were often laid out in regular rows. There is also a tendency for a greater proportion of bodies in these graves to be protected by a coffin, cist, or stone-​lining. These two traditions could appear contemporaneously within a single settlement but were usually spatially distinct. This distinction can be seen within an individual cemetery in the case of Poundbury, Dorset, where the bulk of the cemetery population was aligned west–​east and had few grave goods, with a significant number being placed in wooden, stone, or lead coffins (Farwell and Molleson 1993). Within this group were a number placed in stone-​lined mausolea with internal plaster wall-​painting, some of which may have had Christian symbolism (Sparey-​Green 1993). However, at the periphery of the cemetery were smaller plots where the graves were aligned mainly north–​south and contained grave goods. At a different spatial resolution, this distinction between the two rites is apparent at Ilchester, Somerset, where the main cemetery to the north-​east of the town was dominated by west–​east-​aligned graves with few grave goods (P. Leach 1994: 91–​103). To the south-​east of the town, situated in the back lots of roadside buildings, burials characteristic of the other tradition were situated (P. Leach 1982: 82–​88).

Christianity in Roman Britain    673 Though the two rites were clearly broadly contemporary through the fourth century, it is not clear precisely when the west–​east aligned inhumation tradition developed. The lack of grave goods means that they are often inherently hard to date. However, at Poundbury the cemetery appears to have been founded in the early to mid-​fourth century, while at Butt Road, Colchester, there is clearly a major cemetery reorganization of the burial ground, replacing a more heterogenous tradition that featured a wide range of grave goods and a north–​south alignment, at some point in the late third or early fourth century ad (Crummy et al. 1993: 4–​163; Millett 1995). These changes in burial rite appear generally to have occurred around the turn of the fourth century or perhaps a little later, but it is not easy to show for certain whether these developments are pre-​Constantinian or not. It would be tempting to relate the advent of this burial rite with the spread of Christianity, but the situation is doubtlessly more complicated than this. A major obstacle for such a simple interpretation is the complete lack of any textual evidence for any extensive church interest in defining the nature of the Christian burial rite in this period, with no recorded prescriptions about orientation, spatial organization, or treatment of the body beyond a broad stricture against cremation. However, in the fourth century it is likely that, with the increasing Christianization of the imperial civil service, the church stepped into the gap left by the secular elite’s withdrawal from involvement in this sphere and became increasingly involved in civil administration. This would have allowed the church increasingly to control a range of activities, including burial in major cemeteries. This need not mean that all those buried within the formally arranged cemeteries were Christians; many may simply have been conforming to a burial rite advocated by a socially and politically powerful element within Roman society. Those who were not able to come to an accommodation with this may instead have opted to bury their dead in more peripheral locations. Indeed, the increase of unusual rites, such as post-​mortem decapitation, may even have been an attempt to develop more overtly ‘pagan’ rites in the face of an increasing Christian orthodoxy (for a more developed version of this argument, see Petts 2003: 146–​149). Attempts to distinguish diagnostic evidence for Christian burial are hampered by the lack of overtly Christian epigraphy from fourth-​century Britain. This reflects the wider decline in the epigraphic habit in the third and fourth centuries throughout the western Empire (Handley 2001). A number of fourth-​century funerary inscriptions have been identified as potentially Christian on the basis of parallels between particular epigraphic formulae used in similar Christian contexts elsewhere in the empire (for example, vixit plus minus found on inscriptions from Carlisle and Brougham; titulum posuit found on inscriptions from Hadrian’s Wall, Old Penrith, Templeborough, and York) (Handley 2001: 181–​183; RIB 620, 689, 786-​7, 934, 955, 1667). The extent to which these phrases are peculiarly Christian are open to debate, and Knight (2010) has argued for the need to look for broad ‘Christian symptoms’ rather than direct Christian identity in the epigraphic record. Across the empire there is a slight increase in the use of epigraphy towards the end of the fourth century and into the early fifth century (Handley

674

674   David Petts 2001: 181–​183). There is a small group of late gravestones from the western end of Hadrian’s Wall that appear to date from this period; however, it is still unclear whether they belong in a fourth-​or fifth-​century context, and they have no overt Christian symbolism (Dark and Dark 1996; Todd 1999). In the sub-​Roman/​early medieval period there is a clearer revival of funerary epigraphy in the western fringes of Roman Britain and, to a lesser extent, in lowland Scotland (Thomas 1993; Edwards 2001). While these later stones are clearly being erected in a Christian context, they too carry remarkably limited evidence for overt religious belief. There is still some debate over the precise relationship between these later insular inscriptions and the latest Roman examples (Handley 2001; Knight 2010). These insular stones differ from contemporary continental stones in their notable lack of Christian symbols or imagery (for example, chi-​rho symbols, peacock, or doves).

Conclusions The intractable nature of the evidence for Roman Christianity in Britain means that there is very little consensus on its extent during the fourth century. The success or otherwise of the Romano-​British church has important implications for understanding the growth of Christianity in western and northern Britain in the fifth and sixth centuries. Some scholarship has regarded the fourth century as a period of consolidation, with a firmly established church within the Roman province forming a springboard for its rapid extension beyond the frontiers of the empire in the fifth century (e.g. Thomas 1981; Petts 2003). An alternative perspective sees the church in Roman Britain as remaining institutionally very weak and reliant on direct imperial support. As a consequence, the abdication of the political control of Britannia by the Roman state in the early fifth century resulted in the final failure of an attenuated church (e.g. Frend 1955, 1968, 1979; Watts 1991). This view then requires that early medieval Britain undergoes a period of reconversion carried out by missionaries from mainland Europe. While the limited nature of the evidence is capable of multiple readings, it is clear from the early documentary evidence that, while the Gallic church clearly had an involvement with the British church, this appears to be in the context of a relationship between two established bodies, and there is no indication of any missionary activity. It is also notable that the earliest significant historic sources written within early medieval Britain situated the island’s conversion during Roman rule rather than the post-​Roman period. Nonetheless, it is clear that, despite the likelihood of continuity, there was a major realignment of the geographic focus of the British church, with a probable decline or collapse of the church in the east of the province in the face of Anglo-​Saxon takeover. The best evidence for the church can be found in western Britain and southern Scotland, regions that had previously been peripheral to the focus of Roman control. However, such shifts in emphasis must be understood within the context of wider changes in the geography of power in sub-​Roman Britain. It is important, though, to emphasize that later fourth-​century and

Christianity in Roman Britain    675 fifth-​century sub-​Roman paganism has been under-​researched, and the low material visibility of the sub-​Roman church is paralleled in the lack of a diagnostic archaeology. As well as debates about the extent and nature of continuity, another key area of debate about Romano-​British Christianity is centred on its relationship with contemporary paganism. This relationship can be recognized in a number of separate cultural spheres. It is clear from the juxtaposition of Christian and Gnostic imagery on some mosaics that there may have been a positive interaction between the church and the more literate and theologically sophisticated aspects of classical paganism (Perring 2003). However, it is not always easy to distinguish the deliberate association of compatible aspects of Gnostic and Christian imagery from a more general expression of classical learning (paideia) that even Christian elites were expected to maintain in this period (Petts 2003: 116–​118). Pagan influence can also be seen in other aspects of Christian practice. This can be seen clearly in the silver repoussée votive leaves that formed an element of the Water Newton hoard. These have strong parallels with votive leaves with overt pagan imagery found elsewhere in Britain (Henig 1993; Painter 1999; Crerar 2006). The practice of depositing such votive offerings found a similar material expression in pagan and Christian practice in Roman Britain. An argument has also been made for wider similarities between pagan and Christian depositional practices in the fourth century (Petts 2002; cf. Fulford 2001). Relationships between Christian and pagan may, of course, also have been tense. Evidence for the destruction of pagan shrines, temples, and statuary in the fourth century has often been related to Christian iconoclasm (for example, the Wallbrook Mithraeum: Green 1976: 46; Merrifield 1977: 375). More recently, though, it has been recognized that the evidence for seemingly arbitrary destruction may actually have been more carefully structured and that the fragmentation and preservation of elements of statues may fit into wider attitudes to the preservation and destruction of holy objects (Croxford 2003). While some destruction of pagan sites and objects may have occurred in the context of Christian-​on-​pagan violence, the evidence for complex artefact biographies for much of the cult material is a reminder that there may be other motivations for damaging and dismembering such items. A final dimension to the wider analysis of Christianity in Roman Britain is the need for a better understanding of the extent of regional variation. This needs to encompass a better appreciation of variation in the evidence within Britain, but also a clearer appreciation of how Romano-​British Christianity compares with the evidence for the church elsewhere in the empire in the fourth century. It is now clear that simple cumulative distribution maps, showing the presence of diagnostic Christian material, are not a satisfactory way of mapping the extent of Christianity (Petts 2003: 26–​27). A range of social factors can influence such distributions and maps can mask spatial variability. For example, it is apparent that the majority of hoards with Christian items are found in the east Midlands and East Anglia, whereas the evidence for high-​status Christianity, as found on mosaics, is largely found in the south-​west. However, in both cases there are complex social processes at play. The south-​west of England also has the largest number of late Roman temples, and we are most likely simply seeing an area where religious belief was expressed through the construction of buildings, whereas in East Anglia and the

676

676   David Petts Midlands there is clearly a long-​term tradition of depositing substantial votive hoards (Petts 2003). However, this does not necessarily mean that Christians in the south-​west of England were not using silver plate for religious purposes; it may simply never have entered the archaeological record, instead perhaps being recycled. Such variation in religious practice at an intra-​provincial level should come as no surprise; nor should the fact that such variation can be recognized between provinces (Petts 2014). For example, the distribution of lead tanks with Christian imagery is almost entirely restricted to Britain, as is the use of peacock imagery on belt-​buckles. However, the use of Christian imagery on gravestones is largely a continental phenomenon, and there is a lack of stone sarcophagi bearing distinct Christian symbolism in Britain. There are also similarities between provinces—​for example, both Britain and northern Gaul are largely lacking good archaeological evidence for substantial fourth-​century church structures, with little to compare with the more spectacular evidence from further south or east (Petts 2014). In conclusion, the evidence for Christianity from Roman Britain is not extensive or easy to interpret, although it is useful to remember that it is still the most widely represented individual cult in fourth-​century Britain. While it may never have been the dominant religion in terms of numbers, it was clearly adopted by wealthy individuals, many of whom were likely to have been in positions of political and social power. There is still debate over the extent to which this Roman church was the foundation for the rapid blossoming of Christianity in the insular world in the fifth and sixth centuries. However, simply focusing on the ‘success’ or otherwise of the Roman church should not prevent us from developing a wider understanding of the institution both within and beyond Britannia.

Abbreviation RIB

R. G. Collingwood and R. P. Wright, The Roman Inscriptions of Britain. Volume I: Inscriptions on Stone. Oxford: Clarendon Press, 1965–​to date.

References Primary Sources De Laude Sanctorum, ed. J.-​P. Migne 1845, Patrologia Latina 20 ( Ridgewood, NJ: Gregg, 1965). Gildas, De Excidio (The Ruins of Britain), ed. and trans. M. Winterbottom (Chichester: Phillimore, 1978).

Secondary Sources Ap Simon, A. (1964–​5). ‘The Roman Temple on Brean Down, Somerset’, Proceedings of the Speleological Society, 10: 195–​258. Baur, P. V. C., and Hopkins, C. (1934). The Christian Church at Dura Europos. New Haven, CT: Yale University Press.

Christianity in Roman Britain    677 Bidwell, P., and Speak, S. (1994). Excavations at South Shields Roman Fort, Volume 1. Newcastle: Society of Antiquaries of Newcastle upon Tyne. Birley, R. (2009). Vindolanda: A Roman Frontier Fort on Hadrian’s Wall. Stroud: Amberley. Bowes, K. (2008). Private Worship, Public Values and Religious Change in Late Antiquity. Cambridge: Cambridge University Press. Brown, P. D. C. (1971). ‘The Church at Richborough, Kent’, Britannia, 2: 225–​231. Bryant, R. (1980). ‘Excavations at St Mary De Lode, Gloucester 1978–​1979’, Glevenis, 14: 4–​12. Clarke, L. G. C. (1931). ‘Roman pewter Bowl from the Isle of Ely’, Proceedings of the Cambridge Antiquarian Society, 31: 66–​75. Collingwood Bruce, J. (1880). A Descriptive Catalogue of Antiquities Chiefly British at Alnwick Castle. Newcastle: Andrew Reid. Colyer, C., and Jones, M. (1979). ‘Excavations at Lincoln, Second Interim Report. Excavations in the Lower Town’, Antiquaries Journal, 59: 50–​91. Cosh, S. (2004). ‘A possible date for the Silchester “Church”’, Britannia, 35: 229–​233. Cosh, S., and Neal, D. (2006). Roman Mosaics of Britain. Volume II. South-​West Britain. London: Society of Antiquaries of London. Crerar, B. (2006). ‘Votive Leaves from Roman Britain’, in M. Henig (ed.), Roman Art, Religion and Society:  New Studies from the Roman Art Seminar, Oxford 2005. BAR International Series, 1577. Oxford: Archaeopress, 71. Crerar, B. (2012). ‘Contextualising Romano-​ British Lead Tanks:  A  Study in Design, Destruction and Deposition’, Britannia, 43: 135–​166. Crow, J. (1995). The English Heritage Book of Housesteads. London: Batsford. Croxford, B. (2003). ‘Iconoclasm in Roman Britain?’, Britannia, 34: 81–​95. Crummy, P., Crummy, N., and Crossan, C. (1993). Colchester Archaeological Report 9: Excavations of Roman and Later Cemeteries, Churches and Monastic Sites in Colchester 1971–​88. Colchester: Colchester Archaeological Society. Cunliffe, B. W. (1968) (ed.). Fifth Report on the Excavations of the Roman Fort at Richborough, Kent. London: Society of Antiquaries of London. Curle, A. O. (1923). The Treasure of Traprain Law. Glasgow: Maclehose, Jackson. Dark, K. R., and Dark, P. (1996). ‘New Archaeological and Palynological Evidence for a Sub-​ Roman Reoccupation of Hadrian’s Wall’, Archaeologia Aeliana, 24: 57–​72. Edwards, N. (2001). ‘Early Medieval Inscribed Stones and Stone Sculpture in Wales: Context and Function’, Medieval Archaeology, 45: 15–​39. Engleheart, G. H. (1922). ‘On Some Roman Buildings and Other Antiquities in a District of North-​ West Hants’, Proceedings of the Hampshire Field Club and Archaeological Society, 9: 214–​218. Farwell, D. E., and Molleson, T. I. (1993). Poundbury Volume 2:  The Cemeteries. Dorchester: Dorchester Natural History and Archaeological Society. Ford, S. D. (1994). ‘The Silchester Church: A Dimensional Analysis and a New Reconstruction’, Britannia, 25: 119–​126. Frend, W. H. C. (1955). ‘Religion in Roman Britain in the 4th Century’, Journal of the British Archaeological Society. 3rd ser. 18: 1–​18. Frend, W. H.  C. (1968). ‘The Christianisation of Roman Britain’, in M. W. Barley and R. P.  C. Hanson (eds), Christianity in Britain 300–​700:  Papers Presented to the Conference on Christianity in Roman and Sub-​Roman Britain, 17–​20 April 1967. Leicester:  Leicester University Press, 37–​50. Frend, W. H. C. (1979). ‘Ecclesia Britannica: Prelude or Dead End?’, Journal of Ecclesiastical History, 30: 129–​144.

678

678   David Petts Frere, S. S. (1976). ‘The Silchester Church:  The Excavations by Sir Ian Richmond in 1961’, Archaeologia, 105: 277–​302. Fulford, M. J. (2001). ‘Links with the Past: Pervasive “Ritual” Behaviour in Roman Britain’, Britannia, 32: 199–​218. Gerrard J. (2011). ‘Cathedral or Granary? The Roman Coins from Colchester House (PEP89)’, Transactions of the London and Middlesex Archaeological Society, 62: 81–​88. Goodburn, R. (2000). Chedworth Roman Villa. London: National Trust. Graham, T. (2002). ‘A Rho-​Cross Engraved on a Jet Finger-​Ring from Bagshot, Surrey’, Oxford Journal of Archaeology, 21: 211–​216. Green, M. (1976). A Corpus of Religions Material from the Civilian Areas of Roman Britain. BAR British Series, 24. Oxford: Archaeopress. Handley, M. (2001). ‘The Origins of Christian Commemoration in Late Antique Britain’, Early Medieval Europe, 10/​2: 177–​199. Hawkes, S. C. (1973). ‘A Late Roman Buckle from Tripontium’, Transactions of the Birmingham and Warwickshire Archaeological Society, 85: 145–​159. Heighway, C. (2010). ‘Christian Origins at Gloucester: A Topographical Inquiry’, in M. Henig and N. Ramsay (eds), Intersections: The Archaeology and History of Christianity in England, 400–​1200. BAR British Series, 505. Oxford: Archaeopress, 39–​48. Henig, M. (1993). ‘Votive Objects: Weapons, Miniatures, Tokens and Fired Clay Accessories’, in A. Woodward and P. Leach, The Uley Shrines: Excavation of a Ritual Complex on West Hill, Uley, Gloucestershire, 1977–​9. London: English Heritage and British Museum Publications, 130–​147. Henig, M. (2006). ‘Neither Baths nor Baptisteries’, Oxford Journal of Archaeology, 25: 105–​107. Henig, M., and Brown, C. (2003). ‘A Romano-​British Buckle Plate from East Challow, near Wantage’, Oxoniensis, 67: 363–​365. Johns, C. (1981). ‘The Risley Park Lanx: A Lost Antiquity from Roman Britain’, Antiquaries Journal, 61: 53–​72. Johns, C. (2001). ‘Silver Pendant with Monogram Cross’, in P. Leach, Fosse Lane, Shepton Mallet 1990. Britannia Monograph Series 18. London: Britannia, 257–​260. Johns, C. (2010). The Hoxne Late Roman Treasure:  Gold Jewellery and Silver Plate. London: British Museum Publications. Johns, C., and Painter, K. (1991). ‘The Risley Park Lanx Rediscovered’, Minerva, 2/​62: 6–​13. Johns, C., and Potter, T. (1983). The Thetford Treasure:  Roman Jewellery and Silver. London: British Museum Publications. Johns, C. M., and Potter, T. W. (1985). ‘The Canterbury Late Roman Treasure’, Antiquaries Journal, 65: 312–​352. Joyce, G. (1881). ‘Third Account of Excavations at Silchester’, Archaeologia, 46: 344–​365. King, A. (1983). ‘The Roman Church at Silchester Reconsidered’, Oxford Journal of Archaeology, 2: 225–​237. Knight, J. (2001). ‘Britain’s Other Martyrs: Julius, Aaron and Alban in Caerleon’, in M. Henig (ed.), Alban and Saint Albans:  Roman and Medieval Art, Architecture and Archaeology. Leeds: British Archaeological Association, 30–​35. Knight, J. K. (2010). ‘An Inscription from Bavai and the Fifth Century Christian Epigraphy of Britain’, Britannia, 41: 283–​292. Krämer, W. (1958). Neue Ausgrabungen in Deutschland. Berlin: Verlag Gebr. Mann. Krautheimer, R. (1986). Early Christian and Byzantine Architecture. 4th edn. New Haven, CT: Yale University Press.

Christianity in Roman Britain    679 Leach, P. (1982). Ilchester: Excavations 1974–​5. Bristol: Western Archaeological Trust. Leach, P. (1994). Ilchester Volume 2: Archaeology, Excavations and Fieldwork to 1984. Sheffield Excavation Reports, 2. Sheffield: Collis. Leech, R. (1986). ‘The Excavation of a Romano-​Celtic Temple and a Later Cemetery on Lamyatt Beacon, Somerset’, Britannia, 17: 259–​328. Mawer, F. (1995). Evidence for Christianity in Roman Britain: The Small Finds. Oxford: BAR British Series, 243. Oxford: Tempus Reparatum. Meates, G. W. (1979). The Roman Villa at Lullingstone. Volume 1: The Site. Maidstone: Kent Archaeological Society. Meates, G. W. (1987). The Roman Villa at Lullingstone. Volume 2: The Wall Paintings and Finds. Maidstone: Kent Archaeological Society. Merrifield, R. (1977). ‘Art and Religion in Roman London: An Inquest on the Sculptures of Londinium’, in J. Munby and M. Henig (eds), Roman Life and Art in Britain II. BAR British Series, 41. Oxford: Archaeopress, 375–​406. Middleton, J. H. (1881–​3). [Proceedings:  2 February  1881]. Proceedings of the Society of Antiquaries of London. 2nd ser. 9: 66–​70. Millett, M. (1995). ‘An Early Christian Community at Colchester?’, Archaeological Journal, 152: 451–​454. Moorhead, S. T. (2000). ‘An Inspiration for the Hinton St Mary Head of Christ’, British Museum Magazine, 36 (Spring): 22. Munier, C. (1963). ‘Concilia Galliae A.  314–​A. 506’, Corpus Christianorum CXLVIII Series Latina. Turnholti, Belgium: Brepols. Ó Ríordáin, S. P. (1947). ‘Roman Material in Ireland’, Proceedings of the Royal Irish Academy, 47: 77–​150. Painter, K. S. (1967). ‘The Roman Site at Hinton St Mary, Dorset’, British Museum Quarterly, 32/​ 1–​2: 15–​31. Painter, K. S. (1972). ‘Villas and Christianity in Roman Britain’, in Actas del VIII Congresso Internacional de Arqueologia Cristiana 1969. Barcelona: Consejo Superior de Investigaciones Cientificas Egipciaces, 149–​166. Painter, K. S. (1977). The Water Newton Early Christian Silver. London:  British Museum Publications. Painter, K. S. (1999). ‘The Water Newton Silver: Votive or Liturgical?’ Journal of the British Archaeological Association, 152: 1–​23. Painter, K. (2010). ‘A Roman Silver Jug with Biblical Scenes from the Treasure Found at Traprain Law’, in M. Henig and N. Ramsay (eds), Intersections:  The Archaeology and History of Christianity in England, 400–​ 1200. BAR Reports British Series, 505. Oxford: Archaeopress, 1–​24. Pearce, S. (2008). ‘The Hinton St Mary Mosaic Pavement:  Christ or Emperor?’, Britannia, 29: 193–​218. Penney, S., and Shotter, D. C. A. (1996). ‘An Inscribed Salt-​Pan from Shavington, Cheshire’, Britannia, 27: 360–​365. Perring, D. (2003). ‘Gnosticism in Fourth-​ Century Britain:  The Frampton Mosaics Reconsidered’, Britannia, 34: 97–​128. Petts, D. (2002). ‘Votive Deposition in Late Roman Britain: Pagan or Christian?’, in M. Carver (ed.), The Cross Goes North: Processes of Conversion in Northern Europe. Woodbridge: Boydell Press, 109–​118. Petts, D. (2003). Christianity in Roman Britain. Stroud: Tempus.

680

680   David Petts Petts, D. (2004). ‘Early Medieval or Late Antique? Burial in Western Britain ad 410 to 600’, in R. Collins and J. Gerrard (eds), Debating Late Antiquity ad 300–​700: Proceedings of the York Conference. BAR British Series, 365. Oxford: Archaeopress, 77–​87. Petts, D. (2014). ‘Christianity and Cross-​Channel Connectivity in Late and Sub-​Roman Britain’, in R. Collins, F. Haarer, K. Matthews, S. Moorhead, and D. Petts (eds), Late and Post-​Roman Britain. London: Society for the Promotion of Roman Studies, 73–​88. Philpott, R. (1991). Burial Practices in Roman Britain. BAR British Series, 219. Oxford: Tempus Reperatum. Quensel-​von Kalbern, L. (1999). ‘The British Church and the Emergence of the Anglo-​Saxon Kingdoms’, Anglo-​Saxon Studies in Archaeology and History, 10: 89–​98. Rivet, A. L. F., and Smith, C. (1979). Place Names of Roman Britain. London: Batsford. Salway, P. (1981). Roman Britain. Oxford: Clarendon Press. Sanderson, I. (1993). ‘Exciting Find Helps Throw Light on “Dark Ages”’, Clwyd Archaeology News, 92–​3 (Winter): 2. Sankey, D. (1998). ‘Cathedrals, Granaries and Urban Vitality in Late Roman London’, in B. Watson (ed.), Roman London: Recent Archaeological Work. JRA Supplementary Series 24. Portsmouth, RI: Journal of Roman Archaeology, 78–​82. Sharpe, R. (2001). ‘The Late Antique “passio” of St Alban’, in M. Henig (ed.), Alban and Saint Alban’s: Rome and Medieval Art, Architecture and Archaeology. Leeds: British Archaeology Association, 13–​34. Snyder, G. F. (2003). Ante Pacem: Archaeological Evidence of Church Life before Constantine. Macon, GA: Mercer University Press. Sparey-​Green, C. (1993). ‘The Mausolea Painted Plaster’, in D. E. Farwell and T. I. Molleson (eds), Poundbury Volume 2:  The Cemeteries. Dorchester:  Dorset Natural History and Archaeological Society, 135–​150. Steane, K. (2006). The Archaeology of the Upper City and Adjacent Suburb. Oxford: Oxbow Books. Thomas, C. (1981). Christianity in Roman Britain to ad 500. London: Batsford. Thomas, C. (1993). ‘The Early Christian Inscriptions of Southern Scotland’, Scottish Archaeological Journal, 17: 1–​10. Todd, M. (1999). ‘The Latest Inscriptions of Roman Britain’, Durham Archaeological Journal, 14–​15: 53–​59. Todd, M. (2005). ‘Baths or Baptisteries? Holcombe, Lufton and their Analogues’, Oxford Journal of Archaeology, 24: 307–​311. Toynbee, J. (1964). ‘The Christian Roman Mosaic, Hinton St Mary, Dorset’, Proceedings of the Dorset Archaeological Society, 85: 116–​121. Turner, R. (1999). Excavations of an Iron Age Settlement and a Roman Religious Complex at Ivy Chimneys, Witham, 1978–​83. East Anglian Archaeology, 88. Chelmsford: East Anglian Archaeology. Watts, D. (1991). Christians and Pagans in Roman Britain. London: Routledge. West, S. E., and Plouviez, J. (1976). ‘The Roman Site at Icklingham’, East Anglian Archaeology, 3: 63–​125. Woodward, A., and Leach, P. (1993). The Uley Shrines:  Excavation of a Ritual Complex on West Hill, Uley, Gloucestershire, 1977–​9. London:  English Heritage and British Museum Publications.

Chapter 33

Memories of t h e Past in Roman Bri ta i n Zena Kamash

Introduction In recent years there has been a boom in memory studies, with a particular focus on collective memory. In what follows I will first present a brief overview of the history of memory studies and then discuss the role memory has played in archaeological thought in the 1990s and 2000s. I will then move onto look at the place of memory in the Roman world, before considering the role of memory in Roman Britain. Several threads will run through this discussion, including the temporal direction of memory, the manipulation of memory, and competing memories. These themes revolve around the idea that people tend to look back in order to address concerns for their present and future and, in so doing, are apt to manipulate memories to suit these present concerns. This can lead to competing memories, whereby there is not a single, uncontested version of the past. This has numerous ramifications for our understandings of identity in Roman Britain. Working across all these themes will be an exploration of how memories were enacted in Roman Britain and how these memory practices may be similar to or different from the wider Roman world.

The History of Memory The traditional focus of the western episteme, including Roman writers on memory, was on the individual, private nature of memory. The key Roman texts on memory—​mostly from works on rhetoric—​can be found in the writings of Cicero (de Oratore 2.74.299–​ 300; 2.86.351–​4), Quintilian (Institutio Oratoria 11.2.17–​26), the author of the Rhetorica ad Herrenium (3.16–​24), and Augustine (e.g. Confessions 10.8–​12). The focus of these

682

682   Zena Kamash authors is on how we remember and the methods by which we can enhance our ‘natural’ memory through training and mnemonic techniques. The most famous of the examples used by these authors is that of Simonides of Ceos, who is said to be the inventor of mnemonics. According to the story, Simonides was able to recall who was sitting where at dinner at the House of Scopas at Crannon (Thessaly), and so able to identify the remains of those killed after the roof fell in (for a detailed discussion of this story, see Farrell 1997). In addition, Cicero (de Or. 2.74.299–​300) also foreshadows Borges’s ‘Funes the Memorious’ (2000: 87–​95) and modern neuroscience in his discussion about how it is kinder to be taught how to forget than how to remember (see, e.g., Parker et al. 2006). The main thread running though all these accounts is that memory was seen as a storehouse of faithful representations of past sensory experiences that could be drawn upon as and when required. This was the accepted and unchallenged view of memory throughout western thought until the twentieth century. In the twentieth century, there was a major rethinking of the nature of memory. In the 1950s Maurice Halbwachs (1992), a student of Emile Durkheim, proposed the radical theory of collective memory (Gedächtnisgeschichte). In this new account, memory was not restricted to individual minds, but could be shared and distributed across groups and communities. Collective memory could be seen, then, as ‘what remains of the past in the lived reality of groups, or what these groups make of the past’ (Le Goff 1992: 95; see also Nora 1978). This work has been highly influential in memory studies, in particular in the work of scholars such as Pierre Nora (1984–​92), Aleida Assman (1999) and Jan Assmann (1999). Much of this work shifted the focus away from ‘memory as storehouse’ and instead moved towards the idea that memory is malleable and open to change, and so, by implication, not necessarily a faithful representation of one particular true past. This new malleable, collective memory opened up the possibility of numerous competing narratives of past events. (For a full history of collective memory, see Olick et al. 2011. For critiques of the ‘memory industry’, see Todorov 1995; Confino 1997; Klein 2000.)

The Material Turn: Archaeology and Memory Although the important role the physical world has to play in the creation and summoning of memories was implicit in, for example, Pierre Nora’s concept of ‘lieux de mémoire’, its full potential was largely sidelined in favour of written and oral histories (see, e.g., Le Goff 1992: 52). Towards the end of the twentieth century, however, collective memory studies saw a turn towards materials, considering the ways in which material culture is implicated in both remembering and forgetting. The arrival of a material element to collective memory brought with it a space in which archaeologists could make a contribution (e.g. Radley 1990; Rowlands 1993; Tatum 1995; Williams and Bradley 1998; Alcock

Memories of the Past in Roman Britain    683 2002; Van Dyke and Alcock 2003; Jones 2007; Borić 2010; Lillios and Tsamis 2010). In particular, archaeologists have noted the ability of objects to traverse time and, through their physical and sensual qualities, to make the past present (Jones 2007: 61). In addition, this archaeological work—​particularly that of Williams and Bradley (1998)—​has brought with it the concept of ‘the past in the past’ that is now key in archaeological thought. This concept introduced the idea that past peoples were also conscious of the past and made reference to it in numerous, material ways that we as archaeologists might detect and interpret. This is a theme to which I will return later in this chapter. Implicit in this idea is the relationship between memory and time. Here, I follow Schwartz (1982: 395) in thinking that references to the past, material or otherwise, are often made because of concerns in the present. Furthermore, I would argue that memories are used not just to create present identities, but also, potentially, to project visions and hopes for the future. This facet of memory is made clear by Holtorf and Williams (2006), who introduce the idea of ‘retrospective’ and ‘prospective’ memories in relation to accumulative and created landscapes, respectively. An example of the latter would be a war memorial, where the memories being created are intended for the future. In the first type, retrospective memories, people are looking back at, or to, older elements of the landscape; in cases where the actual ‘meaning’ of that place or ‘content’ of that memory is lost, we may find the creation of landscapes of the imagination and concomitant legendary topographies: This kind of ‘looking back’ is not necessarily about accurately recalling past events as truthfully as possible: it is rather about making meaningful statements about the past in the given cultural context of a present as well as evoking aspirations for the future. (Holtorf and Williams 2006: 238)

While we can point to distinct examples of these kinds of memorial processes in landscapes, they are often found acting together in a single landscape, as will be demonstrated later in this chapter in my discussion of examples of prospective and retrospective memories from Roman Britain. They also point to the fact that the deliberate selection of points referenced in a landscape can be highly manipulative, hiding some aspects of the past and encouraging others to be forgotten. This links with some caveats made by Hella Eckardt (2004), who reminds us that, when we, as archaeologists, look at landscapes, we should not assume that all past objects and monuments held equal significance (an easy trap into which to fall when considering the past in the past). She encourages us to search not ‘simply for re-​use, but for more complex interactions with the past, which include nostalgia, exclusion, destruction and forgetting’ (Eckardt 2004: 46). On the topic of destruction and forgetting, Susanne Küchler’s studies (1987, 1988) of malangan sculpture in Papua New Guinea has demonstrated that the creation and deliberate destruction of sculpture were powerful methods not only of remembering, but also of forgetting. Küchler neatly reminds us that remembering and forgetting can often not be separated. Also central to Küchler’s work is the concept that the action of removing an object from memory can be as important as the object itself (see also Connerton

684

684   Zena Kamash 1998). This highlights the fact that memory cannot be of objects alone, but must be thought about in conjunction with human experience and action—​that is, that memory is distributed between people and things (Renfrew et al. 2004: 2; Jones 2007: 26). Some of the most powerful examples of these actions of remembering and forgetting through objects are found in Holocaust memorials. The Stolpersteine, for example, are small, cobble-​sized bronze plaques that are set into pavements across many cities in northern Europe. Each gives the name of a victim of the Holocaust. One of the most striking commentaries on these Stolpersteine—​and of direct relevance to this discussion of objects, action, and forgetting—​comes from Joseph Pearson (2010): It is not what is written which intrigues me, because the inscription is insufficient to conjure a person. It is the emptiness, void, lack of information, the maw of the forgotten, which gives the monuments their power and lifts them from the banality of a statistic. (emphasis added)

Memory in the Roman World Most recently Karl Galinsky’s Memoria Romana project (2009-​2013) has explicitly sought to explore the role concepts such as Gedächtnisgeschichte might play in our understandings of memory in the Roman world (Galinsky 2009). The project funded numerous Ph.D. scholarships and postdoctoral research projects, whose fruits are beginning to make themselves felt, including my own work on memory and religion in Roman Britain (see Fabrizi 2012; Li Causi 2012; Farrell and Nelis 2013; Raymond 2013; Rebeggiani 2013; Seider 2013; Kamash 2016; Barchiesi, forthcoming; Kamash, forthcoming). The project also organized major conferences in Rome (Memory in Rome and Rome in Memory, which included a contribution from Daniel Libeskind (Galinsky 2014)) and at the Getty Villa (Cultural Memories in the Roman Empire (Galinsky 2016)). This project has been able to use a wealth of previous work on Roman memory as its springboard. In many ways the important role of memory in Roman society has long been recognized in studies of the Roman world and pre-​dates the attention paid to memory in other areas of archaeology. This is due, partly, to the writings of Roman authors on memory and, partly, to the much-​discussed idea of damnatio memoriae—​the act of deliberately removing from public view all images of, and inscribed references to, disgraced emperors; for example, Domitian, or Caracalla’s unfortunate brother, Geta (on this in Britain, see Barrett 1993). Of course, it is also now widely recognized that such actions serve less to make people forget than to act as powerful reminders of that person and his or her actions (Flower 1998, 2006). The act of damnatio memoriae does, however, bring to the fore two main features of memory in the Roman world: first, that memory was central to the lives—​and perhaps most importantly the afterlives—​of Roman people, and, secondly, that the Romans were well practised in the art of manipulating memories.

Memories of the Past in Roman Britain    685 A few, highly selective, examples from in and around the city of Rome will serve to demonstrate these features (a useful and authoritative set of bibliographies on the subject can be found on the Memoria Romana website:  Galinsky 2014). The Via Appia, which connects Bovillae and Rome, is renowned for the elaborate tombs that lined its course. Here the memory of those who had died was kept alive in tombs that bore their names, faces, and sometimes inscriptions detailing their careers and genealogies. These tombs and their prominent position highlight the centrality of vision and visibility in the maintenance of memory in Rome (see Koortbojian 1996). Of all monuments, the Arch of Constantine in Rome perhaps best illustrates the role that monuments could play in memory in the Roman world. This monument was dedicated to the emperor Constantine by the senate for his decennial celebrations. The arch is well known for its use of spolia—​reused architectural components—​including examples of earlier imperial portraiture being recut into the face of Constantine. In his analysis of the monument, and particularly the use of spolia, Gutteridge (2010) suggests that this is a ‘trans-​temporal’ monument that defies the law of time. The memories of vanished buildings and dead emperors cannot be completely erased in this monument but, rather, are recombined into a new set of remembrances that look forwards and backwards simultaneously. Perhaps the best example of Roman manipulation of memory can be seen in Augustus’ building programme in Rome (see Gowing 2005: 132–​159). Augustus deliberately chose to maintain what could be considered as antiques in the forum, such as the umbilicus urbis (the central pit dug by Romulus when he laid out the sacred boundary of the city) and the lapis niger (the spot in front of the senate house where Romulus disappeared—​either murdered or taken to the heavens), reminding the people of his commitment to protecting the old traditions of Rome. He also constructed new buildings, such as the Temple to Divus Julius, linking himself both to the recent past and to a divine lineage. In these ways, Augustus made use of spaces and buildings in Rome to create both a museum of the past and a dynastic monument to his family, and so to manipulate strongly the memories of Rome’s inhabitants.

Remembering in Roman Britain Looking at memory in Roman Britain presents somewhat different challenges and opportunities from studies of the remainder of the empire. As well as the distance of Britain from Rome, major differences lie in the nature of the archaeological evidence and in the history of Romano-​British studies. Although Britain does have some examples of still extant monumental architecture (for example, the amphitheatres at Silchester, Cirencester, and Chester), these are in no way on the same scale as the Mediterranean. This has clear ramifications: not for us the detailed analyses of sculpture or streets lined with funerary monuments. This does not mean, however, that we cannot engage with the debates over memory in the Roman world. On the contrary, I am of the opinion

686

686   Zena Kamash that, because of these differences, scholars of Britain under the Roman Empire have, for a long time, allied themselves closely with prehistoric archaeology and so have taken a more theoretically driven approach, putting us ahead of the memory game in comparison to other parts of the Roman world. Foremost, here, is the work of Howard Williams (1998) on the past in the past in relation to Roman Britain. In addition, we also have the major advantage of access to an ever-​increasing amount of high-​quality excavation data (Fulford and Holbrook 2011), from which to make detailed analyses of the processes of memory in Roman Britain. In what follows I hope to capitalize on these substantial opportunities. I will look at three realms of memory in Roman Britain: the burial of memories; the reorganization of landscapes and memories; and the building of religious memory. First, I will explore how burying people and objects, and in so doing making them invisible, can be part of the memory process. The focus here will be on embodied actions and the use of legendary topographies in the landscape. I will then move on to examine how different memory communities responded to major periods of landscape reorganization with the aim of demonstrating how these analyses can contribute to wider discussions about the creation and maintenance of identity in Roman Britain. Finally, I will explore the Romano-​ Celtic temple phenomenon. I will argue that their construction—​and the memories generated and evoked by them—​was linked to the kinds of memory-​making that are particularly prevalent in times of social instability and that this phenomenon should be seen as part of a broader set of processes that began in the late Iron Age. My aim here is not to be exhaustive, but rather to demonstrate the potential of taking a memory approach in understanding Roman Britain. In particular, I wish to demonstrate that focusing on memory can contribute to, and enhance, wider topical discussions about continuity and change and identity. In so doing, I will be drawing on many of the insights about memory outlined above, with an emphasis on the temporal directions of memory practices in Roman Britain and what these may suggest about the people enacting these practices. In particular, I will make reference to the active interplay between people and things in processes of remembering and forgetting, and to the retrospective and prospective directions of memory. As noted above, one of the biggest differences between Roman Britain and the Mediterranean is the scale of monumentality. This challenges us to consider the ways in which invisibility and absence might be implicated in remembering in Roman Britain. In relation to this, Howard Williams (2004) considers the ephemeral nature of cremation practices in Roman Britain. In this analysis, he rightly points out that the practices associated with cremating the dead may have been more significant as a ‘technology of remembrance’ than the physical presence of a permanent, enduring monument. As with the example of malangan sculpture, embodied action is key here for such ephemera to be endowed with the power of prolonged memory creation. Also important in thinking about the absent and what is not seen is the hold that these may take on people’s imaginations, as noted, for example, in the creation of legendary topographies (Holtorf and Williams 2006) and in the power of the Stolpersteine. Evidence for such can be seen across Britain at the sites of Neolithic

Memories of the Past in Roman Britain    687 and Bronze Age monuments where we find the burial of people and objects in the Roman period (see Aitchison 1988; Dark 1993; Williams 1998; Darvill 2004; Hutton 2011; Vejby 2012). One example of this phenomenon is at Silbury Hill, a Neolithic artificial chalk mound near Avebury, Wiltshire. Field (2003) has noted that each time a part of Silbury Hill has been investigated more Roman than Neolithic finds have been discovered. While this may, in some ways, not be surprising, given that Roman sites are almost always more productive than those dating to the Neolithic, there is, nevertheless, a considerable concentration of Roman activity at this hill. This activity includes the deposition of material in at least three shafts. The finds from these shafts included pottery, coins, bronze objects, glass beads and vessels, iron objects and sculpted masonry, as well as deer, sheep, cattle, pig, fox, and dog bones (Brooke and Cunnington 1896; Brooke 1910; Hutton 2011). This assemblage bears a striking resemblance to the deposits found in the pits and wells of Silchester (Fulford 2001), suggesting that those at Silbury Hill too may have been of ritual significance. In addition to these shafts, ninety-​four coins, a bronze bracelet, and Roman pottery were found in the ditch of the mound itself (Moorhead 2001: 100; Gillings and Pollard 2004: 97–​ 100). These finds are significant because they suggest that, while those engaging in activity at Silbury Hill may not have known its detailed history, and indeed may well have thought it a natural landscape feature, they may have been aware of previous human activity at the mound in the shape of the ditch. Such associations may have leant this curious mound mythological and legendary power in the Roman period, making it suited to the activity seen in the shaft deposits. Chronology seems to be key to the activities at these legendary topographies. Hutton (2011: 15–​16) notes three elements of recurrent chronological patterning in these acts of long-​term memory creation at Neolithic chambered tombs, the Wessex monuments and caves in the Bristol Channel region: (1) deposition of objects seems to have been more widespread in the Romano-​British period than in other periods; (2) there was an apparent reduction of activity in later prehistory, followed by a ‘revival’ under Roman rule; and (3)  the datable finds are overwhelmingly from the mid-​third century ad onwards. This patterning seems to be highly significant. The apparent break in activity in later prehistory adds weight to the idea that we are looking at a form of retrospective memory-​making focused around landscapes of the imagination and legendary topographies. Secondly, the concentration of these activities in the third century ad demonstrates that memory-​making can be more prevalent at certain times than at others. It is difficult to be sure what the concentration in the third century indicates. One possibility is that, as the third century ad was a period of prosperity in Roman Britain, we may be seeing here activities linked to self-​confidence and optimism about the future. In contrast, reactions to Roman domination from ad 43 onwards are widely accepted to be multiple and varied (see Mattingly 2007 on discrepant identities). Jeremy Taylor’s Atlas of Roman Rural Settlement (2007) ably demonstrates, for example, the variety of reactions we find in rural settlements and landscape organization across England. In Norfolk, for example, dramatic changes in the organization of the landscape occurred in the years immediately after the conquest (Taylor 2007: 66, 97–​101, 110). The story is

688

688   Zena Kamash very different, however, in Cornwall and the Mersey Basin, where there seems to be prolonged continuity from the Iron Age into the Roman period, with the major period of reorganization not occurring until the third century ad (Taylor 2007: 57–​59, 69–​7 1, 110–​111). In the Upper and Middle Thames Valley we see a similar pattern of continuity from the Iron Age into the Roman period, followed by a period of major reorganization and change—​this time in the second century ad. This reorganization of the landscape resulted in the construction of large settlement enclosures, trackways, and timber-​ aisled buildings—​for example, at Claydon Pike and Neigh Bridge, Somerford Keynes (Gloucestershire) (Miles et al. 2007: 377). One exception to this general pattern for the Upper and Middle Thames Valley is a small group of sites near Bicester that seem to have been abandoned in the conquest period (Henig and Booth 2000: 106). One possibility is that these sites lay within the tribal area of the Catuvellauni, rather than the Dobunni like the rest of the Upper Thames Valley (Miles et al. 2007: 376), and so the variation that we see here may be the tangible consequence of different reactions to the conquest from one tribe to another. Clearly, whenever they occurred, these episodes of major landscape reorganization would have been unsettling periods of upheaval and disruption that would have required significant social renegotiation of the landscape. Before I look at this in more detail, a caveat must be sounded here, related to the limits of the explanatory power of a memory approach. Memory is a powerful tool for suggesting potential impacts and effects on people and how they have created and re-​created their identities, but it cannot so easily provide an explanation of the causal factors in the past. In other words, while I may be able to suggest some of the effects of the second century ad reorganization on memory communities, I will not be explaining why or how this occurred—​this is a question that is worthy of more attention elsewhere. In addition to ‘angst depots’, another major method of social renegotiation in times of stress is through the medium of religion. At Claydon Pike, the reorganization of the landscape was accompanied by the delineation of a probable sacred precinct in trench 19 (Miles et al. 2007: 164–​165). Some of the activities that seem to have been performed here included the burning of incense—​the strong sensory impacts of which may have been central to memory creation here (on smell, see Drobnik 2006). As the area lacked any clear religious or sacred precursor and seems to have been created anew in the second century ad, it would seem reasonable to suggest that the memory practices associated with this site would have been prospective in nature. So, we may suggest that the response of the inhabitants of Claydon Pike to this upheaval was to create something new around which they could focus memories for the future. At Neigh Bridge, Somerford Keynes, however, there is evidence for a shrine at the river crossing, marked by extraordinarily high numbers of first–​second century ad brooches (Miles et al. 2007: 388). In contrast to Claydon Pike, this religious focus seems to have traversed the period of continuation and reorganization, suggesting that both retrospective and prospective memories may have been at play. The retrospective nature of the memories is further strengthened by the focus on water, which is widely recognized as a long-​term fascination in Britain (see Kamash 2008). What this suggests is that similar-​seeming sets of

Memories of the Past in Roman Britain    689 processes, when probed, may expose divergent memory practices, similar in nature and probably contributing to, Mattingly’s discrepant identities (2007). The question of continuity from the Iron Age into the Roman period touched on here is one of the main debates in studies of religion in Roman Britain. This debate centres mostly on whether there were precursors to Romano-​Celtic temples on particular sites (see, e.g., Hingley 1982, 1985; Henig 1984; Watt 1998; Smith 2001: 13–​16, 33–​75; Kamash et al. 2010). In addition, Williams (1998) has carried out a survey on the reuse of ancient monuments by Romano-​Celtic temples, pointing to numerous instances of Roman period ritual activity in and around Bronze Age barrows and Iron Age hillforts. In what follows, instead of thinking about reuse or continuity on particular sites, I will consider the phenomenon of building Romano-​Celtic temples more generally and explore further how memories of the past could be used to create new social constellations. The earliest known temples of this type in Britain, such as those at Harlow (Essex), Heathrow (Middlesex), and Hayling Island (Hampshire), date to the latest Iron Age immediately preceding the Roman conquest (for Harlow: Wheeler 1928; France and Gobel 1985; Bartlett 1988; for Heathrow: Grimes and Close-​Brooks 1993; for Hayling Island: Downey et al. 1980; King and Soffe 1991, 1994, 1998; Smith 2001: 40–​44). There is no obvious architectural precedent in Britain for this building form; the nearest parallels seem to come from the continent (see, e.g., King 1990). Interestingly, the later Iron Age is a period that sees numerous shifts in the uses of, and practices surrounding, material culture—​particularly that imported from the continent. In the 1980s and early 1990s, these shifts and imports were explained and discussed through the framework of world-​systems theory and core-​periphery models, leading to a focus on elite desire for goods from the Mediterranean and the all-​pervasive allure of Rome (e.g. Cunliffe 1988). Such perspectives have been increasingly challenged of late, with more agency being given to the native Britons in these exchanges. Most recently, J. D. Hill (2007) and Niall Sharples (2010) have championed the idea that the shifts we see in the archaeological record were in response to significant changes in the social context in Britain, and that the various permutations of these adoptions and changes were used to reinforce particular identities and bind people together into distinctive groups. The usual list of ‘new things’ in the late Iron Age includes distinctive burial practices, such as the Aylesford–​Swarling cremations in south-​eastern England, the use of a coinage in some areas, and the use of material culture imported from the continent. At the same time as this introduction of new practices and forms of material culture, many of the hillforts were being abandoned, leaving a large social vacuum. This social vacuum may have been partially filled by oppida settlements, but may also have been filled by Romano-​Celtic temples. Furthermore, by placing the initial impetus for Romano-​Celtic temples within these wider late Iron Age processes of change and flux, we can also give context to the otherwise unanchored appearance of Romano-​Celtic temples in Britain. The shrines in the late Iron Age religious complex at Westhampnett Area 2 are interesting in this regard (Fitzpatrick 1997: 229; 2008: 180, 254, figures 83–8​4). The site comprises 2–​4 shrines, numerous pyre sites, and c. 160 cremations. The closest parallels for the rectangular and sub-​rectangular shrines are found on hillforts, such as Danebury

690

690   Zena Kamash (Structures 2 and 3) and Cadbury Castle (Cunliffe 1984), suggesting a possible link between this style of shrine and hillforts. Sharples (2010: 116–​124) has argued, with specific reference to Wessex, that hillforts were powerful mechanisms for creating communities in the early Iron Age through a combination of employing labour as potlatch and the concomitant feasting that would accompany these building, and rebuilding, episodes (on memory and food, see Sutton 2001). The construction, and any subsequent remodelling, of Romano-​Celtic temples and their complexes would have required a similar mobilization of human effort (which may have been coerced or voluntary) and of feasting. At Marcham/​Frilford, for example, we find large deposits of animal bone and pottery in the top fills of the Iron Age double-​ditched enclosure that lay to the south of the Roman temenos. The pottery in these ditches includes some very unusual imported forms, including raspberry rouletted wares, which are dated to the period immediately post-​conquest. It seems reasonable to suggest that these deposits were the remains of feasting episodes associated with the builders of the temenos and temple. At Hayling Island (Hampshire) the timber shrine was replaced in c. ad 50–​60 by a circular temple (Downey et al. 1980; King and Soffe 1991, 1994, 1998; Smith 2001: 40–​44). The finds from this phase were dominated by animal bones and pottery, which also suggests that the people involved in this rebuilding indulged in grand feasts, similar to those argued for by Sharples (2010) at hillfort sites. Likewise, at Harlow (Essex) a masonry Romano-​ Celtic temple was constructed around 7 metres north of a pre-​Flavian circular feature (Wheeler 1928; France and Gobel 1985; Bartlett 1988). As well as coins being deposited repeatedly and continuously in both phases, there was also repeated deposition of animal bones (Haselgrove 1989: 76; Smith 2001: 85–​86). The animal bone evidence demonstrated continued preference for sheep across both phases and selectivity in side of animal (right-​ over left-​handed elements) (Smith 2001: 86). The sheep also seem to have been killed at particular times of year. This evidence suggests that the feasting may originally have been associated with episodes of building and rebuilding. The apparently annual kill cycle suggests that feasting may then have been used as a device to evoke the memory of labouring episodes and, in turn, to create new memory ties for younger generations. Anthony King (2005: 357) has also noted that the high percentages of sheep/​goat remains at his Group A sites (e.g. Hayling Island, Uley, and Lowbury) show a temptingly similar pattern to Iron Age sites. Although he ultimately goes on to dismiss this pattern as being evidence for cultural similarity—​citing further similarities with sites in Gaul and Germany—​this actually seems to reinforce the argument for linking these temple sites not only to earlier Iron Age practices, but also to the late Iron Age processes of continental borrowing—​that is, this borrowing may have been facilitated by fitting into established patterns of practice that included feasting habits (see also T. Moore, this volume, for discussion of continental influences on Britain). In these ways, then, Romano-​ Celtic temples seem to have been created as a new space for gathering together people and memories. This was, however, no simple renegotiation and replacement. First, while patterns of practice (such as feasting) may be repeated over time, this does not mean that they

Memories of the Past in Roman Britain    691 are immutable and monolithic; rather, they can be simultaneously both old and new, both retrospective and prospective. Secondly, although there are a handful of verified instances of Romano-​Celtic temples that were situated within Iron Age hillfort enclosures (Williams 1998; Smith 2001:  151), the vast majority of Romano-​Celtic temples (over eighty out of ninety known sites) positioned themselves in relation to deeper Bronze Age and Neolithic pasts, avoiding the more recent Iron Age past (Kamash, forthcoming a). One implication behind this spatial configuration and patterning is that the collective memories generated by these sites were being directed physically away from the more controversial recent past and towards an earlier, possibly more idealized, past. Another possibility is that by not building directly over the direct past this may have been a powerful form of protest in itself: leaving behind a physical reminder of what had to be left behind in the face of major social upheavals. These spaces are both abandoned and empty and yet still living in tension with the newly created spaces. We see, then, selectivity and manipulation of memory in referencing the past at, and through, Romano-​Celtic temples—​in terms of both material practices and landscape placement.

Conclusions Overall, using memory as an approach to understanding the patterns of practice that we see archaeologically in Roman Britain opens up a space for exploring some of the concerns (and the effects of those concerns) of the people living in Roman Britain. To some extent we may also be able to use memory to probe the motivations that lay behind those practices, though, as I noted earlier, the power of a memory approach probably does not lie in its ability to provide direct explanations or causal factors—​memory is too slippery for such specificity. Memory is very useful, however, in demonstrating the variety of ways in which people have engaged with the past and the variety of pasts that may have been created through those enacted memories. What people chose to remember or selected as the focus of their memory practices could vary widely even between neighbours, such as the inhabitants of Claydon Pike and Neighford Bridge. What seems to link these choices seems invariably to be related to present concerns and so we often find evidence for a rich interplay between retrospective and prospective memories. In this way, the enactment of memories in Britain seems similar to that in the wider Roman world. The major difference that we see between Britain and the rest of the empire revolves around the links between visibility and memory. Visibility of memories is, perhaps implicitly, central to most discussions of memory in the Roman world. In Britain, however, we find numerous instances of invisible, buried memories whose power lies precisely in their lack of visibility and their absence. Whether this represents an actual difference between Britain and elsewhere in the Roman Empire, or whether this belies a difference in evidence and approach, remains to be seen. This suggests, then, that we may be only at the beginning of memory in the Roman world.

692

692   Zena Kamash

Acknowledgements I owe a great debt of thanks to Professor Karl Galinsky and the Max Planck Society for providing generous funding and support for this work through the Memoria Romana project. My thanks also go to my graduate students for helping me develop various ideas expressed here in our numerous discussions. In addition, I am very grateful to the editors of this volume for their insightful and helpful comments on an earlier version of this chapter.

References Primary Sources Augustine, Confessions, trans. E. B. Pusey. New York: Pocket Books, 1907. Cicero, De Oratore, trans. E. W. Sutton. Loeb Classical Library. Cambridge, MA:  Harvard University Press, 1967. Quintilian, Institutio Oratoria, trans. H. E. Butler. Loeb Classical Library. Cambridge, MA: Harvard University Press, 1920–​2. Rhetorica ad Herrenium, trans. H. Caplan. Loeb Classical Library. Cambridge, MA: Harvard University Press, 1954.

Secondary Sources Aitchison, N. B. (1988). ‘Roman Wealth, Native Ritual: Coin Hoards within and beyond Roman Britain’, World Archaeology, 20: 270–2​83. Alcock, S. E. (2002). Archaeologies of the Greek Past: Landscape, Monuments and Memories. Cambridge: Cambridge University Press. Assmann, A. (1999). Erinnerungsräume: Formen und Wandlungen des kulturellen Gedächtnisses. Munich: C. H. Beck. Assmann, J. (1999). Moses the Egyptian:  The Memory of Egypt in Western Monotheism. Cambridge, MA: Harvard University Press. Barchiesi, A. (forthcoming). The War for Italia:  Conflict and Collective Memory in Vergil’s Aeneid. Berkeley and Los Angeles: University of California Press. Barrett, J. (1993). ‘Chronologies of Remembrance: The Interpretation of Some Roman Inscriptions’, World Archaeology, 25/​2: 236–​247. Bartlett, R. (1988). ‘The Harlow Celtic Temple’, Current Archaeology, 112: 163–​166. Borges, J. L. (2000). Labyrinths. London: Penguin Modern Classics. Borić, D. (2010) (ed.). Archaeology and Memory. Oxford: Oxbow Books. Brooke, J. W. (1910). ‘The Excavation of a Roman Well near Silbury Hill’. Wiltshire Archaeological and Natural History Magazine, 36: 373–​375. Brooke, J. W., and Cunnington, B. H. (1896). ‘Excavation of a Roman Well near Silbury Hill’, Wiltshire Archaeological and Natural History Magazine, 29: 166–​171. Confino, A. (1997). ‘Collective Memory and Cultural History: Problems of Method’, American Historical Review, 102: 1386–​1404.

Memories of the Past in Roman Britain    693 Connerton, P. (1998). How Societies Remember. Cambridge: Cambridge University Press. Cunliffe, B. (1984). Danebury:  An Iron Age Hillfort in Hampshire. Volumes 1 and 2.  CBA Research Report 52. London: Council for British Archaeology. Cunliffe, B. (1988). Greeks, Romans and Barbarians: Spheres of Interaction. London: Batsford. Dark, K. R. (1993). ‘Roman-​Period Activity at Prehistoric Ritual Monuments in Britain and the Armorican Peninsula’, in E. Scott (ed.), Theoretical Roman Archaeology First Conference Proceedings. Aldershot: Avebury, 133–​146. Darvill, T. (2004). Long Barrows of the Cotswolds and Surrounding Areas. Stroud: Tempus. Downey, R, King, A., and Soffe, G. (1980). ‘The Hayling Island Temple and Religious Connections across the Channel’, in W. Rodwell (ed.), Temples, Churches and Religion in Roman Britain. BAR British Series 77(i). Oxford: Archaeopress, 289–​302. Drobnik, J. (2006). ‘Preface: Part VII: Sublime Essences’, in J. Drobnik (ed.), The Smell: Culture Reader. Oxford: Berg, 371–​374. Eckardt, H. (2004). ‘Remembering and Forgetting in the Roman Provinces’, in B. Croxford, H. Eckardt, J. Meade, and J. Weekes (eds), TRAC 2003: Proceedings of the Thirteenth Annual Theoretical Roman Archaeology Conference, Leicester 2003. Oxford: Oxbow Books, 36–​50. Fabrizi, V. (2012). Mores Veteresque Novosque: La Rappresentazione del Passato e del Presente di Roma negli Annales di Ennio. Pisa: ETS. Farrell, J. (1997). ‘The Phenomenology of Memory in Roman Culture’, Classical Journal, 92: 373–​383. Farrell, J. and Nelis, D. (2013) (eds). Augustan Poetry and the Roman Republic. Oxford: Oxford University Press. Field, D. (2003). ‘Silbury Hill’, British Archaeology, 70: 15–​17. Fitzpatrick, A. (1997/2008). Archaeological Excavations on the Route of the A27 Westhampnett Bypass, West Sussex, 1992. Volumes 1 and 2. Salisbury: Trust for Wessex Archaeology. Flower, H. (1998). ‘Rethinking damnatio memoriae: The Case of Cn. Calpurnius Piso in ad 20’, Classical Antiquity, 17: 155–​186. Flower, H. (2006). The Art of Forgetting: Disgrace and Oblivion in Roman Political Culture. Chapel Hill, NC: University of North Carolina Press. France, N. E., and Gobel, B. M. (1985). The Romano-​British Temple at Harlow. Gloucester: Alan Sutton. Fulford, M. (2001). ‘Links with the Past:  Pervasive “Ritual” Behaviour in Roman Britain’, Britannia, 32: 199–​218. Fulford, M., and Holbrook, N. (2011). ‘Assessing the Contribution of Commercial Archaeology to the Study of the Roman Period in England, 1990–​2004’, Antiquaries Journal, 91: 323–​345. Galinsky, K. (2009) Memoria Romana website. (accessed May 2013). Galinsky, K. (2014) (ed.). Memoria Romana:  Memory in Rome and Rome in Memory. Rome: Memoirs of American Academy in Rome. Galinsky, K. (2016) (ed.). Cultural Memories in the Roman Empire. Los Angeles: Getty Publications. Gillings, M., and Pollard, J. (2004). Avebury. London: Duckworth. Gowing, A. M. (2005). Empire and Memory:  The Representation of the Roman Republic in Imperial Culture. Cambridge: Cambridge University Press. Grimes, W. F., and Close-​ Brooks, J. (1993). ‘Excavation of Caesar’s Camp, Heathrow, Harmondsworth, Middlesex, 1944’, Proceedings of the Prehistoric Society, 59: 303–​360.

694

694   Zena Kamash Gutteridge, A. (2010). ‘The Depiction of Time on the Arch of Constantine’, in D. Borić (ed.), Archaeology and Memory. Oxford: Oxbow Books, 158–​170. Halbwachs, M. (1992). On Collective Memory, ed. and trans. L. A. Coser. Chicago: University of Chicago Press. Haselgrove, C. (1989). ‘Iron Age Coin Deposition at Harlow Temple, Essex’, Oxford Journal of Archaeology, 8/​1: 73–​88. Henig, M. (1984). Religion in Roman Britain. London: Batsford. Henig, M., and Booth, P. (2000). Roman Oxfordshire. Stroud: Sutton Publishing. Hill, J. D. (2007). ‘The Dynamics of Social Change in Later Iron Age Eastern and South-​Eastern England c.300 bc–​ad 43’, in C. Haselgrove and T. Moore (eds), The Later Iron Age in Britain and Beyond. Oxford: Oxbow Books, 16–​40. Hingley, R. (1982). ‘Recent Discoveries of the Roman Period at the Noah’s Ark Inn, Frilford, South Oxfordshire’, Britannia, 13: 305–​309. Hingley, R. (1985). ‘Location, Function and Status: A Romano-​British “Religious Complex” at the Noah’s Ark Inn, Frilford (Oxfordshire)’, Oxford Journal of Archaeology, 4/​2: 201–​214. Holtorf, C., and Williams, H. (2006). ‘Landscapes and Memories’, in D. Hicks and M. C. Beaudry (eds), The Cambridge Companion to Historical Archaeology. Cambridge: Cambridge University Press, 235–​254. Hutton, R. (2011). ‘Romano-​British Reuse of Prehistoric Ritual Sites’, Britannia, 4: 1–​22. Jones, A. (2007). Memory and Material Culture. Cambridge: Cambridge University Press. Kamash, Z. (2008). ‘What Lies Beneath? Perceptions of the Ontological Paradox of Water’, World Archaeology, 40/​2: 224–​235. Kamash, Z (2016) ‘From the Individual to the Collective: Memory Practices on Religious Sites in Roman Britain’, in K. Galinsky (ed.) Cultural Memories in the Roman Empire. Los Angeles: Getty Museum Press, 153–169. Kamash, Z. (forthcoming). ‘Memory, Materiality and Religion in Roman Britain’, in T. Clack (ed.), Archaeology, Syncretism and Creolisation. Oxford: Oxford University Press. Kamash, Z., Gosden, C., and Lock, G. (2010). ‘Continuity and Religious Practices in Roman Britain:  The Case of the Rural Religious Complex at Marcham/​Frilford, Oxfordshire’, Britannia, 41: 95–​125. King, A. (1990). Roman Gaul and Germany. Berkeley and Los Angeles:  University of California Press. King, A. (2005). ‘Animal Remains from Temples in Roman Britain’, Britannia, 36: 329–​369. King, A., and Soffe, G. (1991). ‘Hayling Island’, in R. F. L. Jones (ed.), Britain in the Roman Period: Recent Trends. Sheffield: J. R. Collis Publications, 111–​113. King, A., and Soffe, G. (1994). ‘The Iron Age and Roman Temple on Hayling Island’, in A. Fitzpatrick and E. L. Morris (eds), The Iron Age in Wessex: Recent Work. Salisbury: Trust for Wessex Archaeology, 114–​116. King, A., and Soffe, G. (1998). ‘Internal Organisation and Deposition at the Iron Age Temple on Hayling Island’, Proceedings of the Hampshire Field Club Archaeological Society, 53: 35–​47. Klein, K. (2000). ‘On the Emergence of Memory in Historical Discourse’, Representations, 69: 127–​150. Koortbojian, M. (1996). ‘In commemorationem mortuorum: Text and Image along the “Streets of Tombs”’, in J. Elsner (ed.), Art and Text in Roman Culture. Cambridge:  Cambridge University Press, 210–​233. Küchler, S. (1987). ‘Malangan: Art and Memory in a Melanesian Society’, Man, 22: 238–​255.

Memories of the Past in Roman Britain    695 Küchler, S. (1988). ‘Malangan: Objects, Sacrifice and the Production of Memory’, American Ethnologist, 15: 625–​637. Le Goff, J. (1992). History and Memory, trans. S. Randall and E. Claman. New York: Columbia University Press. Li Causi, P. (2012). Il Riconoscimento e il Ricordo: Fama e Memoria nel De Beneficiis di Seneca. Palermo: Palumbo Editore. Lillios, K. T., and Tsamis, V. (2010) (eds). Material Mnemonics: Everyday Memory in Prehistoric Europe. Oxford: Oxbow Books. Mattingly, D. (2007). An Imperial Possession: Britain in the Roman Empire. London: Penguin. Miles, D., Palmer, S., Smith, A., and Perpetua Jones, G. (2007). Iron Age and Roman Settlement in the Upper Thames Valley: Excavations at Claydon Pike and Other Sites within the Cotswold Water Park. Thames Valley Landscapes Monographs 26. Oxford: Oxford Archaeology. Moorhead, T. S. N. (2001). ‘Roman Coin Finds from Wiltshire’, in P. Ellis (ed.), Roman Wiltshire and After. Devizes: Wiltshire Archaeological and Natural History Society, 85–​107. Nora, P. (1978). ‘Mémoire collective’, in J. Le Goff, R. Chartier, and J. Revel (eds), La Nouvelle Histoire. Paris: Retz. Nora, P. (1984–​92). Lieux de Mémoire. Paris: Gallimard. Olick, J., Vinitzky-​Seroussi, V., and Levy, D. (2011) (eds). The Collective Memory Reader. Oxford: Oxford University Press. Parker, E. S., Cahill, L., and McGaugh, J. L. (2006). ‘A Case of Unusual Autobiographical Remembering’, Neurocase, 12: 35–​49. Pearson, J. (2010). ‘Nazi Victims and Stumbling Blocks to Memory’ (accessed May 2013). Radley, A. (1990). ‘Artefacts, Memory and a Sense of the Past’, in D. Middleton and D. Edwards (eds), Collective Remembering. London: Sage Publications, 46–​59. Raymond, E. (2013) (ed.). Vox Poetae: Manifestations Auctoriales dans l’Epopée Gréco-​Latine. Lyon: CECOR. Rebeggiani, S. (2013). ‘Reading the Republican Forum: Virgil’s Aeneid, the Dioscuri and the Battle of Lake Regillus’, Classical Philology, 108/1: 53-​69. Renfrew, C., Gosden, C., and DeMarrais, E. (2004). ‘Introduction: Art as Archaeology and Archaeology as Art’, in C. Renfrew, C. Gosden, and E. DeMarrais (eds), Substance, Memory, Display: Archaeology and Art. Cambridge: McDonald Institute Monographs, 1–​6. Rowlands, M. (1993). ‘The Role of Memory in the Transmission of Culture’, World Archaeology, 25: 141–​151. Schwartz, B. (1982). ‘The Social Context of Commemoration: A Study in Collective Memory’, Social Forces, 61/​2: 374–​402. Seider, A. (forthcoming). Memory in Vergil’s Aeneid: Creating the Past. Cambridge: Cambridge University Press. Sharples, N. (2010). Social Relations in Later Prehistory: Wessex in the First Millennium bc. Oxford: Oxford University Press. Smith, A. (2001). The Differential Use of Constructed Sacred Space in Southern Britain from the late Iron Age to the 4th Century ad. BAR British Series 318. Oxford: Archaeopress. Sutton, D. (2001). Remembrance of Repasts:  An Anthropology of Food and Memory. Oxford: Berg. Tatum, J. (1995). ‘Aunt Elvie’s Quilt on the Bed of Odysseus: The Role of Artifacts in Natural Memory’, Helios, 22/​2: 167–​174.

696

696   Zena Kamash Taylor, J. (2007). An Atlas of Roman Rural Settlement in England. CBA Research Report 151. York: Council for British Archaeology. Todorov, T. (1995). Les Abus de la Mémoire. Paris: Arlea. Van Dyke, R., and Alcock, S. E. (2003) (eds). Archaeologies of Memory. Oxford: Blackwell. Vejby, M. (2012). ‘Reinterpreting the Iron Age and Roman Reuse of Megalithic Tombs in Atlantic Europe’. Unpublished Ph.D. thesis, University of Reading. Watt, D. (1998). Religion in Late Roman Britain: Forces of Change. London: Routledge. Wheeler, R. E.  M. (1928). ‘A Romano-​Celtic Temple near Harlow and a Note on the Type’, Antiquaries Journal, 8: 300–​326. Williams, H. M. R. (1998). ‘The Ancient Monument in Romano-​British Ritual Practices’, in C. Forcey, J. Hawthorne, and R. Witcher (eds), TRAC 97: Proceedings of the Seventh Annual Theoretical Roman Archaeology Conference. Oxford: Oxbow Books, 71–​86. Williams, H. M.  R. (2004). ‘Ephemeral Monuments and Social Memory in Early Roman Britain’, in B. Croxford, H. Eckardt, J. Meade, and J. Weekes (eds), TRAC 2003: Proceedings of the Thirteenth Annual Theoretical Roman Archaeology Conference, Leicester 2003. Oxford: Oxbow Books, 51–​61. Williams, H. M. R., and Bradley, R. (1998) (eds). The Past in the Past: The Re-​use of Ancient Monuments (World Archaeology 30(1)). London: Routledge.

PA RT  I V

LANDSCAPE AND E C ON OM Y

698

Chapter 34

‘By Sm all T h i ng s Reveal e d ’ Rural Settlement and Society Martin Millett

Introduction It is a paradox that, while the vast majority of the population of Roman Britain were farmers, and certainly the bulk of its landscape was given over to some form of agriculture, rural settlement has attracted comparatively little attention, at least until recently. Perhaps even more surprising is that most past scholarship is heavily biased towards one particular social group—​those who dwelt in villas. This is like looking at the modern countryside through the lens of those who live in the Cotswolds and drive Range Rovers! To gain a more balanced perspective, we are faced with a serious problem of a lack of good evidence from the full range of ordinary rural sites. This imbalance has been to some extent rectified in recent years by archaeological work that has been undertaken in advance of building development (see Wilson, this volume), since the locations of such projects have been determined by external factors and hence they provide something that may approximate to a random sample of the landscape. The results of such work have already contributed to one major review of the rural landscape (Taylor 2007) and they are currently being analysed in considerable detail in a project run by the University of Reading and Cotswold Archaeology, which has yet to produce its definitive results (http://​www.cotswoldarchaeology.co.uk/​developer-​funded-​roman-​archaeology-​in-​ england/​). Pending the completion of this work, the present chapter provides an overview of some key issues together with a discussion of research from one small area that perhaps provides some complementary insights. However, this should be read in an awareness of the issue of regional variation, for it is now very clear that there were tremendous differences in the character of rural settlement across Roman Britain, so much so that

700

700   Martin Millett there is really no such thing as the Romano-​British countryside. It is also important to understand that such geographical variation is not simply a matter of environmental constraints but reveals fundamental differences in the social organization and lifestyles of the populations. At the outset we should acknowledge that Roman Britain was primarily a rural society. Assessing the total size of the population is very difficult, but at a rough estimate it amounted to perhaps 3.6 million people of whom about 3.4 per cent were associated with the military, 6.5 per cent were urban dwellers (using the broadest definition of towns), with the remaining 90 per cent living in the countryside (Millett 1990: 185; Mattingly 2006: 356 suggests a total population of about two million). These numbers are based on the evidence from a series of rural surveys undertaken in the latter part of the twentieth century and they probably need to be updated. Taylor, for instance, recorded about 28,000 known rural settlements in England alone (Taylor 2007: 12, table 3.2). Whatever the exact figure, the point of importance is that much of the countryside was teaming with people and very different from some past images of a largely wooded place with clearings for the occasional settlement. The whole landscape had been exploited in a range of different modes by a large population since c. 1000 bc, and, given the reliance on manual labour for farming and other rural activities, any traveller passing along a Roman road through Britain will have been aware of living and working people almost everywhere. Furthermore, as Taylor (2103) has argued, these people must be recognized as active agents in the construction of the cultures of Roman Britain.

The Character of the Countryside Rural areas were a product of a long and varied history, with the types of buildings and the character of the settlement patterns and fields being the result of this development and the inhabitants’ intricate relationship with the natural environment. Thus the countryside was a patchwork onto which we have only comparatively small windows of knowledge. The best are generally provided by landscape-​scale evidence, mostly the product of aerial photography or large-​scale excavation (for instance, in gravel quarries), and now occasional large-​scale geophysical surveys. All these methods are limited by a focus on the features—​such as settlements, fields, and trackways—​which are a product of human exploitation. Although, as we shall see in a moment, these provide key insights into the character of the countryside, it is important to appreciate that there was more to the rural world than those things associated with farming and habitation. On the one hand, there were focal structures that had other functions: forts and frontiers were located in the countryside and there were also a significant number of temples and sanctuaries that will have acted as key points for the gathering of people. On the other, paths, roads, trackways, and navigable rivers created linkages along which people, animals, and things constantly moved back and forth.

‘By Small Things Revealed’: Rural Settlement and Society    701 There were also substantial tracts that were not used for arable farming. Aside from areas for pasture—​largely unenclosed—​there were also moorlands, marshlands, and woodlands, all of which were exploited in different ways. There is, for instance, strong evidence for the production of salt, the digging of peat, and the management of woodland through coppicing and pollarding. Wood and charcoal were, after all, key products, although they are very difficult to trace archaeologically. Equally, the ground itself was mined and quarried for its resources. Metal-​ore extraction, although geologically circumscribed, was of key importance (see also Dungworth, this volume). More widespread was the quarrying of stone for building and other purposes, and the digging for clay for the production of tiles and pots. These examples demonstrate two key points. First, much of what we think of today as industry was rurally located, and, second, the boundary between farming and other modes of exploiting the land was much more blurred than we might now think. It is especially important to appreciate that, in a pre-​industrial society, a large supply of labour was vital to many rural activities. Furthermore, labour requirements were seasonally variable, with many more people needed for harvest than, for instance, for ploughing. Hence, it is very likely that there were complex systems that allowed these demands to be met. Aside from social structures, which may have involved the payment of some dues in labour, we can envisage estates or other communities where activities like quarrying or potting or woodmanship were seasonal and complementary with the demands of arable farming or stock-​raising. This further highlights an aspect of Roman Britain that remains obscure—​the nature of land tenure. The existence of occasional land-​sale documents (Turner 1956, RIB 2443.13; Tomlin 1996) and the odd textual reference (Theodosian Code XI. 7, 2) confirm that at least some property was privately owned, although the extent to which this was the case remains unclear. Some have argued for the existence of forms of multiple ownership or ‘Celtic’ systems of partible inheritance (Stevens 1966)—​although this is somewhat undermined by the discovery of a will formulated using Roman law in Trawsfynydd, Merionethshire (Tomlin and Hassall 2004: 347–​349), while many assume the growth of large private estates in the hands of a later Roman aristocracy (Scott 2000: 106–​120). All or any of these may have applied; we simply do not know, and we certainly cannot assume that there was any single application of supposed Roman legal norms. Indeed, wide variations in the patterns of rural settlement witnessed in the archaeological evidence suggest that there was a range of practice. It is important to acknowledge this level of ignorance, since many past discussions of villas, in particular, have been founded on very simplistic notions of landownership that are almost certainly anachronistic. It is worth pausing for a moment to look at some examples of well-​explored landscapes that illustrate the variation across the country. One of the most remarkable pieces of archaeological work in recent years has been Dominic Powleslend’s geophysical survey along the southern side of the Vale of Pickering in Yorkshire (http://​thelrc.wordpress. com/​lrc-​digital-​atlas/​). This project involved mapping 1,250 hectares of a multi-​period landscape using magnetometry (Figure 34.1). From the point of view of understanding Roman rural settlement, the results are stunning. Although many had previously

702

Figure 34.1  Magnetometer survey of part of the Roman landscape in the Vale of Pickering, North Yorkshire. Across the top of the image a trackway runs east–​west, flanked on either side by settlement enclosures and small fields. Beyond these to the south are the fainter traces of a series of fields bounded by ditches. Superimposed on these, in the northern zone, the back dots represent grubenhaüser (sunken-​featured buildings) of Early Medieval date. To the south the land rises towards the Yorkshire Wolds and the ancient landscape is obscured by deposits of wind-​blown sand. Source: courtesy of Dominic Powlesland, Landscape Research Centre.

‘By Small Things Revealed’: Rural Settlement and Society    703 concluded that the landscape was densely occupied and wholly exploited during this period, there had previously been no such large-​scale survey to allow a full appreciation of this. A snapshot of part of the landscape is shown in Figure 34.1, where we see a trackway with adjacent enclosures running from east to west across the image—​this is part of a system that has been traced for many kilometres, and many of the enclosures were used for human habitation, as is illustrated by the density of features revealed. To the north lie the wetlands at the centre of the Vale, which were seasonally flooded yet provided important pasture. To the south the land rises towards the Yorkshire Wolds, and is partly obscured (and protected) by a blanket of wind-​blown sand. Visible are field enclosures and other features, as well as spots representing Anglo-​Saxon grubenhaüser. The remarkable thing is not the density of landscape use shown, but the fact that it is continuous across the whole area surveyed. This shows how we should now understand the whole of Roman Britain. A contrasting example, based on aerial photographic evidence, is provided by the regional study of the Magnesian Limestone ridge in South and West Yorkshire (Roberts 2010). Here there is also evidence for the use of the whole landscape, but with an extensive system of small fields and enclosures that originated in the later Iron Age, dominating the landscape and indicating a system of dispersed settlements. A few of these developed into villas in the third and fourth centuries, but for the most part this landscape too continued to be dominated by indigenous-​style farmsteads.

The Significance of the Villa It is in the context of this patchwork of intensive and varied rural activities that we should understand villas. The term is derived from Latin but has gained an archaeological meaning of its own, so almost any Roman-​style rural house is labelled with the term—​although very few indeed of those found in Roman Britain would have been seen as villas by a visitor from Italy. In archaeological terms, their physical remains are fairly clearly defined, and many of the houses share common architectural characteristics ultimately derived from continental (although rarely Mediterranean) models. In broad terms, there are perhaps 2,000 such sites, representing perhaps 1 per cent of known rural sites, largely distributed over the southern and eastern counties of England, to one side of a line between the mouths of the rivers Tees and Exe. In this sense they clearly represent both an aspiration towards a particular set of cultural values—​arguably associated with Roman power—​ and also a decision by their owners to expend significant capital resources on this sort of permanent building. We can rarely tell whether their owners were indigenous people, or incomers, or those returning after a period elsewhere: what is clear is that the vast majority show continuity of site from previously occupied settlements. It is equally clear that the buildings themselves provide little or no information about how the wealth deployed to construct them was created, only how it was spent. The resources may have been generated from economic activity (for example, farming, mining, trade) whether locally

704

704   Martin Millett or elsewhere, or they may have been inherited, or extorted: we simply cannot tell. Thus the common jump from villa house to the assumed associated villa estate, and thence to a particular reconstruction of the Romano-​British economy (usually in terms of the exploitation of marketing opportunities in towns generated by the Pax Romana), is self-​ evidently unjustifiable. It is also difficult to reconcile this common explanation with evidence that shows that, in Britain, villas start to be built early in the Roman period (often at a large scale) but spread relatively slowly, becoming on average smaller through time. This suggests the spread of a fashion for a particular type of elite display, not the development of a particular economic mode (Taylor 2011). Equally, the distribution of villas is uneven, with some regional concentrations (for instance, in the Cotswolds) and a general spatial association with towns that we understand to be important local (not provincial) political centres. As such they may perhaps be associated with the houses of magnates who fulfilled roles in civic administration. This is not to deny that most villas had farms and estates attached to them: it is merely to assert that this association is not necessarily causative. Members of the council of a civitas will often have owned farms, for this was the Roman mode of demonstrating wealth: that does not mean that they derived their principal wealth from farming. This suggests that we should be looking at villas, not as a crude economic indicator, but rather for the insights they provide about the social organization of Roman Britain. Attempts to approach this on the basis of a study of villa plans has generated some insights (Smith 1997; cf. Rippengal 1993), in particular through encouraging us to look at the building plans critically, rather than on the basis of loose analogies with aristocratic country houses of more modern periods. In particular, interesting studies have suggested how some major villas of the later Roman period may have worked within the framework of elite Mediterranean classical culture, with an especial focus on dining rituals and a familiarity with classical mythology and literature (Scott 2000; Gerrard 2013: 141–​142). Other studies (Smith 1997), which have emphasized possible patterns of multiple occupancy at certain villas, have met with less acclaim, although they do highlight the difficulty in reading plans in terms of a simple nucleated family. Two contrasting examples illustrate some of the key issues. At one extreme, we have Fishbourne, near Chichester in West Sussex—​often misleadingly referred to as a ‘palace’ (Figure 34.2). Although there is some debate about the precise date of its construction (Cunliffe 1971: 77; Black 1987: 84–​86), it is clear that an enormous villa was constructed over a brief period in the period around ad 70. It was within a complex best interpreted as a late Iron Age oppidum, which seems to have been used for an invasion period military base. The exact relationship between an indigenous ruler, the Roman invaders, and the establishment of civitas government here is contentious, but it is very clear that the villa represents a symbol of acceptance of Roman ways at a very early date. Unlike any other known villa in Britain, its opulent architecture displays close links with the contemporary Mediterranean, suggesting both the expenditure of immense wealth within a very short time and close personal ties between its owner and the imperial household. The principal excavated courtyard is only part of a much larger complex, which includes ancillary structures that are very difficult to understand, but indicate a large-​scale

eam Str

Ditch Aqueduct Road/path

Masonry building Road/path

N

0

10

50m

Figure 34.2  Plan of the Roman villa at Fishbourne, Sussex, in the later first century ad showing the approach from the east and associated structures. Source: drawn by Lacey Wallace, after Cunliffe (1971); Cunliffe et al. (1996); Manley and Rudkin (2003).

706

706   Martin Millett monumentalization of the landscape (Cunliffe et al. 1996; Manley and Rudkin 2003). Interestingly, although the architectural detail and decoration signify Italian influence, the plan is not easy to parallel in Italy, and includes features such as an audience hall that surely betray use by a broader community and not just a pro-​Roman potentate. By contrast, the villa at Frocester is much more typical of other villas in Britain, except that it has been subject to a programme of very thorough long-​term research, which has included a close study of the surrounding landscape (Gracie 1970; Gracie and Price 1979; Price 2000–​10) (Figure 34.3). The site, which lies on the edge of the Severn valley in Gloucestershire just below the Cotswold escarpment, is one of a large group in this part of England. It shows a more gradual evolution from an enclosed Iron Age farm, which was remodelled in the second century, but only became a recognizable villa towards the end of the third century ad. The initial villa house comprised a simple range, with a large reception room near its centre. This was added to in the fourth century with the creation of a façade comprising portico and pavilions, and the addition of a second range of rooms along the back of the house, as well as perhaps a hall at one end. There is very good evidence to suggest that the house was of more than one storey and thus built to create a strong impression to the visitor. This gentrification is also seen within, as several rooms were decorated with mosaic pavements (Neal and Cosh 2010: 136–​139, nos 430.1–​4). The frontal aspect of the architecture was further enhanced around ad 340 with the laying-​out of a formal garden. Thus we see a relatively ordinary farm monumentalized and made into a show piece within a few generations late in the Roman period. The building plan, although showing Roman influence, was characteristic of Britain and the north-​western provinces, and the form again suggests that receiving, entertaining, and impressing visitors were crucial to its owners.

The Diversity of Farmsteads In contrast to villas, ordinary farmstead sites are much more difficult to evaluate. This is partly simply because they were mostly built in timber with the result that the houses, especially the less monumental types, are much less legible as structures. It is clear that, while pre-​Roman Iron Age housing was dominated by the round-​house, a broader range of mostly rectilinear structures occurred in the Roman period (e.g. Roberts 2010: 73–​76). There were pockets of continued round-​house use, mostly in association with other structural forms (Taylor 2007: 31, table 4.4). Equally, there is much variety in other types of architecture, with few really distinctive and widespread types, like the aisled hall or the stone-​built hall-​house. Furthermore, farms vary from small places with a few buildings to more extensive settlements with multiple structures. A common theme is the use of enclosure to define space within settlements, suggesting that the definition of space was important to the inhabitants. The most sustained regional study of rural settlement in Roman Britain is probably the Danebury Environs project (Cunliffe 2008). Here, the main focus was on examining

Late Iron Age

c. ad 50–100

c. ad 100–150

c. ad 150–200

c. ad 200–275

c. ad 275–360

N Building Gravel pit Ditch Ditch postulated or disused Road c. ad 360–400

0

100m

0

300ft

Figure 34.3  Plans showing the development of the Roman villa and associated structures at Frocester Court, Gloucestershire. Source: drawn by Lacey Wallace, based on Price (2000–​10).

708

708   Martin Millett a series of villa sites, but the work included the excavation of one ordinary rural site at Rowbury Farm, which was located within a field system of Bronze Age origin; the settlement itself did not have any intelligible structures surviving from the Roman period, and it appears that it was used for livestock management rather than occupation, going out of use early in the Roman period. The limited evidence from the Roman period led to the suggestion that the site was either a subsidiary of a nearby villa or a peasant settlement (Cunliffe and Poole 2008: 132–​133). What the whole project provided was an intensive review of the development of rural society in one particular area, revealing a range of different histories and structural types, even within the category of the smaller villas, but with apparent evidence for a broad continuity in agrarian traditions from the Iron Age with little evident Roman impact.

The Issue of Nucleated Sites While accounts have traditionally given emphasis to single farms or smaller clusters of building, recent work has shed more light on the existence of larger nucleated settlements. Taylor’s Atlas has demonstrated that this is partly a regional matter, with a pronounced preference for settlement nucleation is some areas like the East Midlands (Taylor 2007: 76, figure 34.6.5). This is reminiscent of the patterns of later periods, when there is a marked contrast between landscapes dominated by villages, on the one hand, and dispersed farmsteads, on the other, associated with different social formations. While this distinction may hold some value for Roman Britain, it would be a mistake to impose such a simple division or to transpose models, derived from the study of the medieval peasant, onto the interpretation of Roman Britain (cf. McCarthy 2013). The social structure of those living in Roman Britain was the product of particular historical circumstances that were very different from those of later periods, so we should be cautious even of referring to nucleated sites as villages. The recently excavated site at Higham Ferrers, beside the river Nene in Northamptonshire, provides an interesting example of an east Midlands nucleated site (Lawrence and Smith 2009). The Roman settlement, which overlies an Iron Age farm, was laid out in the second century but reached its full size in the later third and fourth centuries (Figure 34.4). It comprises a series of rectangular stone-​founded houses, set within enclosures with paddocks and horticultural plots, which lined a minor road that followed the river valley. There was a small shrine at one side of the excavated area, although it is seems most likely that this served the community rather than being a more significant rural sanctuary. The buildings (Figure 34.5) are otherwise relatively undifferentiated, and the carefully analysed evidence from the excavations suggests that this was a primarily agrarian settlement. There is no suggestion of its dependence on any villa or other major house that might have been occupied by the landowner or their agent. This perhaps implies a settlement of the ‘middling sort’, although it is interesting to note the find of slave shackles, which might indicate that we are dealing with a different type of estate centre.

‘By Small Things Revealed’: Rural Settlement and Society    709

Figure 34.4  Plan of the nucleated rural settlement at Higham Ferrers, Northamptonshire, in the late third–​fourth centuries ad. A series of buildings set within enclosures flank the Roman road, which runs along the valley of the river Nene just above the flood plain. Source: illustration from Lawrence and Smith (2009), reproduced courtesy of Oxford Archaeology.

There are many general similarities with a larger example, at Kingscote in Gloucestershire (Timby 1998), which is also dominated by simple rectangular houses, this time organized in a couple of broadly rectilinear groupings. This site is larger, with as many as seventy-​five houses indicated by survey work. In contrast to Higham Ferrers, there is some evidence for a single larger house, which was located to one side of the main group of buildings and seems to date to the fourth century. This had a main hall, not unlike

710

710   Martin Millett

Figure 34.5  Photograph of a domestic building (Figure 34.4, No. 10810) within the nucleated rural settlement at Higham Ferrers, Northamptonshire. The Roman road and the river Nene are visible in the background. Source: illustration from Lawrence and Smith (2009), reproduced courtesy of Oxford Archaeology.

those found in the other houses, but to one side there were a series of private apartments, including one at the focal point of the building, which is perhaps a reception chamber. This had a mosaic floor with central medallion showing Venus (Neal and Cosh 2010: 168–​ 171, no. 438.1). Although differentiated from the other houses and with an apparent function for gatherings, this building does not look like the main residence of a major landowner, although it could have been the base for his or her agent. A possible villa close by might perhaps represent a resident landowner. As with so many such sites, this leaves questions about its social and economic role completely open. The original suggestion that it might have been the centre for an imperial estate seems most improbable, especially now that such nucleated centres are evidently so very widespread in Roman Britain.

Towards a More Dynamic Understanding of the Countryside Traditional approaches to the countryside have very often been based on the study of particular sites, generally starting from a study of their structures, but now generally also involving the integration of evidence from artefacts, seeds, and animal bones (see

‘By Small Things Revealed’: Rural Settlement and Society    711 also Maltby, this volume, and Van der Veen, this volume). The best such studies provide excellent insights into these individual settlements, but they too often fail to place them into a broader provincial context. The real challenge for us today is to obtain a fuller understanding of the human societies who lived in Roman Britain. This requires the peopling of the whole landscape rather than just looking at particular structures, and seeing these people’s lives as dynamic, in terms of both day-​to-​day existence and development through time. Such an understanding can only ever derive from a combination of the analysis of detailed evidence from excavations with a broader reading of features within the landscape. An example of this type of approach is provided by our work in the lowlands on the edge of the Vale of York, in northern England (Halkon and Millett 1999; Millett 2006; Halkon et al. 2014). Here long-​term research has combined landscape study with selective excavation on a series of sites with the aim of exploring more fully the transition from Iron Age to Roman society. The study area is itself geographically varied, with the chalk uplands of the Yorkshire Wolds to the north-​east, its steep scarp slope dominating the lower ground to the south and west, cut into by a series of small valleys of streams that flow into the Humber. The foot of the Wolds is skirted by a bench of good arable land, which merges into a zone of wetland moors beside the Humber estuary. During the Roman period these lowlands were interspersed by the fingers of a now-​silted tidal inlet, the margins of which were dominated by sand ridges, a relic of the area’s glacial history. This topography has generated a highly varied environment that was differentially exploited by its population. Two key themes emerge from its study; first, the strength of the connections between the exploitation of uplands and lowlands, which were not separate, but formed an integral part of the same landscape, and, second, the varied use of the landscape for different types of agriculture and industrial production. Integration into the Roman world affected aspects of the lives of the inhabitants in complex ways. Iron Age settlement in the area had been dominated by a pattern of dispersed rural settlement, with most farms apparently occupied by single nuclear families. At Hayton, on the arable lands at the Wold foot, enclosed farmsteads with round-​houses were mostly focused on the stream valley, along which ran a major droveway that connected to the Wolds. The droveway was defined on either side by substantial earthen dykes against which many of the farmsteads were built. These had small fields or paddocks attached, on the side away from the dyke, and the excavated evidence shows the adoption of mixed arable farming with the exploitation of the usual range of domestic animals. Beyond the stream valley there were areas of larger fields to the south-​east and apparently a zone of open pasture to the north-​west. The droveway that followed the valley up onto the Wolds seems to have been used for the seasonal movement of flocks of sheep, which were kept in the valley over winter and for lambing in spring, but were moved to the uplands for summer pasture. The intensive exploitation of these flocks is witnessed by the widespread consumption of sheep-​meat in feasting, as shown by bone deposits within the excavated settlement sites. In the wetlands towards the Humber, there were similar farm sites, some of which were aggregated into slightly larger hamlets, but significantly these are generally without

712

712   Martin Millett evidence for associated field systems. There was large-​scale smelting of iron ore, which was found in the boggy ground locally. This industrial production required substantial supplies of wood as fuel, and it would appear that much of the wetland landscape was given over to managed woodland to provide this. The woods also provided grazing animals, probably cattle and pigs, creating a range of resources that complemented those of the mainly arable lands and could be exchanged with these nearby communities. Roman intervention in the landscape is marked by the arrival of the military sometime after ad 70, seen locally with the construction of an auxiliary fort at Hayton. The military brought new influences with them, as well as effecting a wholesale restructuring of power relationships. The latter can, to a limited extent, be seen in their preferential consumption of the prime lamb, produced by the local farms. New power relations are also suggested by the way that an Iron Age-​style settlement with an exceptionally large round-​house was soon constructed close up against the front of the fort, suggesting that a local was cosying-​up to the newly dominant authority. Elsewhere, the impact of Rome is less clearly visible, with only very gradual changes in lifestyles on local farms. We see the adoption of some aspects of a new material culture. The previously modest material culture repertoire, with just a limited range of ceramic cooking pots and beehive querns used within timber buildings, was gradually augmented, starting during the phase of military contact. Dress accessories, especially brooches, began to be used. There were also indications of changes in cooking and eating with the introduction of new types of quern, which imply a new way of making the daily bread, and a broadening of the variety of pot shapes and changes in the patterns of use-​deposits on jars, which imply alterations in ways of cooking. After the period of direct military intervention, which ended towards the end of the first century ad, more major changes were brought about by integration into the broader Roman world, which is represented by the construction through the area of a strategically important Roman road which linked Lincoln and York. The route crossed the Humber via a ferry at Brough, and followed the foot of the Wolds through our area, before continuing towards the frontier by way of the legionary base at York. This new road, which in part seems to have followed an existing network of trackways, completely changed the world of the local inhabitants, bringing a vast array of connections to an outside world. This road rapidly became the focus for new ways of life for some who migrated to live beside it, forming new roadside settlements at Hayton, and, within a generation or so, also at Shiptonthorpe nearby. Excavations at Shiptonthorpe give a very clear impression of the occupants’ transition. The earliest phases of the excavated household closely resemble an Iron Age-​style enclosed farmstead, with a variant on the round-​house built right next to the new road within a square enclosure: in other words, a traditional indigenous settlement relocated into an imperial context. This also applies to the modest character of the occupants’ uses of material culture, which remained in an Iron Age-​ style for at least a generation or two. It seems likely that most of the people moving to live by the road at both Shiptonthorpe and Hayton were coming from nearby, and there is a hint in the survey evidence of the abandonment of certain farmsteads, which may

‘By Small Things Revealed’: Rural Settlement and Society    713 reflect this move. However, there are also indications in the ceramics that some might have moved from further afield, from what is now Lincolnshire, perhaps taking advantage of economic opportunities created by the new road system. There is a suggestion of a slightly more rapid change in lifestyles at Hayton, in survey evidence both from the roadside settlement and on the excavated farmstead. Here the round-​houses and enclosures were transformed during the second century with the construction of two new-​style rectangular timber buildings, a small courtyard house, and an aisled hall within a new enclosure. This transformation, which respected the topography of the earlier houses, was accompanied by a broader change in the use of material culture. This included a diversification both in the types of object represented, to include a range of tools, architectural fittings and furniture, as well as in the range of pots used. The latter indicates a wider change in domestic habits, of storage and cooking as well as eating, which shows that the inhabitants had by now adopted a lifestyle typical of many in Roman Britain. The one category of material that remains conspicuously absent is that associated with personal grooming. A comparable overall pattern is repeated at Shiptonthorpe, but a little later in the middle of the third century, illustrating how different families adopted new life-​ways on their own terms and for their own means. As at Hayton, this change in material use was accompanied by a new domestic context, here with the construction of a huge aisled hall, the size of which is surely a strong statement of the social status of its owners (Figure 34.6).

Figure  34.6  Cut-​away reconstruction drawing of the third-​century aisled hall excavated at Shiptonthorpe, East Yorkshire. Source: drawing by Mark Faulkner, from Millett (2006).

714

714   Martin Millett This highlights one of the more striking features revealed by work in this landscape, for, while there is very strong evidence for the integration into the new life-​ways of the peoples who lived close to the Roman road, the same is not true for some of those living further from it. While sites along the road have a profusion of ceramics, coinage, and other metal artefacts (as revealed in particular by finds reported to the Portable Antiquities Scheme), most who lived away from the road seem not to have used such artefacts. Although there are exceptions, the overwhelming pattern is of a continued modest use of material culture. This is particularly surprising in the settlements in the wetlands in the southern part of our study area, as here the Iron Age pattern of large-​ scale iron production seems to have been superseded by pottery manufacture as iron ore resources became depleted. There is thus only the slightest of evidence for Roman-​ period iron-​working, but, by the third century, a number of sites were producing household pottery in kilns that lay on the margins of their settlements. The wide distribution of this pottery illustrates how successful these people were in negotiating new economic opportunities as it spread widely across what is now East Yorkshire and continued to be produced until the later fourth century. It is thus unclear why these producers did not also take up the use of the new material culture. It may be the result of rejection and the maintenance of other value systems, or it might be because the producers were in some way subservient, so that the profits from their pot sales did not come into the production zone but instead stayed elsewhere. There are also hints of a more nuanced pattern of rejection of Romano-​British material culture elsewhere in the area, with some sites quite close to the Hayton road apparently using imported pottery, but not receiving coins or other metal artefacts in any quantity. This reinforces the idea that the people of the region were selective in their adoption of material culture and there were significant variations in their attitudes to different elements of what is too often seen as a single cultural package. On the excavated farm at Hayton personal grooming seems eventually to have come into fashion in the later part of the third century when the enclosure system was remodelled to allow a small bathhouse to be added at one corner of the courtyard building. This was a tiny structure, with rooms only 2 metres × 2 metres, but includes the full basic suite for Roman-​style bathing and used the best of materials, including window glass and painted wall-​plaster. The adoption of both a new architecture and the habit of bathing suggests a very significant change in the daily lives of at least some of the inhabitants. Finally, it was during the middle of the fourth century that the well-​established timber buildings were replaced by a rather rambling range of stone structures, which brought the new architecture fully into the domestic domain. Too little survives of these buildings to understand details of their form, but the scraps of wall-​plaster and decoration indicate that some of the elements traditionally associated with villas were being adopted here. Most surprisingly, a fragment of wood recovered from a well backfilled when the site went out of occupation shows that the building’s furnishing included a cupboard inlaid with bone of a type known from the Mediterranean. So there is evidence for the ways in which the people in this area become more entwined with material and activities associated with the broader empire. But they also

‘By Small Things Revealed’: Rural Settlement and Society    715 continued with a series of their own traditional practices, which were evidently key to their cultural identity. Most obvious of these are certain patterns of burial practice that can be shown to have originated in the Iron Age. In particular, the inhabitants had particular rituals associated with the death of infants at around the time of birth—​a common occurrence in pre-​industrial societies. Those infants who died at this time were given careful burial within a domestic setting (a context from which older infants, children, and adults seem to have been systematically excluded). Some of them were buried around the walls of the houses—​and at Shiptonthorpe this seems to have been particularly associated with the eastern end of the aisled hall. The others were interred at boundaries in the general area of the house. There is also an apparent association between these neonatal burials and a series of feasting deposits, which were found across the site, possibly representing funerary ceremonies. The animals represented in the feasting deposits found at the farm at Hayton were mostly sheep, which is rather unusual in comparison with the evidence for other sites, including Shiptonthorpe. This brings us back to a consideration of how the people living here interacted with society more broadly. In terms of the farming economy, there is very little evidence for substantial change from the Iron Age into the Roman period, so the role of sheep in feasting may be a function of the close connection between the community in this valley and their flocks, which were regularly moving seasonally between lowland streamside pasture and the uplands (Figure 34.7). Other aspects of movement between the different local environments are illustrated by the exploitation of heather healthlands—​presumably nearer the Humber—​perhaps for roofing or for animal bedding, as reflected in the environmental samples from excavated sites. We may also speculate that this link also brought pots manufactured in that area together with woodland products like charcoal to Hayton. Such seasonally influenced movements need to be set in the context of the connectivity created by the Roman road, for both Hayton and Shiptonthorpe were situated at junctions between this and ancient routes from the lowlands up to Wolds. Given that there was demand for animals to supply meat to both the military and those living in the towns, we can envisage that stock would have been gathered at these points before being driven along the road to centres of consumption. Moving animals via the roads would have required frequent watering places as well as regular stops for feeding. Clear evidence of this is provided by a waterhole right beside the road in the excavated area at Shiptonthorpe, and we may envisage such features placed regularly along the route. Less direct evidence is provided by the environmental evidence for the production of hay noted at the Hayton farmstead, which derived from the exploitation of the meadows beside the stream for growing animal fodder, perhaps to provide feed for animals passing along the road. It certainly implies a more intensive stock regime. Linked with the idea of the roadside settlements becoming funnels for supplies that were henceforth moved along the road, we have two different sets of evidence. At Shiptonthorpe there is an unusual focus of corn-​grinding querns, which implies that flour was being milled on a scale beyond the usual domestic level. As this area is not one with much evidence for large-​scale intensive cereal farming, we might tentatively

716

716   Martin Millett

Figure 34.7  Aerial photograph of part of the Roman landscape at Burnby Lane, Hayton, East Yorkshire, looking north-​east. The valley of the Burnby Beck runs down the centre of the photograph, with its relict course showing as a light crop-​mark against the darker green of the flood plain. On either side of this damp ground can be seen a series of settlement enclosures. On the left a length of boundary ditch is also visible. This runs parallel with the stream, and formed one side of a droveway that linked the lowlands to the south-​west with the Wolds to the north-​east, allowing animals to be moved for seasonal grazing. Source: photograph by Peter Halkon, courtesy of the Hayton Project.

suggest that it acted as a gathering point between smaller-​scale producers and consumers without access to their own grain supplies (perhaps from the woodland to the south), so that the small-​scale consumer could trade and take away milled flour. At Hayton, beside the Roman road in an area of probable pastureland that is devoid of settlement features, there is a substantial concentration of metal-​detected finds, which has been interpreted as evidence for the site of a season market or fair, where stock brought down the valley might have been sold. The composition of the assemblage of metal finds includes many objects associated with horse harnesses and carts, as well as a number associated with the military, among which is a lead seal of Legio VI Victrix, which was based at York from the time of Hadrian. These finds indicate that there was probably a lot more to the fair than just the selling of sheep, and we may well envisage a more substantial and varied event, which perhaps involved military buyers. Such a context will have widened the network of contacts of those living in the valley and provided them with the opportunities to be active participants in a broader cultural

‘By Small Things Revealed’: Rural Settlement and Society    717 milieu. As such it provides an unusual insight into the sorts of social contact that must have been at the centre of the lives of most rural inhabitants of Roman Britain.

Conclusions It is clear that the study of rural settlement in Roman Britain is at a key turning point. The proliferation of new evidence provided both by developer-​funded work and long-​ term field research has provided a catalyst for new thinking about the subject which has at long last moved the focus away from villas. It is equally important that emphasis is now being placed not on individual sites but on the study of the overall landscape, not simply as a background canvas, but as a highly variegated framework within which different people lived their own lives. Finally, as Jeremy Taylor (2013) has shown, the rural population must be recognized as the active creators of their own culture, not simply as passive recipients of that imported from elsewhere in the empire or, more locally, from the towns of the province.

Abbreviation RIB  R  . G. Collingwood and R. P. Wright, The Roman Inscriptions of Britain. Volume I: Inscriptions on Stone. Oxford: Clarendon Press, 1965–​to date.

References Primary Sources Theodosian Code. The Theodosian Code and Novels and the Sirmondian Constitutions. A Translation with Commentary, ed. C. Pharr. Princeton: Princeton University Press, 1952.

Secondary Sources Black, E. W. (1987). The Roman Villas of South-​ East England. BAR British Series 171. Oxford: Archaeopress. Cunliffe, B. W. (1971). Excavations at Fishbourne. Leeds: Research Reports of the Society of Antiquaries of London. Cunliffe, B. W. (2008). The Danebury Environs Roman Programme: A Wessex Landscape during the Roman Era. Volume 1: Overview. Oxford University School of Archaeology Monograph 70. Oxford: Oxford University School of Archaeology. Cunliffe, B., Down, A., and Rudkin, D. (1996). Chichester Excavations IX: Excavations at Fishbourne 1969-​88. Chichester: Chichester District Council. Cunliffe, B. W., and Poole, C. (2008). The Danebury Environs Roman Programme: A Wessex Landscape during the Roman Era. Volume 2 Part 5: Rowbury Farm, Wherwell, Hants. 2003.

718

718   Martin Millett Oxford University School of Archaeology Monograph 71. Oxford: Oxford University School of Archaeology. Cotswold Archaeology 2014 ‘Realising the research potential of developer-​funded Roman Archaeology in England’ http://​www.cotswoldarchaeology.co.uk/​developer-​funded-​ roman-​archaeology-​in-​england/​. Accessed 27/​7/​14. Gerrard, J. (2013). The Ruin of Roman Britain:  An Archaeological Perspective. Cambridge: Cambridge University Press. Gracie, H. S. (1970). ‘Frocester Court Roman Villa, Gloucestershire. First Report’, Transactions of the Bristol and Gloucestershire Archaeological Society, 89: 15–​86. Gracie, H. S., and Price, E. (1979). ‘Frocester Court Roman Villa, Gloucestershire. Second Report’, Transactions of the Bristol and Gloucestershire Archaeological Society, 97: 9–​64. Halkon, P., and Millett, M. (1999) (eds). Rural Settlement and Industry:  Studies in the Iron Age and Roman Archaeology of Lowland East Yorkshire. Yorkshire Archaeological Report 4. Leeds: Yorkshire Archaeology. Halkon, P., Millett, M., and Woodhouse, H. (2014) (eds). Hayton, East Yorkshire: Archaeological Studies of the Iron Age and Roman Landscapes. Leeds: Yorkshire Archaeology. Landscape Research Centre: Digital Atlas. http://​thelrc.wordpress.com/​lrc-​digital-​atlas/ (accessed 27 July 2014). Lawrence, S., and Smith, A. (2009). Between Villa and Town: Excavations of a Roman Roadside Settlement and Shrine at Higham Ferrers, Northants. Oxford Archaeology Monograph 7. Oxford: Oxford Archaeology. McCarthy, M. (2013). The Roman Peasant: Towards a Study of People, Landscape and Work during the Roman Occupation of Britain. Macclesfield: Windgather Press. Manley, J., and Rudkin, D. (2003). Facing the Palace: Excavations in Front of the Roman Palace at Fishbourne, Sussex, UK 1995–​99. Lewes: Sussex Archaeological Society. Mattingly, D. (2006). An Imperial Possession: Britain in the Roman Empire. London: Allen Lane. Millett, M. (1990). The Romanization of Britain:  An Essay in Archaeological Interpretation. Cambridge: Cambridge University Press. Millett, M. (2006) (ed.). Shiptonthorpe, East Yorkshire: Archaeological Studies of a Romano-​ British Roadside Settlement. Yorkshire Archaeological Report 5.  Leeds:  Yorkshire Archaeology. Neal, D. S., and Cosh, S. R. (2010). Roman Mosaics of Britain. Volume 4: Western Britain. London: Society of Antiquaries. Price, E. (2000–​10). Frocester: A Romano-​British Settlement, its Antecedents and Successors. Volumes 1–​4. Stonehouse: Gloucestershire and District Archaeological Research Group. Rippengal, R. (1993). ‘Villas as a Key to Social Structure? Some Comments on Recent Approaches to the Romano-​British Villa and Some Suggestions towards an Alternative’, in E. Scott (ed.), Theoretical Roman Archaeology:  First Conference Proceedings. Aldershot: Avebury, 79–​102. Roberts, I. (2010). Understanding the Cropmark Landscapes of the Magnesian Limestone. Leeds: West Yorkshire Archaeological Services. Scott, S. (2000). Art and Society in Fourth-​Century Britain: Villas in Context. Oxford University School of Archaeology Monograph 53. Oxford: Oxford University. Smith, J. T. (1997). Roman Villas: A Study in Social Structure. London: Routledge. Stevens, C. E. (1966). ‘Social and Economic Aspects of Rural Settlement’, in A. C. Thomas (ed.), Rural Settlement in Roman Britain. London: Council for British Archaeology, 108–​128.

‘By Small Things Revealed’: Rural Settlement and Society    719 Taylor, J. (2007). An Atlas of Roman Rural Settlement in England. CBA Research Report 151. York: Council for British Archaeology. Taylor, J. (2011). ‘The Idea of the Villa:  Reassessing Villa Development in South-​ East Britain’, in N. Roymans and T. Derks (eds), Villa Landscapes in the Roman North. Amsterdam: Amsterdam University Press, 179–​194. Taylor, J. (2013). ‘Encountering Romanitas:  Characterising the Role of Agricultural Communities in Roman Britain’, Britannia, 44: 171–​190. Timby, J. R. (1998). Excavations at Kingscote and Wycomb, Gloucestershire. Cirencester: Cotswold Archaeological Trust. Tomlin, R. S. O. (1996). ‘A Five Acre Wood in Roman Kent’, in J. Bird, M. Hassall, and H. Sheldon (eds), Interpreting Roman London: Papers in Memory of Hugh Chapman. Oxford: Oxbow Books, 209–​216. Tomlin, R. S.  O., and Hassall, M. W.  C. (2004). ‘Roman Britain in 2003 III. Inscriptions’, Britannia, 35: 335–​349. Turner, E. G. (1956). ‘A Writing Tablet from Somerset’, Journal of Roman Studies, 46: 115–​118.

720

Chapter 35

Rur al Transf ormat i on i n the Urbanized L a nd s c a pe Martin Pitts

Introduction: Questions and Models Considerable attention has been devoted to questions of rural change in Roman Britain. To date, this has tended to crystallize around two interrelated issues: the economic relationship between town and country, and the extent to which ‘Roman culture’ was adopted in rural areas. Particularly influential on both counts is the following passage by R. G. Collingwood:  Economically, the towns were parasitic on the country-​side. They had to be fed by it, and the goods they produced, together with the services they rendered as markets and trading-​centres, were no adequate return for the food they consumed and the expenditure which they demanded for the upkeep of their public services. They had their industries; but these consisted only to a small extent in the production of goods needed in the country; most of them were luxury-​trades whose produce was mostly used in the towns themselves. They did a large business in retail trade, selling pottery made in Gaul and other imports, but here again, the total quantity of these goods which found its way into the country districts was the barest fraction of what the towns consumed. From a strictly economic point of view the towns were a luxury. Their function was cultural and political. They stood for the decencies and elegances of civilized life, and they provided a link between the Roman government and the mass of the people, to whom those decencies and elegances were things out of reach. Their populations, rich and poor alike, thus formed a privileged section of the people, privileged to enjoy the blessings of romanization at the expense of the country-​ folk. (Collingwood and Myres 1937: 198–​199)

Seventy-​five years on, Collingwood’s model has endured remarkably well. One reason for this is the lack of data available to previous generations properly to test

Rural Transformation in the Urbanized Landscape    721 Collingwood’s statement (Millett 1982). Nonetheless, as discussed below, the most recent data fit Collingwood’s model rather well. While there have been critics over the years, notably Wacher’s restatement of the positive contribution of Romano-​British towns in the development of rural settlement and agricultural improvements (Wacher 1995: 70), the core idea of cities surviving as rural parasites continues to strike a chord with the extant evidence. The basic logic of the parasitical model resonates with attempts to explain Romano-​British rural change in response to integration into imperial systems and the transformative pressures of taxation (Hingley 1982; Millett 1990a), as well as post-​colonial emphasis on the exploitative dimensions of Roman imperialism, culminating in sharp divisions in culture and identity experienced between urban and rural communities (Mattingly 2006). Despite an emerging degree of consensus on Collingwood’s vision, the general issue of the character and extent of rural transformation in the urbanized landscapes of central, southern, and south-​east Britain is far from settled. A shifting research agenda in recent decades has led to a change from studies prioritizing the economy (for example, villa development and pottery marketing) to those favouring cultural change (for example, the social use of space and consumption), resulting in necessarily partial explanations of Rome’s impact. Given this state of play, the most compelling general models of rural change that integrate both economic and cultural processes remain those that were first proposed in the 1980s and early 1990s. Notable examples include Haselgrove’s account (1982) of political centralization in late Iron Age south-​east Britain, Hingley’s structural analysis (1982) of the impacts of Roman imperialism, and Millett’s interpretative essay (1990a) on the Romanization of Britain. These are typical examples of processual grand narrative, drawing upon Immanuel Wallerstein’s ‘World-​Systems Analysis’ (1974), which invokes the concepts of ‘centre’ and ‘periphery’ as analytical categories. While the most explicit centre–​periphery models have fallen out of regular usage—​ owing to a combination of significant elements being refuted by further archaeological investigation and a shift towards post-​processual studies eschewing macroeconomic models in favour of local and symbolic perspectives—​such models deserve fresh examination in the light of new data being unearthed. The most relevant processual models for understanding rural change in Roman Britain are those of Hingley (1982) and Millett (1990a). Both are effective in linking the global pressure of Rome’s imperial regime through taxation to local developments at the tribal or civitas level that dictated the subsequent character of urban–​rural relationships. For Hingley, Rome’s mission as a colonizing power was to exploit its subject territory, to facilitate a flow of surplus goods and produce to be directed to the ‘core’ of the empire. To achieve this, particular forms of social organization and economic relationships were fostered by Rome, involving recurring relations of dominance and dependency at multiple scales—​between the provinces and the ‘core’ at the inter-​ provincial scale and between town and country at the tribal level. Hingley’s model envisaged the empire in terms of a pyramid structure, with relations between the imperial core at the pyramid’s apex and rural locations at the base, being articulated through tribal centres (for example, Verulamium), which were, in turn, subordinate

722

722   Martin Pitts to provincial centres (for example, London). Flows of basic agricultural produce, raw materials, and slaves moved up the pyramid in response to the stimulus of taxation, with quantities being retained or converted to cash at the intermediate levels for local consumption, and a reciprocal downward flow of luxury and manufactured goods from the apex to the base. Crucially, the reciprocal exchange of manufactured goods for agricultural raw materials was unequal, with the latter produced in the periphery requiring greater expenditure of energy (largely in the form of labour) than the exchange commodities produced in the core. In short, this is a more formalized and systemic realization of Collingwood’s statement of major Romano-​British towns (that is, the civitas/​tribal capitals) functioning as parasites. Major towns were required adequately to facilitate the exploitation of the rural populace, serving as nodal points for the conversion of agricultural surplus to tax and the wider dissemination of imported luxury and mass-​produced items. In contrast to Hingley, Millett (1990a, b) highlighted Rome’s laissez-​faire attitude to provincial administration, emphasizing the role of pre-​conquest elites as the primary agents of change. Providing that sufficient centralized social organization existed prior to Roman annexation, friendly members of the pre-​existing tribal aristocracy could retain their local power and govern on Rome’s behalf in return for the adoption of a Roman constitution and the guarantee of tax revenues, which were payable by civitas (Millett 1990a: 125). As the power of the local elite became increasingly dependent on Roman patronage, the desire to use and appropriate Roman symbols became more prevalent throughout society, eventually becoming status indicators that were to be progressively emulated by the lower social orders (Millett 1990b). Millett (1990a: 117–​ 123) argued that the ‘trickle-​down’ of such Roman aspiration was evident in rural areas through the flourishing of villas in the second and third centuries, in addition to significant distributions of imported samian pottery outside the urban sphere (see also Evans, this volume). Despite the reduced emphasis placed on imperial exploitation in Millett’s model, it does not contradict Hingley (1982). Both accounts recognize the role of taxation in structuring relations between town and country and both suggest that the adoption of Roman goods and symbols in the countryside was an inevitable and desirable outcome. Whereas the early formulations of Hingley and Millett are broadly consistent with one another, they appear to diverge from Collingwood on the rural penetration of Roman material culture, who was at pains to stress the failure of Romanization in the countryside (either material or ideological), a view that has been more recently expressed by David Mattingly (2006) and others. Clearly, questions of rural change in Roman Britain cut across several issues of wider importance and controversy—​the extent to which the colonial regime was exploitative or facilitating; whether change in material culture was a by-​product of economic process or cultural participation; the degree to which non-​elite communities had agency and choice; and the precise mechanisms by which rural communities were socially and economically connected to the larger whole of the Roman Empire. Addressing these concerns requires not only detailed empirical exploration of the available data (of which a burgeoning corpus has been amassed since the debates of the 1980s and early 1990s),

Rural Transformation in the Urbanized Landscape    723 but also an integrated approach that considers such evidence in the light of both economic and cultural terms (cf. Fulford 1982; Millett 1982).

Models and Data There currently exists a large and diverse array of archaeological data for the evaluation of theories of rural transformation in Roman Britain. However, this is a comparatively recent development in the history of the subject. In the mid–​late twentieth century, for example, attempts to understand rural change were predominantly focused on charting the origins and development of ‘villa’ buildings in the countryside, which came to be seen as evidence for the Romanization of the indigenous population (e.g. Rivet 1964: 99–​130; Richmond 1969). Rivet’s statement that ‘we cannot have a villa without a town’ (Rivet 1964: 105) was, in part, reified by the emergent pattern of the earliest villas developing in close proximity to major towns in the south and south-​east. The excavations at early villas such as Park Street (near Verulamium, modern St Albans) and Lockleys (Welwyn, both in Hertfordshire and in the civitas of the Catuvellauni) became type-​ sites to illustrate the Romanization of the ‘native’ elite. This process was characterized at both sites by the direct replacement of Iron Age round-house dwellings by rectilinear ‘Romanized’ house forms within a generation of the Claudian conquest. Such explanations remain broadly current, with Millett (1990a: 94) regarding the phenomenon as an outgrowth of the desire by their local owners to appear Romanized rather than an indicator of increased prosperity and living standards under the new regime (contra Frere 1987: 258). The privileged position of villas in accounts of rural change was to change irrevocably in the 1970s and 1980s. In the first place, these decades witnessed an upsurge in the numbers of non-​villa rural settlements excavated, reflecting the historical reality that villas would have been vastly outnumbered by their non-​villa counterparts (Hingley 1989: 4–​ 5). The persistence of rural settlements lacking overtly ‘Romanized’ architectural features highlighted a gap in traditional narratives as well as in explanations of cultural change amongst non-​elite communities. In addition to a more balanced coverage of settlement types, advances in the recording and quantification of small-​finds assemblages (artefacts and ecofacts) provided glimpses of a more holistic picture of urban–​rural relationships in Roman Britain, in terms of patterns of diet and ‘marketing’ distributions. For example, King (1978, 1984) showed entrenched differences in the proportions of major domesticates in the animal bone assemblages of so-​called Romanized settlements (towns and villas, favouring cattle) and ‘un-​Romanized’ settlements (low-​status farmsteads, favouring sheep and goat) (for further discussion of animal husbandry, see Maltby, this volume). King attributed these changes to Romanization as a form of cultural preference, with the influence of garrison units previously stationed in Gaul and Germany providing the basis for local emulation. In terms of the pottery, there were still insufficient data properly to evaluate potential differences in urban and rural supply at

724

724   Martin Pitts this time (Millett 1982). Nevertheless, Swan’s corpus (1984) of Romano-​British kiln sites, in addition to distribution studies of later Roman products such as Oxfordshire (Fulford and Hodder 1975) and Farnham wares (Millett 1979), suggested consistent patterning, notably a shift away from the urban marketing of ceramics in the early Roman period to the development of late Roman rural-​based industries that took advantage of river transport to tap into markets further afield (for further discussion of the pottery evidence, see Evans, this volume). Building on the developments of the previous two decades, scholarship from the 1990s onwards shows a greater level of sophistication and specialization. On a theoretical level, the influence of postcolonial literature (e.g. Said 1993) and its applications in Roman archaeology (e.g. Webster and Cooper 1996; Mattingly 1997) prompted a rethinking of historical models of Romanization as a top-​down civilizing process, with debates over the contextual meaning of material culture taking increased prominence (e.g. Webster 1997). Whereas the issue of rural change remains firmly on the agenda, with increased emphasis on ‘bottom-​up’ perspectives and non-​elite communities (e.g. James 2001), questions of urban–​rural interdependence, economic growth, and marketing have been arguably sidelined in favour of the dominant agenda of cultural concerns—​for example, Romanization, creolization (Webster 2001) and identity (Mattingly 2004). At the same time, the ever-​increasing corpus of assemblage data from individual excavations demanded quantitative study and the application of new methodologies for appropriate analysis and synthesis (e.g. Evans 1995, 2001; Cool and Baxter 1999, 2002; Lockyear 2000). As a consequence, while archaeological data from Roman Britain are routinely examined in order to shed light on cultural difference, the economic underpinning of such patterns is seldom prioritized. The most sophisticated models that address economic concerns remain, for the most part, those formulated in the early 1990s (e.g. Millett 1990a)—​ironically an era in which barely sufficient data were available to test the models being proposed. Despite the relative lack of attention to the integration of findings and model testing, the patterns emerging from more recent synthetic studies of Romano-​British finds assemblages are remarkably consistent. While the scope for significant variation within settlements may be accepted (e.g. Eckardt 2007; Willis 2007), most striking is the differentiation of major urban settlements (and military sites) from lesser nucleated (for example, small towns and villages) and rural settlement classes—​the latter tending to be less distinctive as a discrete category and subject to greater variation depending on regionality and/​or the particular classes of evidence under scrutiny. A brief summary of the basic patterns is as follows. In terms of coin loss, the major towns show a consistently higher rate than all other classes of settlement, apart from sites associated with the army (Reece 1993). Faunal assemblages from major urban centres and military sites tend to have higher proportions of pig and cattle remains than other settlement types, with progressively lower percentages in assemblages in the order of secondary urban sites, villas, and rural sites (King 1999, 2001). Analysis of butchery practice shows a high frequency of cut-​marks indicative of specialist animal carcass processing in towns, with a comparative paucity of equivalent marks in rural assemblages (Maltby 1989, 1994,

Rural Transformation in the Urbanized Landscape    725 2007). The synthesis of available data on plant remains shows military communities and major urban centres (especially London) to have been the biggest consumers of exotic and imported plants, with access to such foodstuffs being more limited on rural sites in the south-​east, and virtually non-​existent for rural sites in the rest of the province (van der Veen et al. 2008). A similar pattern emerges for the spread of lighting equipment, with little uptake of lamps outside military communities and the major civilian centres of London and Colchester (Eckardt 2002). Finally, the increasingly wide-​ranging and statistically robust analyses of pottery consumption in Roman Britain highlight the supply of both imported wares (such as samian and amphorae) and specialist consumption technology (drinking and dining forms as opposed to basic cooking wares) to be most prevalent in assemblages from major towns (and military sites), with a progressive fall-​off in frequency further down the settlement hierarchy (e.g. Willis 1996, 2011; Evans 2001, 2005; Pitts 2008).

Urban Hinterlands—​Beyond Resistance and ‘Romanization’ The overarching picture that emerges from recent surveys of consumption in Roman Britain (as briefly summarized above) is, arguably, one involving the maintenance of a wide gap between urban and non-​urban populations. It should not be underestimated that the principal evidence for such a gap consists of differences in the supply of foodstuffs and everyday manufactured objects—​such patterns are most likely to have been experienced in the guise of entrenched differences in social practice and culture between urban and rural communities. This observation is acutely apparent in recent attempts to investigate urban hinterlands in Roman Britain. Gaffney and White’s survey-​driven study of Wroxeter’s hinterland revealed negligible material change in the countryside in terms of the spread of urban forms of consumption: ‘the town appears rich, dynamic, and “Romanized”; the countryside appears impoverished, static and conservative’ (Gaffney and White 2007: 282–​283). While the scale and character of such results are likely to be, in part, attributable to the social organization (or lack thereof) of the Cornovii prior to conquest, described as ‘decentralized and egalitarian’ by Millett (1990a: 100) (a classification disputed by Gaffney and White), it is, nonetheless, arguable that comparable patterns emerge in areas with centralized and hierarchical pre-​conquest societies, once such pre-​existing differences are accounted for. In contrast to the Wroxeter hinterlands project, the study of the East Anglian hinterlands of Roman London and Colchester (Perring and Pitts 2013) privileged the analysis of consumption patterns from excavated finds assemblages, taking particular advantage of the region’s numerous large and well-​dated published pottery groups. The results of this project revealed that, although there was a strong uptake of new forms of material culture in the Essex countryside (relative to regions outside the south-​east),

726

726   Martin Pitts in quantitative terms the urban–​rural difference remained striking. Correspondence analysis of nearly 100 stratified pottery assemblages (quantified by estimated vessel equivalent) from London, Essex, and Cambridgeshire shows very little change in the pronounced separation of patterns of urban and rural pottery use over c. 200 years, despite a moderate uptake of supposedly ‘Romanized’ vessel forms throughout the region and period (Figure 35.1 and Table 35.1). The earliest assemblages from settlements with pre-​conquest origins are distinguished by the presence of imported Gallo-​Belgic fine wares and older Gallic-​influenced pedestalled vessels (for example, Heybridge and Kelvedon, upper-​left quadrant of Figure 35.1), with the tendency for large drinking vessels in such ceramic repertoires fitting with the regional predilection for feasting activities (Pitts 2005, 2010). However, such colourful practices appeared to die out within a generation of the Claudian conquest. From the Flavian period onward, communities living outside the major urban centres appear to have been content to use a seemingly watered-​down, locally produced version of what was being used by city-​dwellers—​far fewer imports, fewer vessels associated with the presentation and serving of food and drink, and, compared to the earlier first century, a relative shift in emphasis from drinking vessels to dining vessels within the subset of pottery used for consumption. In terms of pottery usage, Figure 35.1 shows no significant or consistent differences between assemblages from minor urban settlements, villas, and low-​status rural sites. Such relative homogeneity outside the ‘true’ urban sphere echoes findings in other classes of evidence, such as plant remains, whereby lesser urban centres showed greater affinity with rural sites and little differentiation between villas and lower-​status sites (van der Veen et al. 2008). Further indication of the character of impact of urbanism on the wider settlement landscape can be obtained by considering particular types of material culture that can act as proxies for other more illuminative materials (for example, foodstuffs). One particularly striking pattern from the Essex urban hinterlands project (Perring and Pitts 2013) was the relatively high levels of lids in pottery assemblages from London and Colchester (plotted in Figure 35.1 alongside other diagnostic urban staples such as flagons and mortaria), a trend that is apparent throughout the early Roman period. Further investigation showed lids made up around 4–​10 per cent of the typical ‘true’ urban assemblage, with less than 4 per cent on most non-​urban sites (Figure 35.2). What makes this a seemingly odd finding is that the vast majority of lids were made of local sandy grey coarse wares, probably at multiple kiln sites throughout the region (for example, production of this fabric is attested at Ardleigh, near Colchester and Heybridge, somewhat further afield) (Going and Belton 1999; Biddulph et al. 2015). Although such grey wares invariably form a substantial component of most coarse ware assemblages regardless of settlement-​type in the region during the early Roman period, the most common products in the fabric (various cooking jars) are more synonymous with rural and minor urban settlements than major centres such as London and Colchester. Such patterns, at the least, suggest a particular preference for lids among urban communities, perhaps for cultural reasons linked to the preparation of food. However, if this was the case, the scarcity of imported lids

1.0

0.5

WDW2-4 WIT3-4 STN5 KEL1-3 Major urban WAM3-5 BOR7-8 STN3-5 STN3 RHM4-5 KEL5-6 Minor urban HEY2-3 STN6-7 STN7-8 High-status rural CHM-TP4-6 CAN3-6 HEY3-6 CHM5 Low-status rural BRT5-6 CHM-TP8-9 WAM10-11 FDT3-5 BRT5-9 KEL3-4 ARD3-5 RET6-7 NSH8-10 ARD8 HEY3 WDW3-4 BRT8-9 HEY8-10 HEY3-4 CHM-MA3-6 STH8-9 BRT10-11 KEL3-5 BOR10 STN3-7 HEY7-8 CHM-MA8 CAN3-7 CHM-TP9-10 CHG3-6 RET8-9 CHM-MA6 CHM6-7 CHM-MA9-10 WIT8-10 CHM-MA10-11 ARD6-7 HEY4-6 LOK11 HEY10-11 BOR9-10 ARD4-6 HEY9-10 KEL4-6 CHM8-10 RET10-11

0.0

COL8 COL8-10 LON-E6-8

COL4

–0.5

LON-E8-9

COL4-6 COL9 COL6-8 COL6-7 COL6-9 COL4-5 COL7 COL5-6 LON-W5-8 LON-W5-6 LON-E5-7 COL6 LON-E5-6 LON-W5 LON-W6-8 LON-W5-7 LON-W7-9 LON-W3-4

LON-E7-8

LON-W6-7

LON-E9

LON-W9-10

LON-W7-10

–0.5

COL9-10

0.0

0.5

1.0

1.5

Pedestal jars

1.5

GB beakers Pedestal tazze GB jars 1.0 GB platters

Non-urban and minor urban

Platters

0.5

Jars

GB flagons 0.0

–0.5

–1.0

Miniatures Storage jars

Cups

Dishes

Beakers

IMP platters IMP mortaria

Major urban Bowls

SMN platters

IMP flagonsSMN cups IMP cups Amphorae IMP lids IMP bowls

Lids

SMN dishes

IMP beakers

SMN bowls Flagons Mortaria IMP dishes IMP jars

–0.5

0.0

0.5

1.0

1.5

Figure 35.1  Correspondence analysis highlighting differences in the composition of pottery assemblages (top) by vessel form (bottom) in the hinterland of Roman London and Colchester, c. 50 bc–​ad 250. Note: Details of abbreviations are outlined in Table 35.1. © Martin Pitts.

728

728   Martin Pitts Table 35.1 Codes used in Figures 35.1 and 35.2 Code

Site

Ceramic phase

Date range

ARD

Ardleigh

1

50–​15 bc

BOR

Great Holts Farm, Boreham

2

15 bc–​ad 20

BRT

Braintree

3

ad 20–​55

CAN

Cambridge, New Addenbrookes

4

ad 55–​60

CHG

Chignall

5

ad 60–​75/​80

CHM

Chelmsford, street frontage

6

ad 75/​80–​100

CHM-​MA

Chelmsford, Mansio

7

ad 100–​125

CHM-​TP

Chelmsford, temple

8

ad 125–​140

COL

Colchester

9

ad 140–​160/​70

FDT

Fen Ditton

10

ad 160/​70–​210

HEY

Elms Farm, Heybridge

11

ad 210–​250

KEL

Kelvedon

Fabric code

Fabric

LOK

Little Oakley

SMN

Samian ware

LON-​E

London, East

IMP

Misc. imports

LON-​W

London, West

GB

Gallo-​Belgic ware

NSH

North Shoebury

RET

Curry Hill, Rettendon

RHM

Moor Hall Farm, Rainham

STH

Strood Hall

STN

Stansted

WAM

Wendens Ambo

WDW

Woodham Walter

WIT

Ivy Chimneys, Witham

from the continent in early urban assemblages is unusual. An alternative explanation is that lids were required specifically to aid the transportation of foodstuffs or other products from the country to the city, facilitating urban consumption. This hypothesis accords with the suggestion that major towns would have formed the primary market for new varieties of fresh fruit, vegetables, and herbs that became widespread in the south-​east (van der Veen et al. 2008) (see also van der Veen, this volume).

Rural Transformation in the Urbanized Landscape    729 18

COL9

LON-W9-10

16 14

% lids

12 10 8 6 4

Major urban Minor urban High-status rural Low-status rural

LON-E6-8 COL6-9

LON-W7-9

LON-W6-7 LON-E9 COL7 LON-W6-8 COL4 LON-W5-8 HEY4-6 LON-E5-6 COL6-7 LON-E5-7 COL9-10 LON-E8-9 COL8 ARD6-7 LON-W5-7 COL4-6 LON-W5-6 LON-E7-8 COL6 LON-W7-10

2

COL6-8

COL5-6 COL4-5

HEY7-8 HEY3 FDT3-5 ARD3-5 CHM6-7 WDW3-4 HEY3-4 LOK11 KEL3-5 HEY2-4 BRT5-9 HEY10-11 HEY8-10 CHM8-10 CHM-TP9-10 CHM-MA10-11

LON-W5

20

30

ARD8

CHG3-6

ARD4-6

40

50

% jars

60

70

CHM-TP4-6 RHM4-5 STN5 BRT8-9 KELI-3 CAN3-6

80

90

Figure 35.2  Pottery assemblages from Roman London, Essex, and Cambridgeshire plotted according to the percentage prevalence of lids versus jars, c. 50 bc–​ad 250. Source: after Perring and Pitts (2013). © Martin Pitts.

Returning to the processual models of Hingley (1982) and Millett (1990a), the situation described for the south-​east fits well with the idea of rural produce (containers and/​or their contents) being sent to major towns in response to the stimulus of taxation—​either to be converted to cash in urban markets or to be taken directly in kind. Furthermore, although imports and more specialist mass-​produced goods available in urban markets clearly did make it into the countryside, the relative weakness of rural consumption in emulating urban consumption highlights the essential asymmetry and inequality of urban–​rural exchange as postulated by Hingley. Similarly, in terms of the consumption of regional produce, Colchester appeared much more dependent on the output of its immediate hinterland, whereas London received flows of goods from a wider variety of sources within the province, seemingly confirming the latter’s elevated position in Hingley’s pyramidal structure of imperial relationships. Although the gap between urban and non-​urban consumption was largely maintained into the third century, relations of urban–​rural interdependence were not static. Indeed, the available evidence suggests a tendency for increased integration in pottery supply across all settlement types by the late second century (Perring and Pitts 2013). While this period coincides with a wave of villa construction in the region, the villa assemblages themselves are not greatly differentiated from those of non-​villa rural settlements. It thus remains to be seen whether or not the appearance of villas represented the ‘trickle-​down’ of Roman culture into the countryside or simply the investment in rural property by absentee landlords who relied on local tenants to manage their holdings.

730

730   Martin Pitts Assessing the cultural implications of the various changes in rural production and consumption in East Anglia is not straightforward. The relatively weak rural consumption of imports must be set in the context of sweeping changes to coarse ware production, which, although heavily influenced by continental forms (via Gaul) prior to the conquest, took on a decidedly new ‘provincial’ character by the late first century ad. In the light of the infamous Boudiccan uprising of ad 60/61, it is unlikely that such changes in production reflected a widespread local desire to adopt the ways of the colonial settlers at Colchester. Instead, as the case of the coarse ware lids suggests, such changes probably reflected the realignment of production to suit the needs of the imposed urban communities—​implemented either via the remaining indigenous elites or by the next generation of veteran colonists. The impact on the countryside of fulfilling the basic calorific demands of the new urban communities should not be underestimated—​as well as being provided with the basic foodstuffs, the urban populations also required appropriate vessels in which to cook and serve food and drink in concordance with their own cultural needs and practices. Against the backdrop of a new impetus for rural production comes the question of the agency of communities living outside the major urban centres. Were they content to adopt the new types of pottery used by urban dwellers and carry on with their lives as before, or did they do so consciously to emulate the tribal aristocracy and to identify themselves with the new provincial order (as per Millett 1990b)? On this issue the models of Hingley (1982) and Millett (1990a, b) are less in tune with the available evidence. Here it is crucial to evaluate both the extent and meaning of the so-​called trickle-​ down of imported ceramics that enjoyed conspicuous display in the graves of local elites in the initial post-​conquest period, such as Folly Lane, St Albans (Niblett 1999), and Stanway, Colchester (Crummy et al. 2007). While it is true that imports such as samian ware clearly did exhibit widespread penetration outside the major towns, the quantities involved rarely rivalled those of urban assemblages. On this point, Millett (1990a: 123) expressed the view that: Larger-​scale distributions like that of samian ware were less heavily centred on the towns than might be expected. This indicates an ample sufficiency of supplies and so the demands of any consumer could be filled. Where these objects were not being received the deficiency was of demand, not supply.

In the light of data now available, it appears that the gap between urban and non-​ urban consumers was greater than Millett originally envisaged. Either the supply of imports was not sufficient to meet rural demand, the pots themselves were too expensive for a large proportion of rural consumers, or the demand for such goods in the countryside was simply not as strong as that of the major towns. These are not necessarily mutually exclusive possibilities. While samian was evidently imported in quantity to Britain, supplies were not limitless. As the premium ceramic table ware of the period, such imports would inevitably have carried higher prices than basic coarse wares. A full suite of samian vessels probably involved a substantial outlay of income for the average

Rural Transformation in the Urbanized Landscape    731 Romano-​British peasant and could hardly have been a priority for people living at, or close to, the subsistence level. Addressing the uses to which imported ceramics were put in the Romano-​British countryside is especially pertinent to understanding the desire for such commodities. In cultural terms, the growth in the consumption of Gallo-​Belgic wares (a repertoire featuring many forms that imitated samian originals) at a wide range of locations in Essex up to the time of the conquest demonstrates a pre-​existing taste for sophisticated ceramic forms beyond the circle of elite usage (e.g. Cool 2006: 152–​171). Although it might be expected that this demand for imported fine wares would be maintained in rural areas after the conquest, the shift in elite dining and display to the increasingly built-​up and rarefied urban environment effectively raised the bar of exclusivity that only a minority could hope to emulate. Whereas the use of imports in late Iron Age feasting practices seemingly placed few restrictions on appropriate architectural settings, an elegant banquet held in a British round-house would have done little to impress Roman officials accustomed to the levels of luxury experienced in the Mediterranean. Despite evidence for the more widespread uptake of imports, the changed political landscape nevertheless afforded limited opportunities for true emulation and empowerment on the part of provincials living outside the major towns. This did not preclude the possibility of imports being used to engender a sense of social distinction, luxury, and sophistication within local rural communities. Indeed, where continuity in the uptake of Gallo-​Belgic to samian fine wares is apparent at indigenous settlements post-​conquest (for example, Baldock, Hertfordshire, and Heybridge, Essex), evidence for the deposition of such vessels suggests continuity with pre-​Roman and early Roman food ways into the late second century ad (Pitts 2007, 2008). A wholesale shift towards a greater emphasis on dining forms (for example, dishes and bowls) across all assemblages is more likely to reflect the changed priorities of the surviving local elite and the needs of the urban market, rather than the widespread adoption of Roman dining practices and mores by Essex peasants (Perring and Pitts 2013). As Cooper (1996: 95) succinctly stated, ‘many of the people who used Roman-​style pottery had no particular desire to emulate Roman ways but were simply taking what was available to them and busily adapting it as part of the material expression of their own culture’.

Health, Well-​being, and Social Inequality It is undeniable that Roman imperialism had a profound impact on the lives of people who resided in the British countryside. Although the arrival of Rome presented new opportunities for plurality and cultural expression—​for example, through new forms of dress and items of personal adornment (Crummy and Eckardt 2003; Eckardt 2005)—​it is assumed that the new demands of taxation and the provisioning of imposed urban

732

732   Martin Pitts and military populations meant that for the majority this impact would have been more exploitative than facilitating (e.g. Mattingly 2006). This idea forms another component of R. G. Collingwood’s model of the parasitical city, but is again one that remains largely unproven. After all, the entrenched differences in urban and rural patterns of consumption do not provide direct evidence of the degree of human exploitation nor of the extent to which the consequences of Roman imperialism may have affected the quality of rural life. An important question in this respect is whether or not the observed discrepancies in consumption patterns between town and country are indicative of differences in social inequality and the distribution of wealth between urban and rural communities. As Jongman (2007: 596) puts it: social inequality matters for an understanding of the lives of many ordinary Romans. It matters for our understanding of the [economic] growth that did occur, because the wealth of the elite (often paraded by ancient historians as a sign of prosperity) may not have been a sign of a prospering economy after all, but instead of effective exploitation of the poor.

Moreover, recent epidemiological research highlights the startling reality that modern societies with high levels of inequality in the distribution of income also show higher rates of poor health and lower life expectancies, even in wealthy industrialized nations with advanced healthcare provision such as the United States and the United Kingdom (Wilkinson 2005; Wilkinson and Pickett 2010). The implications of these findings with respect to the ancient world are clear: if significant social inequality existed, it should be manifest in patterns of poor health, even if evidence of income is not available to measure inequality directly. Unfortunately, effectively gauging the level of poor health in ancient populations is fraught with challenges. In addition to the familiar archaeological problems of ensuring that the various biological traits visible through the scrutiny of human remains are consistently recorded and quantified between different excavated cemeteries, it is also necessary to mitigate for the effects of other factors that could have a major impact on interpretation. For example, many skeletal conditions that might be associated with hard labour during life are also more likely to occur in particularly long-​lived individuals as a general sign of ageing. Conversely, rather than indicating an individual who had lived a healthy life, a skeleton showing no apparent evidence of disease or trauma could simply have resulted in the case of an unhealthy individual rapidly succumbing to infectious disease, leaving no time for lesions to form on the skeleton—​the so-​called osteological paradox (Wood et al. 1992). In addition, the prevailing ritual of cremation in the early Roman period severely limits the potential insights into changing patterns of health from human remains in Roman Britain, owing to the destructive nature of the rite. Nevertheless, if such problems are taken into account, an initial study of the potential relationship between social inequality and poor health in late Romano-​British cemeteries has yielded some striking results (Pitts and Griffin 2012).

Rural Transformation in the Urbanized Landscape    733 Ancaster 0.5 Baldock 4 Baldock 1

Dimension 2

0.0

Lankhills

Boscombe

Kempston

Baldock 3

Cassington

Cirencester, North Colchester

Chignall St Albans

Derby

Radley, Barrow Hills

Minor urban

Alchester

Rural

Stanton Harcourt

Queensford Mill, Dorchester-on-Thames

Cirencester, South Icklingham

Gloucester, Kingsholm

Dorchester, Tolpuddle Hall

Bletsoe

Dunstable

–0.5

Gloucester, Gambier-Parry Lodge

Major urban

London, West Tenter St

Radley

Leicester, Newarke St –0.5

0.0

0.5

Urban Minor Urban Rural 1.0

Dimension 1

Figure  35.3  Multidimensional scaling analysis of the prevalence of non-​work-​related health conditions in late Roman Britain. Source: after Pitts and Griffin (2012). © Martin Pitts.

Figure 35.3 presents a picture of the overall similarities and differences in health in late Roman Britain by measuring the statistical similarity of 30 late Roman cemetery populations from central and southern England in terms of multiple health indicators, and plotting the results using multidimensional scaling. Each cemetery selected was composed of at least 20 individuals, with data available for at least three of the following eight palaeopathological conditions chosen for their relationship with diet, infectious disease, and/​or as general indicators of poor health: rib periostitis, caries, dental calculus, ante-​mortem tooth loss (AMTL), dental abscesses, periodontal disease, enamel hypoplasia, and cribra orbitalia (for discussion of medicine, see Baker, this volume). Although the results exhibit a degree of overlap when considered by settlement type, a general pattern is visible whereby urban cemeteries are plotted towards the lower left, minor urban cemeteries in a line following the central vertical axis, and rural cemeteries

734

734   Martin Pitts to the upper right. Regression analysis indicates that dimension 1, accounting for the greatest variability in the sample, has a statistically significant relationship with settlement type as well as the prevalences of all the dental health indicators considered apart from caries, including enamel hypoplasia, an indicator of general health. In short, the results show that, as rates of poor health increase, so too does the likelihood that the cemetery in question is associated with a rural settlement. If we take stock of the results outlined in Figure 35.3, the emergent pattern of (relatively) healthy urban populations in contrast to their rural counterparts is somewhat at odds with the squalid picture of Roman urban life painted by ancient historians (e.g. Scobie 1986; Garnsey 1999) and palaeopathologists (Redfern and Roberts 2005). At this point it is important to consider that late Roman towns in Britain were probably not especially densely occupied, which thus reduced the scope for infectious disease and other problems related to inadequate sanitation. In contrast to their predecessors and continental counterparts, late Romano-​British towns were probably relatively sparsely occupied centres of government with large private residences (Millett 1990a: 221–22​3; Mattingly 2006: 325–3​50). In this context, it is not unlikely that the provision of urban amenities outweighed the negative health implications of urban living—​the benefits of Roman civilization highlighted by Collingwood that only the privileged urban dwellers were able to access. Given that dental differences linked to diet form an important component of the patterns displayed in Figure 35.3, it is likely that a particular benefit of urban living was access to better food and dental treatment. The dental differences imply a greater dependency on carbohydrates (that is, cereals) in the countryside and suggest a higher meat component in the urban diet, which is consistent with the higher proportions of animals bred solely for meat (that is, pigs) and evidence for butchery practices indicating the production of filleted beef in urban contexts (Maltby 2007). Access to better food is difficult to measure other than in terms of general health, although this again is consistent with what might be expected in a system involving the transfer of agricultural surplus to urban centres. Indeed, the patterns fit with Galen’s (De alimentorum facultatibus 1.13) second-​century observations on Roman Asia Minor, where rural populations were forced to subsist on poorer-​quality cereals, having sent their best produce to the towns. In the context of recurrent differences between urban and rural settlements in virtually every class of material evidence available, it should come as no surprise that patterns of health in Roman Britain similarly indicate that rural populations were worse off. This did not mean that the country was not a healthy place to be for the wealthy few, but, rather, that the numbers of people enjoying such a lifestyle would have been negligible in the wider context. The wretched patterns of health observed among the small cemetery population associated with the extensive villa at Chignall (Essex) are entirely in keeping with the picture of an overworked workforce of slaves or bonded tenants, as opposed to a wealthy Romano-​British family (Clarke 1998). So far the available data suggest that the major fault-​lines of inequality in the urbanized landscape of Roman Britain existed between urban and rural populations, with minor towns forming something of a compromise between the two extremes. However,

Rural Transformation in the Urbanized Landscape    735 if inequality itself was a cause of such health differences, according to the epidemiological literature (e.g. Wilkinson 2005) we would expect such differences to be underpinned by higher levels of inequality within the most badly affected (rural) communities. This is especially relevant to the strong relationships between settlement type and conditions such as neoplastic disease (cancer) in late Roman Britain observed by Pitts and Griffin (2012), which were likely to have been influenced by differences in living conditions considering the absence of modern carcinogens in the environment. Unfortunately, there is no straightforward means of measuring the level of income inequality between different communities in Roman Britain. To explore this problem, Pitts and Griffin (2012) calculated the level of inequality in the provision of grave furnishings in a sample of late Roman cemeteries, with the basic assumption that a more unequal society would treat its members unequally in death, thus providing a theoretical proxy for inequality in life. Although somewhat speculative, the results outlined in Figure 35.4 show that, on the whole, late Roman urban cemeteries were characterized by lower inequality than their non-​urban counterparts, measured according to the Gini coefficient. The Gini coefficient is a commonly used indicator of inequality in modern societies, yielding a number between 1 and 0, where 0 denotes perfect equality and 1 a scenario of perfect inequality, with one person having everything and everyone else nothing. Significantly, the most notable result to buck this trend was from the urban cemetery of Cirencester, which behaved more like a rural settlement according to indicators such as coin loss (Reece 1993). In the wider context, the picture of inequality in late Roman Britain outlined by Pitts and Griffin (2012) is certainly consistent with Willem Jongman’s argument that the material benefits of the Roman system were increasingly unequally distributed under the late Empire (Jongman 2007: 597). Also included in Figure 35.4 are three early Roman cremation cemeteries from Hertfordshire, which typically show higher levels of furnishing as well as lower levels of inequality in their distribution compared to most late Roman cemeteries. While at face value these patterns appear to validate Jongman’s vision, caution is warranted in the absence of corroborative health data. Indeed, although the early Roman cemeteries in question are not small (with at least 100 individuals in each), the rite of cremation appears less widespread than inhumation in the late Roman period, and it is likely that, in some cases, cremation may have been the exclusive preserve of particular subsets of society, such as collegia (Biddulph 2005) (see also Weekes, this volume, for discussion of burial practice). As we do not know exactly how representative the cemetery populations were of living communities, it is difficult to assess the true significance of the results in Figure 35.4. In material terms, the earlier rite of cremation was intrinsically more egalitarian, with incremental differences in the numbers of pots and brooches per cremation (perhaps indicative of group membership), whereas later Roman inhumation often involved a much wider gulf between the large numbers of simple coffin burials and those interred in more expensive stone coffins or mausolea. Despite the necessary cautions and qualifications, the broad overlap with patterns of health and other evidence for differences between urban and rural communities in late Roman Britain is unlikely to be simple coincidence. One of the main elements contributing

736

736   Martin Pitts 3.5

Braughing

C1-C2 cremation Urban Minor urban Rural

3.0

Mean no. of furnishings per grave

Welwyn 2.5 King Harry Lane

Lankhills

2.0

1.5 Colchester Poundbury 1.0

Ilchester London, East London, West Tenter St Baldock

Bradley Hill Alchester

0.5

Stanton Harcourt Chignall Kempston Dunstable Cirencester Radley, Barrow Hills

Queenford Farm, Dorchester-on-Thames 0.2

0.3

0.4

0.5 0.6 Gini (all furnishings)

0.7

0.8

0.9

Figure  35.4  Comparison of the Gini coefficient of inequality with the mean number of furnishings per grave for selected Romano-​British cemeteries Source: after Pitts and Griffin (2012), with additions. © Martin Pitts.

to the differences in inequality was whether or not a coffin was provided for the deceased. The inability for rural communities to provide what appears to have been regarded as a standard component of the prevailing burial rite reinforces the picture of exploitation beyond the urban sphere. Against this backdrop, it is tempting to view the longer-​term effects of Roman imperialism in Britain in terms of increased oppression and inequality for the rural majority and greater wealth and luxury for the globalized and mobile minority—​most notably evidenced in the flourishing of increasingly palatial villa complexes in the fourth century (Millett 1990a: 227; Perring 2002; contra Hingley 1982: 34–​41). This, of course, remains an impressionistic view that requires further testing. Nevertheless, as the determinants of economic and societal well-​being in the modern world are ever more based on the distribution of wealth, so should the success or failure of former historical eras. Indeed, the degree of social inequality in the Romano-​British countryside should be a firm priority for future research, especially given the tendency for postcolonial narratives to assume high levels of exploitation.

Rural Transformation in the Urbanized Landscape    737

Conclusions R. G. Collingwood’s statement on the parasitical nature of Roman urbanism in Britain was for the most part a deduction made on the basis of limited evidence. Despite its somewhat prophetic nature in the light of the wealth of information now available, it was, above all, a product of its time. In the same book Collingwood explained the identification of contemporary early twentieth-​century Britons with Rome as a civilizing power (and not the ‘native’ Britons), owing to the nineteenth-​ century British experience as a colonial power in central Africa (Collingwood and Myres 1937: 178–​179). In a similar vein, the present account of rural transformation in Roman Britain has deployed two yardsticks of early twenty-​first-​century globalization—​consumption and social inequality—​to make sense of the Roman past. Indeed, it is inevitable that archaeologists and historians will continue to interpret the past through the lens of their contemporary social contexts, and this should not be regarded as an obstacle to understanding, so long as it is done thoughtfully and critically. Most importantly, Collingwood acknowledged the inherent inequalities in the manifestation of Roman imperialism in newly conquered provincial territories—​ not unlike the effects of modern globalization in the developing world (e.g. Bauman 1998), the different forms and underlying mechanisms notwithstanding. That such inequalities would have been felt most acutely in the countryside is unsurprising given the pre-​eminent role of agriculture in the Roman economy and is in broad concordance with the available archaeological evidence—​if not yet proven outright.

References Primary Sources Galen, De alimentorum facultatibus (On the Properties of Foodstuffs. Introduction, Translation and Commentary), trans. O. Powell. Cambridge: Cambridge University Press, 2003.

Secondary Sources Bauman, Z. (1998). Globalization: The Human Consequences. Cambridge: Polity Press. Biddulph, E. (2005). ‘Last Orders:  Choosing Pottery for Funerals in Roman Essex’, Oxford Journal of Archaeology, 24: 23–​45. Biddulph, E., Compton, J., and. Martin, T. S. (2015). ‘The Late Iron Age and Roman Pottery’, in M. Atkinson and S. J. Preston (eds), Heybridge: A Late Iron Age and Roman Settlement. Excavations at Elms Farm 1993–5. Volume 2. Internet Archaeology 40. http://dx.doi. org/10.11141/ia.40.1 (accessed 30 May 2016). Clarke, C. P. (1998). Excavations South of Chignall Roman Villa, Essex, 1977–​81. East Anglian Archaeology Report 83. Chelmsford: East Anglian Archaeology. Collingwood, R. G., and Myres, J. N. L. (1937). Roman Britain and the English Settlements. Oxford: Oxford University Press.

738

738   Martin Pitts Cool, H. E.  M. (2006). Eating and Drinking in Roman Britain. Cambridge:  Cambridge University Press. Cool, H. E. M., and Baxter, M. J. (1999). ‘Peeling the Onion: An Approach To Comparing Vessel Glass Assemblages’, Journal of Roman Archaeology, 12: 72–​100. Cool, H. E.  M., and Baxter, M. J. (2002). ‘Exploring Romano-​British Finds Assemblages’, Oxford Journal of Archaeology, 21: 365–​380. Cooper, N. (1996). ‘Searching for the Blank Generation:  Consumer Choice in Roman and Post-​Roman Britain’, in J. Webster and N. Cooper (eds), Roman Imperialism: Post-​Colonial Perspectives. Leicester Archaeology Monographs 3. Leicester: Leicester Archaeology, 85–​98. Crummy, N., and Eckardt, H. (2003). ‘Regional Identities and Technologies of the Self: Nail Cleaners in Roman Britain’, Archaeological Journal, 160: 44–​69. Crummy, P., Benfield, S., Crummy, N., Rigby, V., and Shimmin, D. (2007). Stanway: An Élite Burial Site at Camulodunum. Britannia Monograph Series 24. London: Britannia. Eckardt, H. (2002). Illuminating Roman Britain. Montagnac: Editions Monique Mergoil. Eckardt, H. (2005). ‘The Social Distribution of Roman Artefacts: The Case of Nail-​Cleaners and Brooches in Britain’, Journal of Roman Archaeology, 18: 139–​160. Eckardt, H. (2007). ‘Contexts in Colchester’, in R. Hingley and S. Willis (eds), Roman Finds: Context and Theory. Oxford: Oxbow Books, 142–​149. Evans, J. (1995). ‘Roman Finds Assemblages, towards an Integrated Approach?’, in P. Rush (ed.), Theoretical Roman Archaeology: Second Conference Proceedings. Aldershot: Avebury, 33–​58. Evans, J. (2001). ‘Material Approaches to the Identification of Different Romano-​British Site Types’, in S. James and M. Millett (eds), Britons and Romans: Advancing an Archaeological Agenda. CBA Research Report 125. York: Council for British Archaeology, 26–​35. Evans, J. (2005). Pottery in Urban Romano-​British Life’, in A. MacMahon and J. Price (eds), Roman Working Lives and Urban Living. Oxford: Oxbow Books, 145–​166. Frere, S. S. (1987). Britannia: A History of Roman Britain. London: Pimlico. Fulford, M. G. (1982). ‘Town and Country in Roman Britain–​a Parasitical Relationship?’, in D. Miles (ed.), The Romano-​British Countryside: Studies in Rural Settlement and Economy. BAR British Series 103. Oxford: Archaeopress, 403–​419. Fulford, M. G., and Hodder, I. (1975). ‘A Regression Analysis of Some Late Romano-​British Pottery: A Case-​Study’, Oxoniensia, 39: 26–​33. Gaffney, V., and White, R. [with Goodchild, H.] (2007). Wroxeter, the Cornovii, and the Urban Process. Final Report on the Wroxeter Hinterland Project 1994–​1997. Volume 1: Researching the Hinterland. JRA Supplementary Series 68. Portsmouth, RI: Journal of Roman Archaeology. Garnsey, P. (1999). Food and Society in Classical Antiquity. Cambridge:  Cambridge University Press. Going, C., and Belton, J. (1999). ‘Roman Pottery’, in N. R. Brown (eds), The Archaeology of Ardleigh, Essex: Excavations 1955–​1980. East Anglian Archaeology Report 90. Chelmsford: East Anglian Archaeology, 125–​157. Haselgrove, C. C. (1982). ‘Weath, Prestige and Power: The Dynamics of Late Iron Age Political Centralisation in South-​East England’, in C. Renfrew and S. Shennan (eds), Ranking, Resource and Exchange. Cambridge: Cambridge University Press, 79–​88. Hingley, R. (1982). ‘Roman Britain: The Structure of Roman Imperialism and the Consequences of Imperialism on the Development of a Peripheral Province’, in D. Miles (ed.), The Romano-​ British Countryside:  Studies in Rural Settlement and Economy. BAR British Series 103. Oxford: Archaeopress, 17–​52. Hingley, R. (1989). Rural Settlement in Roman Britain. London: Seaby.

Rural Transformation in the Urbanized Landscape    739 James, S. (2001). ‘“Romanization” and the Peoples of Britain’, in S. Keay and N. Terrenato (eds), Italy and the West: Comparative Issues in Romanization. Oxford: Oxbow Books, 187–​209. Jongman, W. (2007). ‘The Early Roman Empire:  Consumption’, in W. Scheidel, I. Morris, and R. Saller (eds), The Cambridge Economic History of the Greco-​ Roman World. Cambridge: Cambridge University Press, 592–​618. King, A. C. (1978). ‘A Comparative Survey of Bone Assemblages from Roman Sites in Britain’, Institute of Archaeology Bulletin, 15: 207–​232. King, A. C. (1984). ‘Animal Bones and the Dietary Identity of Military and Civilian Groups in Roman Britain, Gaul and Germany’, in T. F. C. Blagg and A. C. King (eds), Military and Civilian in Roman Britain. BAR International Series 136. Oxford: Archaeopress, 187–​217. King, A. C. (1999). ‘Diet in the Roman World:  A  Regional Inter-​Site Comparison of the Mammal Bones’, Journal of Roman Archaeology, 12: 168–​202. King, A. C. (2001). ‘The Romanization of Diet in the Western Empire’, in S. Keay and N. Terrenato (eds), Italy and the West: Comparative Issues in Romanization. Oxford: Oxbow Books, 210–​223. Lockyear, K. (2000). ‘Site Finds in Roman Britain:  A  Comparison of Techniques’, Oxford Journal of Archaeology, 19: 397–​423. Maltby, M. (1989). ‘Urban–​Rural Variations in the Butchering of Cattle in Romano-​British Hampshire’, in D. Serjeantson and T. Waldron (eds), Diet and Crafts in Towns. BAR British Series 199. Oxford: Archaeopress, 75–​106. Maltby, M. (1994). ‘The Meat Supply in Roman Dorchester and Winchester’, in A. Hall and H. Kenward (eds), Urban–​Rural Connexions: Perspectives from Environmental Archaeology. Oxford: Oxbow Books, 85–​102. Maltby, M. (2007). ‘Chop and Change:  Specialist Cattle Carcass Processing in Roman Britain’, in B. Croxford, N. Ray, R. E. Roth, and N. White (eds), TRAC 2006: Proceedings of the Sixteenth Annual Theoretical Roman Archaeology Conference. Oxford:  Oxbow Books, 59–​76. Mattingly, D. J. (1997) (ed). Dialogues in Roman Imperialism. JRA Supplementary Series 23. Portsmouth, RI: Journal of Roman Archaeology. Mattingly, D. J. (2004). ‘Being Roman: Expressing Identity in a Provincial Setting’, Journal of Roman Archaeology, 17: 5–​25. Mattingly, D. J. (2006). An Imperial Possession: Britain in the Roman Empire. London: Penguin. Millett, M. (1979). ‘The Dating of Farnham Pottery’, Britannia, 10: 121–​138. Millett, M. (1982). ‘Town and Country: A Review of Some Material Evidence’, in D. Miles (ed.) The Romano-​British Countryside:  Studies in Rural Settlement and Economy. BAR British Series 103. Oxford: Archaeopress, 421–​431. Millett, M. (1990a). The Romanization of Britain. Cambridge: Cambridge University Press. Millett, M. (1990b). ‘Romanization:  Historical Issues and Archaeological Interpretation’, in T. F. C. Blagg and M. Millett (eds), The Early Roman Empire in the West. Oxford: Oxbow Books, 35–​41. Niblett, R. (1999). The Excavation of a Ceremonial Site at Folly Lane, Verulamium. London: Society for the Promotion of Roman Studies. Perring, D. (2002). The Roman House in Britain. London: Routledge. Perring, D., and Pitts, M. (2013). Alien Cities: Consumption and the Origins of Urbanism in Roman Britain. Spoilheap Publications Monograph 7. London: Spoilheap Publications. Pitts, M. (2005). ‘Pots and Pits: Drinking and Deposition in Late Iron Age South-​East Britain’, Oxford Journal of Archaeology, 24: 143–​161.

740

740   Martin Pitts Pitts, M. (2007). ‘Consumption, Deposition and Social Practice: A Ceramic Approach to Intra-​ Site Analysis in Late Iron Age to Roman Britain’, Internet Archaeology. Pitts, M. (2008). ‘Globalizing the Local in Roman Britain: An Anthropological Approach to Social Change’, Journal of Anthropological Archaeology, 27: 493–​506. Pitts, M. (2010). ‘Artefact Suites and Social Practice:  An Integrated Approach to Roman Provincial Finds Assemblages’, Facta:  A  Journal of Roman Material Culture Studies, 4: 125–​152. Pitts, M., and Griffin, R. (2012). ‘Exploring Health and Social Well-​Being in Late Roman Britain: An Inter-​Cemetery Approach’, American Journal of Archaeology, 116: 253–​276. Redfern, R., and Roberts, C. (2005). ‘Health in Romano-​ British Urban Communities: Reflections from the Cemeteries’, in D. Smith, M. Brickley, and W. Smith (eds), Fertile Ground: Papers in Honour of Susan Limbrey, Symposia of the Association of Environmental Archaeology. Oxford: Oxbow Books, 115–​129. Reece, R. (1993). ‘British Sites and their Roman Coins’, Antiquity, 67: 863–​869. Richmond, I. (1969). ‘The Romano-​British Countryside’, in P. Salway (ed.), Roman Archaeology and Art: Essays and Studies by Sir Ian Richmond. London: Faber and Faber, 133–​180. Rivet, A. L. F. (1964). Town and Country in Roman Britain. London: Hutchinson. Said, E. (1993). Culture and Imperialism. London: Chatto and Windus. Scobie, A. (1986). ‘Slums, Sanitation, and Mortality in the Roman World’, Klio, 68: 399–​433. Swan, V. G. (1984). The Pottery Kilns of Roman Britain. London: Royal Commission on Historical Monuments. Van der Veen, M., Livarda, A., and Hill, A. (2008). ‘New Plant Foods in Roman Britain: Dispersal and Social Access’, Environmental Archaeology, 13: 11–​36. Wacher, J. (1995). The Towns of Roman Britain. London: Routledge. Wallerstein, I. (1974). The Modern World System I. Capitalist Agriculture and the Origins of the European World-​Economy in the Sixteenth Century. London: Academic Press. Webster, J. (1997). ‘A Negotiated Syncretism: Readings on the Development of Romano-​Celtic Religion’, in D. J. Mattingly (ed.), Dialogues in Roman Imperialism. Portsmouth, RI: JRA Supplementary Series 23. Portsmouth, RI: Journal of Roman Archaeology, 165–​184. Webster, J. (2001). ‘Creolizing the Roman Provinces’, American Journal of Archaeology, 105: 209–​225. Webster, J., and Cooper, N. (1996) (eds). Roman Imperialism:  Post-​Colonial Perspectives. Leicester Archaeology Monographs 3. Leicester: Leicester Archaeology. Wilkinson, R. G. (2005). The Impact of Inequality:  How to Make Sick Societies Healthier. London: Routledge. Wilkinson, R. G., and Pickett, K. (2010). The Spirit Level: Why Equality is Better for Everyone. London: Penguin. Willis, S. (1996). ‘The Romanization of Pottery Assemblages in the East and North-​East of England during the First Century ad: A Comparative Analysis’, Britannia, 27: 179–​221. Willis, S. (2007). ‘Roman Towns, Roman Landscapes:  The Cultural Terrain of Town and Country in the Roman Period’, in A. Fleming and R. Hingley (eds), Prehistoric and Roman Landscapes: Landscape History after Hoskins Volume 1. Macclesfield: Windgather Press, 143–​164. Willis, S. (2011). ‘Samian Ware and Society in Roman Britain and Beyond’, Britannia, 42: 167–​242. Wood, J. W., Milner, G. R., Harpending, H. C., and Weiss, K. M. (1992). ‘The Osteological Paradox:  Problems of Inferring Prehistoric Health from Skeletal Samples’, Current Anthropology, 33: 343–​370.

Chapter 36

The Devel opme nt of Tow ns Adam Rogers

Introduction Towns and urban development have formed a major part of the history of studies of Roman Britain, and some of the towns are among the most heavily investigated Roman settlements. For centuries Roman material has been encountered within living cities, and, in some instances, there are Roman remains still standing. The nineteenth century saw the commencement of some large-​scale excavation programmes, especially at town sites unoccupied by later settlement, including Silchester, Caerwent, Verulamium, and Wroxeter. The investigation of these green-​field sites was instrumental in the formation of models for urban development, but interpretations were also influenced by the cultural context and social preoccupations of the day (cf. Hingley 2008). The post-​war years saw a massive increase in knowledge relating to those towns now lying beneath modern settlements through urban redevelopments and the formation of urban archaeological units. Much of this work was collated for the first time in Wacher’s important narrative (1975) on the Roman towns of Britain. There are now many important studies and syntheses of urban development taking different perspectives (e.g. Millett 1990; Wacher 1995; M. J. Jones 2004; Hingley 2005; Creighton 2006; Mattingly 2006; Rogers 2008), but our knowledge and understanding of town development remains open to constant revision as new discoveries are made and analyses are undertaken. Especially important in more recent years has been the development of theoretical approaches in Roman archaeology and the potential that this has for studying settlement. Towns were new to Britain in the Roman period, although there were also a range of large settlements and complexes in prehistory. There has, however, tended to be a divide between the approaches taken by prehistorians and Romanists in addressing social themes relating to processes of settlement development, their relationship to human identities and experiences, and other issues. While hillforts were important

742

742   Adam Rogers for much of the Iron Age, in the second and first centuries bc another group of sites appeared—​often referred to collectively as oppida and conventionally divided into ‘enclosed’ and ‘territorial’ oppida based on the nature of the earthworks evident (e.g. Haselgrove 1999). Some Roman towns were located in or near oppida, but there is still considerable uncertainty as to their real nature and function and whether we can think of oppida as a clear category of site or whether there was more variety and variation of sites and circumstances than has so far been fully appreciated (cf. Haselgrove and Moore 2007). While it would be unwise to attempt to apply Classical or later urban models to these sites, moreover, including in debating the role of the earthworks, it might be possible to think of them as taking another form of trajectory in urban development (cf. Sharples 2010: 173); oppida may also have had more impact on the urban geographies of Roman Britain than is often acknowledged, and it is important for Roman archaeology to draw on landscape theory. As such it is important for studies of Roman urban development to engage in theoretical debates relating to landscape archaeology as well as theories concerning interconnected issues including identity and experience. The divide between Roman and Iron Age is as much about archaeological perspectives as historical events, and theoretical frameworks can help us to adopt more nuanced approaches to the transition from Iron Age to Roman Britain.

Categorizing Towns: Regionality Central to archaeology have been attempts to categorize the past:  be it by periods, objects, or settlements. Romano-​British studies have created lists of characteristics of what defines certain types of settlements, drawing on knowledge of archaeology, textual sources, and the Roman Empire as a whole. Inevitably, however, these attempts will also have imposed modern perspectives and preoccupations onto the past, and there is always a danger in oversimplifying our understanding of settlements through these definitions. The towns of Roman Britain are usually defined by their planned and organized appearance, their size, the presence of a variety of buildings, a concentration of population, and the administrative and economic roles that they had (e.g. Mann 1996; Burnham et al. 2001). But it is undoubtedly the legal categorizations known through written records that have been most instrumental in defining urbanism and urban development in Roman Britain. It is important to recognize, however, that there may well also have been other forms of urbanism in Britain at this time, alongside the towns included in more conventional studies. It is known that there was a legal hierarchy of urban settlements, with coloniae at the top, municipia, and then civitas-​capitals or centres. The colonia was a chartered town, by the first century ad generally founded for the settlement of discharged veterans, and it had a constitution that was modelled on Rome and adopted Roman law. In Britain, the three initial colonia were founded at Colchester (Camulodunum), Gloucester (Glevum), and Lincoln (Lindum), on the sites of legionary fortresses that have traditionally

The Development of Towns    743 formed important parts of the narrative of the conquest and occupation of the province (e.g. Frere 1967; Wacher 1975). Legio XX is conventionally assigned to the fortress at Colchester, founded around ad 43–​4 and replaced by the colonia from around ad 49. In the ad 60s Legio IX Hispania was assigned to Lincoln (used later by Legio II Adjutrix) and another, as yet undetermined, legion to Gloucester, although Legio II Augusta is thought to have occupied the fortress at some stage (Wacher 1995: 150). While the dates have been the subject of much debate, the colonia at Lincoln was probably founded in the ad 80s (M. J. Jones 2002) and at Gloucester in the 90s (Copeland 2011). Even this most familiar of narratives of development, however, is not safe from reassessment, not least because so little is actually known about the composition of the army in Britain at any one time or of its movements (Mattingly 2006: 130). At Lincoln, for example, the possibility of an earlier Claudian military base has been raised, which is based on some evidence for an early military cemetery (M. J. Jones 2003a: 38–​39), but until more positive structural evidence comes to light this must remain a subject for debate. The coloniae in Britain have generally been treated as Classical-​style settlements with Roman populations; they have been described as ‘bulwarks of loyalty’ (Collingwood and Richmond 1969: 95) and ‘instrument(s) of civilisation’ (Richmond 1963: 55). Many of the inhabitants will have been Roman citizens, but there will also have been other inhabitants, especially local indigenous people, who would not have been citizens, indicating a mix of people and perspectives with social implications. Below the colonia in the hierarchy was the municipium, which was also a chartered town, but far fewer inhabitants automatically received citizenship, with only the ex-​magistrates usually having the right to acquire citizen status. Verulamium was described by Tacitus (Ann. XIV.33) as a municipium, but exactly when, and even if, the town actually acquired this status and the reasons behind it are uncertain; it may well have been an upgrade from civitas-​capital made in the ad 70s (Niblett 2001: 66). York (Eboracum) was described by the fourth-​century writer Aurelius Victor as a municipium at the time of Septimius Severus’ visit in the early third century (Aur. Vict. Caes. XX; Ottaway 2004: 83). It is not entirely clear how much value can be placed on the terms used in these accounts, since they may simply have been used for convenience or individual preference. However, an inscription on an altar dating to ad 237 (AE 116 [1922]), set up in Bordeaux by Marcus Aurelius Lunaris for the goddess Tutela Boudiga, indicates that by this date at least York had become a colonia. He described himself as a sevir Augustalis (priest of the deified emperor) from the coloniae of both Lincoln and York. It is possible in the case of York that the settlement was later granted the status of colonia by the emperor, perhaps after petitioning for it, as occurred elsewhere (cf. Boatwright 2000), and was not founded in the same sense as the earlier coloniae in Britain. The largest group in the urban hierarchy was what are usually termed civitas-​capitals or civitas centres and regarded as the centres of the civitates: the means by which land and people were divided up for administration. Although lower in status, the constitution, like the other towns, was also ultimately modelled on Rome, with laws, annual magistracies, and a town council. We still know frustratingly little about the nature of town councils or other forms of organization and management within the towns of

744

744   Adam Rogers Roman Britain. It seems likely that the circumstances differed within each town depending on local peoples, histories, and the nature of the urban development. The province was large, and, because of this, the regions, local conditions, and peoples are important factors to consider in the development of towns and the meanings that will have been attached to them as places. It is the recorded names of the civitas-​capitals that have conventionally been taken as evidence of processes of negotiation with local people as civitates and towns were formed: for example (Figure 36.1), Silchester (Calleva Atrebatum); Winchester (Venta Belgarum); Caistor-​by-​Norwich (Venta Icenorum); Leicester (Ratae Corieltauvorum); Cirencester (Corinium Dobunnorum); Dorchester (Durnovaria); Exeter (Isca Dumnoniorum); Wroxeter (Viroconium Cornoviorum). The existing ‘tribal’

Coria Luguvalium Carvetiorum Isurium Brigantum Cataractonium

Eburacum Petuaria Parisiorum Colonia Domitiana Lindensium

Ratae Corieltauvorum Venta Viroconium Durobrivae Icenorum Cornoviorum Colonia Claudia Victricensis Colonia Nervia Camulodunensium Glevensium Moridunum Demetarum Caesaromagus Corinium Venta Silurum Dodunnorum Verulamium Londinium Calleva Atrebatum Durovernum Cantiacorum Venta Belgarum Lindinis Durotrigum Noviomagus Reg(i)norum Durnovaria Durotrigum Isca Dumnoniorum

Figure 36.1  Map of southern Britain with the main settlements mentioned in the chapter. Source: © A. C. Rogers.

The Development of Towns    745 organization of Britain prior to the conquest, however, remains a focus of debate, since the main evidence comes from texts written by Roman authors and inscriptions that date to after ad 43. It seems likely that Rome had an influence on the creation of these entities either before or after ad 43, but the nature of this remains uncertain (e.g. Moore 2006, 2011; Sharples 2010). Group identities and allegiances may well have been more fluid in the late Iron Age, perhaps based on individuals’ fluctuating capabilities of encouraging the support of others, and more localized settlement and household identities will also have been important. Ongoing work at the Bagendon/​Ditches complex north of Cirencester is certainly indicating that the negotiations for power and the beginnings of urban development at Corinium are likely to have been more complex than traditional models of tribal oppidum to town (Moore 2011). In the second century further towns appear to have been founded: Caerwent (Venta Silurum), Carmarthen (Moridunum Demetarum), and the less-​understood settlements of Brough-​on-​Humber (Petuaria Parisiorum) and Aldborough (Isurium Brigantum). These remind us of the complex ongoing developments and changes that took place in Roman Britain, although the nature of the processes and negotiations that may have taken place with their foundation remain obscure to us. Settlement size has also been used to categorize towns and also to comment on the perceived success of Roman urbanism in Britain. The towns of Roman Britain are generally fairly small in size, with many of the main enclosed areas falling to around 40 hectares, including Canterbury, Leicester, Exeter, Silchester, and Lincoln. Larger towns included Cirencester at around 97 hectares, Verulamium at 80 hectares, and London at 133 hectares; the enclosed area at Gloucester, however, was only around 16.6 hectares. Settlement, however, will also have extended beyond the enclosures (cf. Esmonde Cleary 1987). Another category of settlement in Romano-​British archaeology is what is generally referred to as the ‘small town’ (e.g. Burnham and Wacher 1990), although these are not always any smaller than the civitas-​capitals or coloniae. Although it seems clear that they did not have the same legal status (at least in terms of that imposed by the empire) within the province, they may have had a more prominent role within the civitates than has generally been acknowledged (cf. Laurence 2001; Hingley 2005). Mattingly (2006: 281) has suggested that the territories assigned to the civitas-​capitals may have been smaller than is conventionally attributed to them, although there remains frustratingly scant evidence to support arguments either way. Since it is likely that all land was attributed to some kind of authority for the purposes of tax-​gathering, it would mean that other forms of settlements and landownership with the civitates were also important. These issues remain debatable and highlight the constant need for discussion without taking the evidence for granted. Hingley (1997) has also argued that ‘small towns’ may reflect more local indigenous interpretations and expressions of urbanism, perhaps not unlike how we might be able to interpret some of the oppida. This might help to explain the apparent diverse range of contexts of ‘small-​town’ development and function and reminds us of the possibility of different forms of urbanism in the Roman period. The debate is a reminder of the necessity to avoid the ‘one-​model-​fits-​all’ approach to urban development and

746

746   Adam Rogers instead examine the complex processes and negotiations that must have occurred in the local context of each site. As noted for Verulamium and York, and possibly also London (cf. Tomlin 2006), urban status could change. It has been argued that some ‘small towns’ possibly sought and were granted status as civitas-​capitals in the later Roman period, including Water Newton (Durobrivae) and Ilchester (possibly the Lindinis in the Ravenna Cosmography; or the Civitas Durotrigum Lindinesis on RIB 1672 and 1673 from Cawfields on Hadrian’s Wall) (Leach 1982; Fincham 2004; Putnam 2007; cf. Fulford 2006). These arguments are based principally on settlement size and the buildings they contained, as there is no textual evidence for these changes. The town at Chelmsford in Essex (Caesaromagus) is often included in discussions of civitas-​capitals (e.g. Wacher 1995), but in other accounts it is treated as a ‘small town’ (e.g. Mattingly 2006: 259). Wacher (1995: 207) suggested that it may have been a largely unsuccessful civitas-​capital, because of his belief that the colonia at Colchester would not have controlled the civitas. It might be that attempting to fit the settlement within the category of either civitas-​capital or ‘small town’ is inadequate, especially since there is little consensus as to what we mean by ‘small town’ and how these settlements functioned. In the north the military establishment of Carlisle (Luguvalium Carvetiorum) may have become a civitas-​capital in the third century (Burnham and Wacher 1990: 58; M. J. Jones 2004: 174; Mattingly 2006: 261).

Town Development and Texts One of the earliest forms of text relating to places in Britain are the names incorporated into Iron Age coin designs, which have long been associated with the oppida known in the vicinity of some of the Roman towns (cf. Hingley 2008). Coin distributions have been studied to assess ownership of oppida, and the work of antiquarians has been instrumental in our recognition of Iron Age and Roman sites and the development of concepts of ‘Romanization’ and ‘civilization’. The name Calleva (Silchester) first appears on the coinage of Eppillus dating to the beginning of the first century ad (Creighton 2006: 64), the self-​proclaimed son of Commius (who is mentioned in Caesar’s Bellum Gallicum IV.21). The coin distribution then changes to those that refer to Epatticcus, son of Tasciovanus and brother of Cunobelin (Creighton 2006: 64), which is often taken to imply some kind of violent conquest of the area or a dynastic union, perhaps through marriage (ibid.). Verulamium is also first named on coinage, with the issues of Tasciovanus inscribed with VER, and Camulodunum on the coinage of Cunobelin as CAMV (Potter 2002: 21; Williams 2005: 33) (see also Walton and Moorhead, this volume, for further discussion on coinage). The names Verlamion and Camulodunon are sometimes suggested for the Iron Age period settlements in order to distinguish them from the Roman towns (Potter 2002: 21). There is no definite evidence for the exact form of these names, but it seems likely that they did differ in the Iron Age. Cassius Dio mentions that Camulodunum was Cunobelin’s capital in his Histories (LX.21), but caution is needed as to whether he would

The Development of Towns    747 have understood the social context of this place at the time he was writing about (cf. Haselgrove 2004; Haselgrove and Moore 2007; Sharples 2010). There are few surviving texts that mention the towns of Roman Britain, and there is very little in the way of any descriptions of the towns in the sources that do exist. The narrative of the Boudiccan Revolt of ad 60/​61 refers to the towns of Camulodunum, Verulamium, and Londinium (Tac. Ann. XXXI–​XXXIII) and, in so doing, provides a brief comment on Camulodunum with its statue of Victory, a curia, a theatre, and a temple where much of the population sought safety during the revolt, but which was ultimately destroyed. Tacitus also recorded that there was no ditch or rampart surrounding the town at this time, which made it easier to attack. As part of the same narrative, he mentions that Londinium was at that time full of merchants and trading vessels (Ann. XXXIII), although he did not see this for himself and was writing at a later date, the Annales probably being written in the early second century ad. There are also other forms of texts that refer to the towns of Roman Britain but provide little or no information as the nature of the settlements. Ptolemy’s Geographia, probably written in the ad 140s, consists of a list of place names and other geographical features from across the empire, which had probably been drawn from a number of sources from different periods. Another collection of place names is the Ravenna Cosmography compiled around ad 700 and using a number of maps that are now lost. Despite the sparse details, its late date, and the many mistakes and ambiguities, perhaps through copying errors, it has been a hugely influential text in Romano-​British studies, since it lists the colonia and assigns tribal names to the civitas-​capitals. It was used by Haverfield (1912: 60–​61) and others to create their reconstructions of Roman urban development in Britain with the civitas system and the processes of conquest, tribal domination, and Romanization. There is also the Antonine Itinerary, long studied in relation to Britain (Hingley 2008), probably compiled in the early third century. It consists of a collection of itineraries or routes across the provinces, with fifteen relating to Britain and mentioning over 100 place names. Many of these names can be recognized, and have also been identified archaeologically, but there are also a number that remain enigmatic. While these sources are important, as lists they provide no information on the individual circumstances surrounding the creation of each settlement or the lives of the people involved.

Debating Urbanization and Romanization Many antiquarians and scholars have been important in developing our understanding of Roman urban development, but it could be argued that Francis Haverfield was one of the first in the context of modern archaeological scholarship (see also Millett and Swift, this volume). His The Romanization of Roman Britain (1912) described the towns

748

748   Adam Rogers of Roman Britain in terms of the ‘five municipalities of the privileged Italian type’ and their dependent territories (Colchester, Verulamium, Lincoln, Gloucester, and York) and what he termed the cantonal system, which was one of the important ‘influences which aided the Romanization of the country’ and ‘permitted the complete remodelling of the native institution by the interpenetration of Italian influences’ (Haverfield 1912: 58). Though he did recognize that urbanization would have been ‘a subtler and more complex process than mere absorption in Rome’ (Haverfield 1912: 64), with towns containing native as well as Roman elements, his arguments did focus principally in the Roman perspective with an emphasis on Romanization: ‘Britain shared in that expansion of town-​life which formed a special achievement of the Roman Empire’ (Haverfield 1912: 64). Frere’s analysis (1967) of towns also drew on the Classical perspective and argued that urbanism was not particularly successful: Britain ‘never produced towns of fully Mediterranean character’. Frere’s analysis, however, considered Roman towns to have been important for the ‘spread of education and Roman civilisation’ (Frere 1987: 192) and the preceding oppida as no more than ‘amorphous collection(s) of roundhouses and unorganised squalor’ (Frere 1987: 192). He also thought of Camulodunum and Verulamium as acting as models for urban development where buildings, town life, and organization could be copied by local peoples in other areas of Britain, and this meant that these two towns received government assistance in their construction. Wacher’s monumental The Towns of Roman Britain (1975) was keen to place an emphasis on the observed pattern of urban development as the result of a well-​planned programme of conquest and expansion. In his book, the towns were divided into a series of chronological chapters entitled ‘The Coloniae’ (Colchester, Lincoln, Gloucester, and York); then ‘The First Civitas Capitals’ (Canterbury, Chelmsford, and Verulamium); ‘From Client Kingdom to Civitas’ (Caistor-​by-​Norwich, Chichester, Silchester, and Winchester); ‘Flavian Expansion’ (Cirencester, Dorchester, Exeter, Leicester, and Wroxeter); and finally ‘Hadrianic Stimulation’ (Caerwent, Carmarthen, Brough-​on-​ Humber, and Aldborough). This was the predominant model of the time, which placed an emphasis on the foundation of towns when forts were abandoned and the army moved on (e.g. G. Webster 1966). The creation of towns was predominantly the result of military strategy, even though at not every town is there evidence of any preceding military settlement. Millett’s The Romanization of Britain (1990) can be seen as an important work of transition in Roman urban studies that sought to place more significance and distinction between the towns with definite military origins and those with indigenous involvement, emphasizing, in particular, the role of the oppida in the development of the civitas system. For Millett, the four coloniae and London had an essentially anomalous position within the settlement system, with the three early coloniae representing deliberate acts of policy by the Roman administration. The civitas system, on the other hand, developed with the aid of the local elite of tribal groups. He constructed a detailed argument contrasting military and native origins of urbanism, but, as the title of the book indicates, his work was not a critique of Romanization but rather a reanalysis of the way in which it took place.

The Development of Towns    749 More recent studies are suggesting that the distinction between military and indigenous origins, moreover, was probably in fact more blurred (cf. Hingley 2005:  271). Mattingly (2006) has argued against there having been an overall plan in urban development in Britain; instead he sees the current picture of urban development as one that resulted from a variety of negotiations between Rome and the peoples of Britain that contributed to the complex individual histories of each settlement. But he did not deny Rome’s power to make decisions:  for Mattingly, the pre-​Roman political geography would not necessarily have been of primary importance; instead what was significant was the resultant circumstances, as people may have surrendered, made alliances, or were subdued. The need to create a new communications network may also have been considered a priority, and settlements were able to develop that were close to this network. Placing more emphasis on pre-​conquest activity, Creighton’s work (2006) has provided an important reanalysis of the chronological sequence of a number of towns, arguing for the importance of the relationship between Rome and local elites vying for power and the development of some of these places before ad 43. The observed developments at these places, however, were not simply about Romanization, since the elites were able to draw on Roman elements and adapt them to emphasize their own power and achievements. Our understanding of urban development, then, is clearly influenced by what perspective is taken, and so it would, perhaps, be most productive to acknowledge the actions and viewpoints of the full range of people associated with each town. Any study of Roman urbanism, then, must tie into debates relating to ‘Romanization’ (Haverfield 1912; Millett 1990; Hingley 2000; Mattingly 2006). Increasingly, alternative terms and ways of understanding the processes of change, continuity, and development that took place in Britain during the Roman period are being sought (e.g. J. Webster 2001; Mattingly 2004, 2006; Hingley 2005) and the historiographical context of the term scrutinized (e.g. Hingley 2000, 2008). Individual negotiations with the presence of Rome indicate that change and continuity were far from uniform processes across Britain, and these non-​uniformities must take into account not only elite but also non-​ elite perspectives. There were peoples who did not wish to engage in the Roman presence, and they may have been either encouraged or forced to live in a town or left more to their own devices in the countryside. There is a tendency in studies of the development of Roman urbanism in Britain to focus on the Roman perspective or that of the indigenous elite, but other perspectives of these monumental processes taking place within the landscape must also be considered.

Town Location and Place Most descriptions of Roman town development in Britain include a discussion of the locations of the settlements, including the strategic, economic, and practical benefits of these sites. Drawing on the empiricist traditions of landscape archaeology, these descriptions have tended to emphasize the practical value and rationality of landscapes

750

750   Adam Rogers relating to urban development. At Colchester, for example, the fortress was in a well-​ chosen location occupying a ridge of land overlooking the Colne Valley and (on a neutral site) within the oppidum, because it was believed to have been placed away from any existing settlement (e.g. Wacher 1995; Crummy 1997); Roman Gloucester ‘overlooks the lowest point where the Severn could be bridged until the Industrial Revolution’ and was the main route into south Wales so was of ‘key strategic importance’ (Hurst 1988: 48). These practical considerations are important, but they provide an external perspective and give little or no consideration to the social value of these locations as places and their longer-​term biographies as foci of human actions, histories, experiences, emotions, and stories. Landscapes were not simply functional, rational, and mundane: ‘natural’ features in the landscape, for example, could have cultural values associated with them and form just as important elements of the settlement record as earthworks and buildings (cf. Rogers 2008). More emphasis can be given to the meanings associated with the locations in which towns were sited through a study of the history of these places and the cultural values that could be given to features in that landscape. Regardless of the political geography created in the Roman period, meanings associated with specific places will also have been drawing on the existing significance associated with them. Known place names may be able to provide some initial hints as to the possible existing meanings attached to some of the places where the towns were founded and the way in which this had an impact on the urban experience. The name Calleva (Silchester), for example, has been interpreted as meaning ‘(the settlement in the) woods’ (Rivet and Smith 1979: 291). It is possible that the oppidum developed here through clearing a space in the woodland or there was perhaps even a sacred grove here. The name Durovernum (Canterbury) appears to relate to wet conditions here, with ‘verno-​’ meaning alder swamp or marsh, suggesting that this location was also culturally meaningful (Rivet and Smith 1979: 53). The meaning of Corinium (Cirencester), though not entirely clear, may relate to a botanical feature also in a wetland area, while Isca (Exeter) is derived from the river (Rivet and Smith 1979: 353, 376). These features were not just mundane points of identification but will have been imbued with meanings and cultural values. Studies of the coloniae have tended to focus on the fortress or town itself rather than their wider setting. The colonia at Colchester was placed within a ‘territorial’ oppidum consisting of a number of dykes, enclosures, and other foci (Hawkes and Crummy 1995). Although the town developed on the site of the fortress, there were other foci connected with the oppidum that continued to be important and formed part of the urban landscape, including what became a temple and theatre complex at Gosbecks, where there was already a number of enclosures, including what is often termed ‘Cunobelin’s farmstead’, and a foci of temples at Sheepen (Crummy 1984, 1997; Niblett 1985). The Roman-​ period urban landscape drew upon and became part of the pre-​existing geography. There has also been one tentative suggestion that the Iron Age coins and pottery found on excavations within the town may indicate some earlier form of settlement or activity here, perhaps even a religious sanctuary (Brooks 2006). The coloniae not located near or within obvious oppida will also have been placed within landscapes of pre-​existing use and meaning. The Roman fortress at Lincoln

The Development of Towns    751 overlooked the valley where the river Witham met the river Till, and there was a large natural body of water known as the Brayford Pool where the two rivers met. There were a number of causeways across the Witham valley in prehistory, which encouraged ritual activity, including at Fiskerton (Field and Parker Pearson 2003) and at Stamp End (Jones and Stocker 2003), where the famous Iron Age Witham shield and sword were found, having been deposited in the water. There also appears to have been late Iron Age and early Roman activity on what was once an island within the Brayford Pool (Steane 2001; Jones and Stocker 2003), as well as some traces of Iron Age activity beneath the fortress itself (Jones and Stocker 2003: 28–​30). To neglect the importance of this landscape before the fortress was constructed will have an impact on the way in which we understand the town and the urban experience here. The civitas-​capitals are often considered according to whether there was a fort (Exeter, Wroxeter, and Cirencester) or oppidum (Verulamium, Winchester, Silchester) preceding the development of the town, but, in some cases, the forts are not proven (for example, Cirencester), and in no cases is the Iron Age activity yet fully understood. At Leicester there is also growing evidence being documented of Iron Age activity concentrating near the river Soar beneath the later town, and this may have been some kind of oppidum or other settlement, including round-houses and evidence of coin manufacture (Clay and Mellor 1985; Kipling et al. 2007). It is knowledge of activity preceding towns, and the understanding of these places that it brings, that is most likely to change through archaeological investigation, as ongoing excavation and survey work at and around Silchester, Cirencester, Aldborough, London, and other towns are indicating. Work at Caistor-​by-​Norwich, for example, has revealed considerable traces of activity beneath the town, including gullies and ring ditches suggestive of round-houses and other features (Bowden and Bescoby 2008). Structural or activity remains need not be the only indications of the importance of places in the past. Rivers, lakes, woods, and hills were not mundane but integral parts of the cultural landscape, and they will also have had an impact on the nature of urban development and the urban experience (cf. Rogers 2011a, b, 2013). For example, the location of London at the highest point where the river was still tidal allowed the development of important port facilities (Milne 1985). Apart from economic or strategic factors relating to London’s location and development (cf. Perring 1991; Milne 1995; Wacher 1995), however, it is possible to analyse the landscape setting from additional perspectives, which will also have formed an important part of the urban biography. Although there does not appear to have been much settlement here in the late Iron Age, it does not mean that there would not have been cultural values attached to this place, especially in relation to its watery setting and islands situated across the Thames in the vicinity of Southwark (Cowan et al. 2009). There are some traces of late Iron Age activity on the islands at Southwark, and the rivers and streams here appear to have invoked religious activity, which continued during the Roman period, especially noticeable in the Walbrook, which ran through the centre of the later town and the marshy area of the Upper Walbrook valley (Bradley and Gordon 1988; Merrifield 1995; Butler 2006; Merrifield and Hall 2008). Towns could be disruptive, even built over pre-​existing

752

752   Adam Rogers features, but also provide continuities of place by being incorporated into the existing geographies and meanings in the landscape.

Town Construction Processes As well as being placed within landscapes of existing meaning, the foundation processes of the towns themselves will not have been mundane. There are likely to have been foundation rituals drawing on both indigenous and Roman traditions. They may not now be easily recognized in the archaeological record, but excavations in the centre of Roman Dorchester have discovered what appear to have been pits and shafts containing unusual collections of objects and animal remains, which have been interpreted as representing ritualized deposits, perhaps even connected with the foundation of the town itself (Woodward and Woodward 2004) (Figure 36.2). Certainly within the Roman tradition there were rituals associated with urban foundation including taking the auspices and the construction of an enclosure or templum (cf. Rykwert 1976); an interpretative possibility which has also been raised for the evidence of an early enclosure beneath the forum at Verulamium (Creighton 2000: 210). Towns were ritualized places through their foundation, creation, definition of space, and the actions of people within and around them. While the exact nature and interpretation of the pits and shafts at Dorchester are still open to speculation, they remind us of the need to recognize activities that we might not necessarily expect or assume in urban development or urban life in Roman Britain, drawing on both local traditions and those introduced from elsewhere in the empire. This includes the ritual and religious activities that took place in many different contexts outside temples and shrines. It is also likely that there will also have been origin myths drawing on the histories and stories that were attached to the landscapes and places where towns developed and perhaps in some cases were created in order to provide legitimacy for the towns in the landscape. Where towns emerged on the sites of fortresses at Colchester, Lincoln, and Gloucester, there is now good archaeological evidence of the physical processes involved in the conversions of these settlements, including the demolition of fortress structures, the construction of new buildings, and also the reuse of others. Excavations within Gloucester in the late 1960s and 1970s, especially at 13–​17 Berkeley Street, have uncovered remains relating to the reuse of fortress buildings as well as the construction of new half-​ timbered and clay-​walled structures (Hurst 1988, 1999a, b). In some cases it appears that the barrack blocks were demolished but no new buildings were constructed in their place for some time. The main area of the town remained small, moreover, since it continued to lie within the original confines of the fortress, although there are remains of settlement known beyond it (Heighway 1983; Heighway and Bryant 1999). Excavations within Colchester have revealed a similar process of conversion, although here the town did expand beyond the original area of the fortress in a more planned way. Excavations at Lion Walk demonstrated that some of the barrack blocks were converted to other

The Development of Towns    753

Durnovaria

N

River Fr

ome

Forum Site Baths

Limit of excavation

Pits and shafts

Timber fencing Enclosure? Enclosure?

Stone footings 0

10

20

30

m

Figure 36.2  Plan of pits, shafts, and timber-​fenced enclosures excavated within the centre of Roman Dorchester (Durnovaria) at the Greyhound Yard site, 1981–​4. Source: adapted from Woodward and Woodward (2004: figure 1). © A. C. Rogers.

uses, while further buildings or parts of structures were demolished and the land put to other uses, including cultivation (Crummy 1984, 1988, 1992) (Figure 36.3). These towns were clearly not grand and monumental in the Classical style from the outset. The first public buildings at Colchester were constructed in the area of the fortress annexe outside the main fortress, probably because the fortress was too crowded, and this may be the reason why the defences were levelled. This conversion and construction process, however, was not speedy and may have taken a number of years, since excavations indicated that the fortress ditch had been neglected for some time before it was levelled (Crummy 1988: 42); the Temple of Claudius, moreover, may well not have been begun until Claudius’ death in ad 54, although this need not necessarily have been the case (Drury 1984). Like Gloucester, the defences of the fortress were maintained at

754

754   Adam Rogers

River Col

ne

N

Duncan’s Gate

Possible site of gate (St Peter’s Street)

North Gate

THEATRE TEMPLE East Gate

Balkerne Gate

Lion Walk Culver excavations Street excavations South Gate Head Gate

Possible site of gate (Priory Street)

Outline of fortress ditch

0

100

200

300

400 m

Figure 36.3  Plan of the relationship between the fortress and later colonia at Colchester; the walls around the town were constructed in the early second century ad; the theatre and temple were constructed within the fortress annexe. Source: adapted from Crummy (1993: figure 2.9). © A. C. Rogers.

Lincoln, and there was probably a fairly similar process of conversion (M. J. Jones 1988, 2002, 2003b; Steane 2006). What differs here, however, is that there is more evidence known regarding the settlement that developed outside the walls (often referred to as the ‘Lower City’), and it appears to have been more formalized, eventually itself being walled, running down the hill from the fortress to the river Witham (M. J. Jones 2002). The development of the colonia at York was an exception in that the fortress appears to have remained in use throughout the Roman period, and a civilian settlement, or canabae, developed beside it. Across the river Ouse a much larger settlement developed, which was eventually promoted to the status of colonia (Ottaway 1999, 2004). Nothing is yet known about the forum/​basilica complex within the town; one suggestion is that the fortress principia may have had this role, but it may still await discovery, since there have not been many opportunities for excavation in the central area of the town. Some of the civitas-​capitals appear to have had a fairly similar process of development to that of the coloniae. At both Exeter and Wroxeter the towns were founded on the sites of fortresses, and they adapted and expanded the street grids (Bidwell 1979; Henderson 1988; White and Barker 1998). Cirencester differed in there only having been a possible small fort (Leaholme) and little is known about its interior, if it did exist (Wacher

The Development of Towns    755 and McWhirr 1982; McWhirr 1988; Holbrook 1998). The processes that sparked the developments at Canterbury, Leicester, Caerwent, Caistor-​by-​Norwich, Dorchester, Carmarthen, Aldborough, and Brough-​on-​Humber are less well understood. While there is likely to have been a military presence at Carmarthen and Brough-​on-​Humber, more archaeological investigation is needed at these towns (Wacher 1969; James 2003). The possibility of a military base at the other towns is often raised, although there is, so far, no conclusive proof; significant pre-​Roman activity, however, is known at Canterbury, Leicester, and Caistor-​by-​Norwich, which will have formed a significant element of these urban biographies. It has often been the conventional argument in discussing urban development that there was direct involvement from Rome in urban planning and construction, including through the utilization of the military. One of the difficulties is that there are very few inscriptions surviving from towns in Roman Britain to provide information as to who funded and built the urban infrastructure. Frere (1967) emphasized the role of the military and drew on his own excavations at Verulamium, where he identified an early (c. ad 49) rectangular timber-​built row of shops in insula XIV (Frere 1972, 1983). He argued that they were similar in plan and appearance to military barrack blocks, and so they must have been designed by military architects, built by craftsmen lent by the government, and military stockpiles would have been used in their construction. Millett (1990: 70–​7 1), however, argued that the comparison with barrack blocks is only superficial and that there was no evidence that the row was planned and roofed as a single unit. He also emphasized the fact that there will have been sophisticated technology already available in Britain, along with established sources of timber supplies, which would have made such construction possible. This evidence, therefore, cannot be used to support the emphasis on official involvement and assistance in town construction. Creighton’s work (2006) has refocused attention to the evidence and role of phases of activity that pre-​dated ad 43. At Silchester, for example, there is evidence of early elite activity but which was probably influenced by Rome or the near continent. Excavations on the site of the basilica have revealed a number of phases of activity that pre-​date the foundation of the town. The earliest evidence related to traces of at least three roundhouses, dating to around 15 bc–​ad 25, which were then replaced by rectangular structures and metalled streets and apparently surrounded by a timber palisade (Fulford and Timby 2000: 8, 20–​24). What might be regarded as fairly high-​status material came from this area, including coins, copper-​alloy brooches, and toilet instruments. Pottery included terra nigra and terra rubra wares and some early samian ware dating to around ad 30 onwards. There were also large numbers of pig bones and oyster shells (Fulford and Timby 2000: 20-2​4), perhaps indicating special feasting activities. Upon reaching the early levels in the excavations within insula IX of the town, there appears to have been a fairly dense occupation with Iron Age timber rectangular buildings and streets (Fulford 2003, 2011). The elites here may well have had contact with Rome, but their use of the material need not necessarily mean that they considered themselves to have been Romanized. While there are earthworks at Silchester conventionally associated with the oppidum, there remains general uncertainty about their nature and purpose and their

756

756   Adam Rogers relationship with the settlement evidence at the site. It would probably be a mistake to make too many assumptions about their nature and function without further evidence, including dateable material; our understanding of these sites and sequences of development can still clearly be advanced through further work. It could be said that the inner earthwork enclosed an area of around 32 hectares and partly underlies the Roman town, while the outer earthwork is more discontinuous and later in date, perhaps representing the continuation of Iron Age forms of monumentalization and special organization in the Roman period; it may also have been part of further dyke systems to the north-​west of the town (Fulford 1984: 79–​83) (Figure 36.4). At Camulodunum a small fort has been identified at Gosbecks (Hawkes and Crummy 1995), which has been interpreted as having been part of the ad 43 invasion pre-​dating the fortress here or, perhaps, built after the fortress had been converted into a colonia around ad 49 or following the Boudiccan Revolt (Hawkes and Crummy 1995: 101). The interpretation remains difficult, because the fort has not been excavated, but its early plan and position led Creighton (2006: 61–​64) to suggest instead that it may even have been pre-​ad 43 in date and influenced by the Roman army. The local leader may have been organizing his/​her own forces in the Roman style. Alternatively, the fort may have been garrisoned with Roman auxiliaries before Roman annexation. Either way, it is clear that there remain many ambiguities in our understanding of urban origins and construction than conventional narratives have allowed for. It appears that London’s early and large-​scale development may have been largely an exception among the towns. Here it appears that there may have been some military involvement in the development of the site, with the construction of a monumental street system and timber port facilities (Milne 1985; Millett 1994; Brigham 1998). The large size of London compared with other towns and the number of early public buildings also indicates its importance. In much of the literature on Roman Britain, London is known as the provincial capital, but, although there is evidence of Roman officials living here, such as the procurator, there is no direct evidence that it was also the home of the provincial governor; a permanent fort incorporated as part of the town (identified at Cripplegate) dating to the late first or early second century ad, however, might support this claim (Perring 1991; Milne 1995; Shepherd 2012; cf. Millett 1998). Development was slow in most towns, and it is likely that there was little monetary input from Rome; official funding was limited even for the development of the colonia. With even less imperial money going to the civitas-​capitals, it seems most likely that local elites would have been involved in funding the projects. An inscription (RIB 3123) from Verulamium dating the completion of the forum/​basilica to the end of ad 79 may have been an exception among towns in terms of its early date. Documents and inscriptions from other parts of the empire indicate that slaves, prisoners, and the military could all be used for construction projects. Surviving sections of the statute from the town of Urso (modern Osuna) in Spain (the Lex Coloniae Genetivae Juliae), moreover, stipulates that the citizens of the town had to be engaged in five days of construction work a year (Hardy 1912: 7; González and Crawford 1986; Duncan-​Jones 1990: 174), but it is unclear to what extent this applied to other areas, though it was apparently also

The Development of Towns    757 N

‘Ou

ter Ear th

wo rk’

West End Brook

‘In

ne

rE

ar

th

wo

rk ’

Town wall circuit

Silchester 0

Brook

100 200 300 400 m

Figure 36.4  Plan of the relationship between the town wall circuit and enigmatic earthworks at Silchester (Calleva Atrebatum) possibly associated with the earlier oppidum here. Source: Adapted from Fulford (1984: figure 85).

the case in Egypt (Duncan-​Jones 1990: 174). If the military was present in a number of towns for the creation of road networks and other tasks, there is little direct evidence for their involvement, and it is unlikely that they will have remained in towns for long (unless retired).

758

758   Adam Rogers The client ruler Togidubnus has formed a part of the narratives of urban construction of Chichester and also, to a lesser extent, Silchester and Winchester. He is mentioned by Tacitus (Agr. XIV): ‘quaedam civitates Cogidumno regi donatae (is ad nostrum usque memoriam fidissimus mansit) [some of the estates were given to King Cogidubnus, who lived down to our day a most faithful ally]’. The difficulty here is that the discrepancy in the name may in fact indicate different people or that an error may have got into the text; confusingly, also, the inscription of the name is also damaged, with the first letter missing. He appears to have been awarded land, but nothing else is known about his life or career (Todd 2004: 47); there were clearly disruptions in landownership here, though. The dedication slab was found at Chichester in the eighteenth century with what appears to be his name (RIB 91), but there is no evidence that he was also involved in Winchester or Silchester, these suggestions being based mainly on the similar dates in which the towns developed and a number of datable fragments of architectural stonework from Chichester. As well as RIB 91, which is a pre-​Flavian dedication of a temple to Neptune and Minerva in honour of the imperial house built on behalf of Togidubnus and the guild of town craftworkers, there was another inscription (RIB 92) dedicated to Nero, perhaps from a major building or statue, and a dedication to the Matres Domesticae, which may have been another early temple (Fulford 2006: 97). Unfortunately, very little is really known about the early development of the town, and the street grid does not appear to have been laid down until the ad 70s. There may, instead, have been some kind of early monumental complex here (cf. Fulford 2006; Mattingly 2006: 269). Another form of indigenous elite involvement in the physical creation of the town has been argued by Creighton (2006) in his analysis of the relationship between urban spaces and monumental burial chambers. At Camulodumum there was the Lexden tumulus dating to the late first century bc and a larger burial complex at Stanway, just west of Gosbecks, and dating from the late first century bc to around ad 60 (Crummy 1997: 22–2​4; Crummy et al. 2007). At Verulamium there was a rich burial chamber, which lay uphill across the river Ver and dated to around the ad 40s–​50s (Niblett 1999), and there were Iron Age burial mounds outside Cirencester in the Tarbarrow Field, in an area that also appears to have been used for Roman period burials and religious activity (O’Neil and Grinsell 1960; Holbrook 2008; Biddulph and Welsh 2011). The recent survey work at Silchester has also produced traces of what may have been elite burials associated with the oppidum, although as yet no excavation has taken place here, and more work is needed to confirm this. The burials that have been excavated demonstrate knowledge of, and contact with, the Roman Empire, as indicated by the items within the burials; but it must be recognized that they may have been as much about using Roman goods and settlement forms for expressing their own identity as of adopting Roman ways. Creighton (2006) has suggested that perhaps even the urban planning and establishment of street grids acted as a form of commemoration associated with the elites that succeeded in achieving power. Within a number of towns, including the colonia, many of the earliest public buildings were in fact constructed in timber rather than stone. At Lincoln, for example, excavations on the site of the fortress principia have demonstrated that, once that it was

The Development of Towns    759 demolished, the stone forum/​basilica complex did not appear until the second century, following the construction of an earlier timber building (Steane 2006). An early timber forum/​basilica complex has also been identified at Silchester (Fulford and Timby 2000), as well as a theatre at Canterbury (Bennett 1988) and amphitheatres at Cirencester (Holbrook 1998) and London (Bateman et al. 2008). There were also many timber domestic structures in the early phases of towns, often in the form of ‘strip’ buildings, as at London (e.g. Hill and Rowsome 2011). At London, however, a number of sites in the western area of the town have also provided evidence of round-houses, perhaps indicating the continuation of local building traditions as people moved into the town from the surrounding area (Perring and Roskams 1991; Hill and Rowsome 2011; Perring 2011). The occupants of these early round-houses may even have come into the town in order to work as labourers during its construction. It would be overly simplistic to interpret these structures as inferior, or of lower status, to rectangular buildings and necessarily representing the houses of poor labourers or slaves. These round-houses formed part of a long tradition of construction, which was fully integrated into human expressions of being and identity; the structures and spaces within them formed a microcosm of the way in which their world was organized, experienced, and understood (cf. Harding 2009; Sharples 2010). It was, of course, not just monumental buildings that formed these towns, moreover, but the people using and interacting in these spaces. These early phases of town development formed significant parts of the biography of these places and the creation of space. It is important not to apply modern expectations and perceived Roman values to our treatment of these sites, many of which draw on Iron Age traditions. That it could be argued that towns never reached a high level of monumentality in Roman Britain may also relate to a continuation of pre-​existing ways of expressing wealth and using space. Local traditions in building and organizing labour and the significance behind using and manipulating different materials may well also have continued to be important, even if they were now building different styles of structures. Sharples (2010: 116–​124), for example, has argued that the construction of Iron Age oppida, hillforts, and other monuments would have been important cultural events, with people encouraged to come together and contribute through motives relating to social interaction, feastings, and rituals. Construction processes and methods themselves could also be highly symbolic, often involving the deliberate choice of the use of specific materials, sometimes brought from long distances, to incorporate into the structures, as seen at hillforts and other prehistoric monuments (Miles et al. 2003; Sharples 2010: 116–​124). Building construction had more than simply practical and economic meaning, since it involved the act of creation of both physical structures and new spaces and could reflect the identities and expressions of those carrying out the work (cf. Gardner 2007; Revell 2009). The construction processes would have formed an important part of the biography of the building and just as significant a part of its meaning as the finished structure itself (cf. McFadyen 2006). Towns can be seen as major construction events, and they also transformed the land through moving earth, rock, and water in terracing, land drainage, reclamation, and other activities—​as identified through excavation at London (e.g. Hill and Rowsome 2011; Leary and Butler 2012), Winchester

760

760   Adam Rogers (e.g. Ford and Teague 2011), and other towns (Rogers 2012). These actions altered places and formed an important part of the urban reality with an impact on the way in which these landscapes were used and experienced.

Conclusion It is becoming increasingly difficult to write about the development of towns in Roman Britain, because no single narrative is possible. Though there were clearly processes involved relating to the expansion of Roman control and military strategy, emphasis also needs to be placed on the importance of each individual urban biography starting before the Roman conquest and the different people involved in and experiencing each town development, from local people to incomers and the military. More nuanced accounts of each town are required, drawing on changing understandings of late Iron Age and Roman period activity at both local and regional levels. It is this that makes urban development such a challenging and important subject for understanding Roman Britain.

Abbreviations AE    L’Année Epigraphique. Paris: Ernest Leroux/​Presses Universitaires des France, 1888–​to date). RIB  R  . G. Collingwood and R. P. Wright, Roman Inscriptions of Britain. Oxford: Clarendon Press, 1965–​to date.

References Primary Sources Aurelius Victor, Caesares, trans. H. W. Bird. Liverpool: Liverpool University Press, 1994. Caesar, Bellum Gallicum, trans. H. J. Edwards. London: Heinemann, 1917. Cassius Dio, Roman History, trans. E. Cary. London: Heinemann, 1961. Ptolemy, Geographia, trans. E. L. Stevenson. Mineola, NY: Dover, 1991. Tacitus. Agricola, trans. M. Hutton. London: Heinemann, 1970. Tacitus, Annales, trans. J. Jackson. London: Heinemann, 1937.

Secondary Sources Bateman, N., Cowan, C., and Wroe-​Brown, R. (2008). London’s Roman Amphitheatre: Guildhall Yard, City of London. MoLAS Monograph 35. London: Museum of London Archaeology Service. Bennett, P. (1988). ‘No. 25 Watling Street’, Archaeologia Cantiana, 106: 129–​130.

The Development of Towns    761 Biddulph, E., and Welsh, K. (2011). Cirencester before Corinium: Excavations at Kingshill North, Cirencester, Gloucestershire. Oxford: Oxford Archaeology Ltd. Bidwell, P. T. (1979). The Legionary Bath-​House and Basilica and Forum at Exeter. Exeter Archaeology Reports Volume I. Exeter: Exeter City Council and the University of Exeter. Boatwright, M. L. (2000). Hadrian and the Cities of the Roman Empire. Princeton: Princeton University Press. Bowden, W., and Bescoby, D. (2008). ‘The Plan of Venta Icenorum (Caistor-​by-​ Norwich): Interpreting a New Geophysical Survey’, Journal of Roman Archaeology, 21: 324–​334. Bradley, R., and Gordon, K. (1988). ‘Human Skulls from the River Thames, their Dating and Significance’, Antiquity, 62: 503–​509. Brigham, T. (1998). ‘The Port of Roman London’, in B. Watson (ed.), Roman London: Recent Archaeological Work. Journal of Roman Archaeology Supplementary Series 24. Portsmouth, RI: Journal of Roman Archaeology, 23–​34. Brooks, H. (2006). ‘Colchester before Colchester’, in P. Ottaway (ed.), A Victory Celebration: Papers on the Archaeology of Colchester and Late Iron Age–​Roman Britain presented to Philip Crummy. Colchester: Colchester Archaeological Trust, 5–​10. Burnham, B., and Wacher, J. (1990). The Small Towns of Roman Britain. London: Batsford. Burnham, B., Collis, J., Dobinson, C., Haselgrove, C., and Jones, M. (2001). ‘Themes for Urban Research, c.100 bc to ad 200’, in S. James and M. Millett (eds), Britons and Romans: Advancing an Archaeological Agenda. CBA Research Report 125. York: Council for British Archaeology, 67–​76. Butler, J. (2006). Reclaiming the Marsh:  Archaeological Excavations at Moor House, City of London. Pre-​ Construct Archaeology Monograph 6.  London:  Pre-​ Construct Archaeology Ltd. Clay, P., and Mellor, J. E. (1985). Excavations in Bath Lane, Leicester. Leicestershire Museums Archaeological Report 10. Leicester: Leicester Museums. Collingwood, R. G., and Richmond, I. (1969). The Archaeology of Roman Britain. London: Methuen. Copeland, T. (2011). Roman Gloucestershire. Stroud: The History Press. Cowan, C., Seeley, F., Wardle, A., Westman, A., and Wheeler, L. (2009). Roman Southwark Settlement and Economy:  Excavations in Southwark 1973–​ 91. MoLAS Monograph 42. London: Museum of London Archaeology Service. Creighton, J. (2000). Coins and Power in Late Iron Age Britain. Cambridge:  Cambridge University Press. Creighton, J. (2006). Britannia: The Creation of a Roman Province. London: Routledge. Crummy, P. (1984). Excavations at Lion Walk, Balkerne Lane and Middleborough, Colchester, Essex. Colchester Archaeology Report 3. Colchester: Colchester Archaeological Trust. Crummy, P. (1988). ‘Colchester (Camulodunum/​ Colonia Victriciensis)’, in G. Webster (ed.), Fortress into City:  The Consolidation of Roman Britain First Century ad. London: Batsford, 24–​47. Crummy, P. (1992). Excavations at Culver Street, the Gilberd School and Other Sites in Colchester 1971–​85. Colchester Archaeology Report 6. Colchester: Colchester Archaeological Trust. Crummy, P. (1997). City of Victory:  The Story of Colchester—​Britain’s First Roman Town. Colchester: Colchester Archaeological Trust. Crummy, P., Benfield, S., Crummy, N., Rigby, V., and Shimmin, D. (2007). Stanway: An Élite Burial Site at Camulodunum. Britannia Monograph Series 24. London: Society for the Promotion of Roman Studies.

762

762   Adam Rogers Drury, P. J. (1984). ‘The Temple of Claudius at Colchester Reconsidered’, Britannia, 15: 7–​50. Duncan-​Jones, R. (1990). Structure and Scale in the Roman Economy. Cambridge: Cambridge University Press. Esmonde Cleary, A. S. (1987). Extra-​Mural Areas of Romano-​British Towns. BAR British Series 169. Oxford: Archaeopress. Field, N., and Parker Pearson, M. (2003). Fiskerton: An Iron Age Timber Causeway with Iron Age and Roman Votive Offerings. Oxford: Oxbow Books. Fincham, G. (2004). Durobrivae: A Roman Town between Fenland and Upland. Stroud: Tempus. Ford, B. M., and Teague, S. [with Biddulph, E., Hardy, A., and Brown, L.] (2011). Winchester—​ a City in the Making:  Archaeological Excavations between 2002 and 2007 on the Sites of Northgate House, Staple Gardens and the former Winchester Library, Jewry St. Oxford: Oxford Archaeology Ltd. Frere, S. S. (1967). Britannia: A History of Roman Britain. London: Routledge and Kegan Paul. Frere, S. S. (1972). Verulamium Excavations, Volume I. Reports of the Research Committee of the Society of Antiquaries of London 28. London: The Society of Antiquaries of London. Frere, S. S. (1983). Verulamium Excavations, Volume II. Reports of the Research Committee of the Society of Antiquaries of London 41. London: The Society of Antiquaries of London. Frere, S. S. (1987). Britannia: A History of Roman Britain. 3rd edn. London: Routledge and Kegan Paul. Fulford, M. (1984). Silchester Defences 1974–​80. Britannia Monograph Series 5. London: Society for the Promotion of Roman Studies. Fulford, M. (2003). ‘Julio-​ Claudian and Early Flavian Calleva’, in P. Wilson (ed.), The Archaeology of Roman Towns: Studies in Honour of John S. Wacher. Oxford: Oxbow Books, 95–​104. Fulford, M. (2006). ‘Corvées and civitates’, in R. J.  A. Wilson (ed.), Romanitas:  Essays on Roman Archaeology in Honour of Sheppard Frere on the Occasion of his Ninetieth Birthday. Oxford: Oxbow Books, 65–​7 1. Fulford, M. (2011). ‘“Calleva: Silchester in the Iron Age”: From Julius Caesar to Claudius: The Late Iron Age/​Roman Transition’. Unpublished conference paper, University of Leicester, November. Fulford, M., and Timby, J. (2000). Late Iron Age and Roman Silchester: Excavations on the Site of the Forum-​Basilica 1977, 1980–​86. Britannia Monograph Series 15. London: Society for the Promotion of Roman Studies. Gardner, A. (2007). Archaeology of Identity:  Soldiers and Society in Later Roman Britain. Walnut Creek, CA: Left Coast Press. González, J., and Crawford, M. H. (1986). ‘The Lex Irnitana:  A  New Copy of the Flavian Municipal Law’, Journal of Roman Studies, 76: 147–​243. Harding, D. W. (2009). The Iron Age Round-​House: Later Prehistoric Building in Britain and Beyond. Oxford: Oxford University Press. Hardy, E. G. (1912). Roman Laws and Charters. Oxford: Clarendon Press. Haselgrove, C. (1999). ‘The Iron Age’, in J. Hunter and I. Ralston (eds), The Archaeology of Britain. London and New York: Routledge, 113–​134. Haselgrove, C. (2004). ‘Society and Polity in Late Iron Age Britain’, in M. Todd (ed.), A Companion to Roman Britain. Oxford: Blackwell Publishing, 12–​29. Haselgrove, C., and Moore, T. (2007). ‘New Narratives of the Later Iron Age’, in C. Haselgrove and T. Moore (eds), The Later Iron Age in Britain and Beyond. Oxford: Oxbow Books, 1–​15.

The Development of Towns    763 Haverfield, F. (1912). The Romanization of Roman Britain. Oxford: Clarendon Press. Hawkes, C. F. C., and Crummy, P. (1995). Camulodunum 2. Colchester Archaeology Report 11. Colchester: Colchester Archaeological Trust. Heighway, C. (1983). The East and North Gates of Gloucester. Gloucester:  Western Archaeological Trust. Heighway, C., and Bryant, R. (1999). The Golden Minster: The Anglo-​Saxon Minster and Later Medieval Priory of St Oswald at Gloucester. CBA Research Report 117. York:  Council for British Archaeology. Henderson, C. (1988). ‘Exeter (Isca Dumnoniorum)’, in G. Webster (ed.), Fortress into City: The Consolidation of Roman Britain in the First Century ad. London: Batsford, 91–​119. Hill, J., and Rowsome, P. (2011). Roman London and the Walbrook Stream Crossing: Excavations at 1 Poultry and Vicinity, City of London. MoLAS Monograph 37. London:  Museum of London Archaeology Service. Hingley, R. (1997). ‘Resistance and Domination:  Social Change in Roman Britain’, in D. J. Mattingly (ed.), Dialogues in Roman Imperialism:  Power, Discourse and Discrepant Experiences in the Roman Empire. Journal of Roman Archaeology Supplementary Series 23. Portsmouth, RI: Journal of Roman Archaeology, 145–​166. Hingley, R. (2000). Roman Officers and English Gentlemen. London and New York: Routledge. Hingley, R. (2005). Globalizing Roman Culture:  Unity, Diversity and Empire. London and New York: Routledge. Hingley, R. (2008). The Recovery of Roman Britain 1586 to 1906:  A  Colony so Fertile. Oxford: Oxford University Press. Holbrook, N. (1998). Cirencester:  The Roman Town Defences, Public-​Buildings and Shops. Cirencester: Cotswold Archaeological Trust. Holbrook, N. (2008). ‘Cirencester and the Cotswolds: The Early Roman Evolution of a Town and Rural Landscape’, Journal of Roman Archaeology, 21: 305–​323. Hurst, H. (1988). ‘Gloucester (Glevum)’, in G. Webster (ed.), Fortress into City: The Consolidation of Roman Britain in the First Century ad. London: Batsford, 24–​69. Hurst, H. (1999a). ‘Topography and Identity in Glevum colonia’, in H. Hurst (ed.), The Coloniae of Roman Britain: New Studies and a Review. Papers of the Conference held at Gloucester on 5–​6 July, 1997. Journal of Roman Archaeology Supplementary Series 36. Portsmouth, RI: Journal of Roman Archaeology, 113–​135. Hurst, H. (1999b). ‘Civic Space at Glevum, in H. Hurst (ed.), The Coloniae of Roman Britain: New Studies and a Review. Papers of the Conference held at Gloucester on 5–​6 July, 1997. Journal of Roman Archaeology Supplementary Series 36. Portsmouth, RI: Journal of Roman Archaeology, 152–​160. James, H. (2003). Roman Carmarthen. London: Society for the Promotion of Roman Studies. Jones, M. J. (1988). ‘Lincoln (Lindum)’, in G. Webster (ed.), Fortress into City: The Consolidation of Roman Britain in the First Century ad. London: Batsford, 145–​166. Jones, M. J. (2002). Roman Lincoln: Conquest, Colonia and Capital. Stroud: Tempus. Jones, M. J. (2003a). ‘The Roman Military Era’, in D. Stocker (ed.), The City by the Pool: Assessing the Archaeology of the City of Lincoln. Oxford: Oxbow Books, 36–​53. Jones, M. J. (2003b). ‘The colonia Era: Archaeological Account’, in D. Stocker (ed.), The City by the Pool: Assessing the Archaeology of the City of Lincoln. Oxford: Oxbow Books, 56–​138. Jones, M. J. (2004). ‘Cities and Urban Life’, in M. Todd (ed.). A Companion to Roman Britain. Oxford: Blackwell Publishing, 162–​192.

764

764   Adam Rogers Jones, M. J., and Stocker, D. (2003). ‘Settlement in the Lincoln Area in the Prehistoric Era:  Archaeological Account’, in D. Stocker (ed.), The City by the Pool:  Assessing the Archaeology of the City of Lincoln. Oxford: Oxbow Books, 19–​33. Kipling, R., Parker, D., and Cooper, L. (2007). ‘Leicester: Bath Lane, Merlin Works’, University of Leicester Archaeological Services News (November), 5–​6. Laurence, R. (2001). ‘The Creation of Geography: An Interpretation of Roman Britain’, in C. Adams and R. Laurence (eds), Travel and Geography in the Roman Empire. London and New York: Routledge, 67–​94. Leach, P. (1982). Ilchester:  Volume 1 Excavations 1974–​ 1975. Bristol:  Western Archaeological Trust. Leary, J., and Butler, J. (2012). Roman Archaeology in the Upper Reaches of the Walbrook Valley:  Excavations at 6–​8 Tokenhouse Yard, London EC2. Pre-​Construct Archaeology Monograph 14. London: Pre-​Construct Archaeology Ltd. McFadyen, L. (2006). ‘Building Technologies, Quick Architecture and Early Neolithic Long Barrow Sites in Southern Britain’, Archaeological Review from Cambridge, 21/​1: 117–​134. McWhirr, A. (1988). ‘Cirencester (Corinium Dobunnorum)’, in G. Webster (ed.), Fortress into City: The Consolidation of Roman Britain in the First Century ad. London: Batsford, 74–​90. Mann, J. C. (1996). Britain and the Roman Empire. Aldershot: Variorum. Mattingly, D. (2004). ‘Being Roman: Expressing Identity in a Provincial Setting’, Journal of Roman Archaeology, 17: 5–​25. Mattingly, D. (2006). An Imperial Procession: Britain in the Roman Empire. London: Penguin. Merrifield, R. (1995). ‘Roman Metalwork from the Walbrook: Rubbish, Ritual or Redundancy?’, Transactions of the London and Middlesex Archaeological Society, 46: 27–​44. Merrifield, R., and Hall, J. (2008). ‘In its Depths what Treasures: The Nature of the Walbrook Stream Valley and the Roman Metalwork Found Therein’, in J. Clark, J. Cotton, J. Hall, R. Sherris, and H. Swain (eds), Londinium and Beyond: Essays on Roman London and its Hinterland for Harvey Sheldon. York: CBA Research Report 156. York: Council for British Archaeology, 121–​127. Miles, D., Palmer, S., Lock, G., Gosden, C., and Cromarty, A. M. (2003). Uffington White Horse and its Landscape:  Investigations at White Horse Hill Uffington, 1989–​95 and Tower Hill Ashbury, 1993–​4. Oxford: Oxford Archaeological Unit. Millett, M. (1990). The Romanization of Britain:  An Essay in Archaeological Interpretation. Cambridge: Cambridge University Press. Millett, M. (1994). ‘Evaluating Roman London’, Archaeological Journal, 151: 427–​435. Millett, M. (1998). ‘Introduction: London as Capital?’, in B. Watson (ed.), Roman London: Recent Archaeological Work. Journal of Roman Archaeology Supplementary Series 24. Portsmouth, RI: Journal of Roman Archaeology, 7–​12. Milne, G. (1985). The Port of Roman London. London: Batsford. Milne, G. (1995). Roman London. London: Batsford/​English Heritage. Moore, T. (2006). Iron Age Societies in the Severn-​Cotswolds: Developing Narratives of Social and Landscape Change. BAR British Series 421. Oxford: Archaeopress. Moore, T. (2011). ‘Detribalizing the Later Prehistoric Past: Concepts of Tribes in Iron Age and Roman Studies’, Journal of Social Archaeology, 11: 334–​360. Niblett, R. (1985). Sheepen: An Early Roman Industrial Site at Camulodunum. CBA Research Report 57. York: Council for British Archaeology. Niblett, R. (1999). The Excavation of a Ceremonial Site at Folly Lane, Verulamium. Britannia Monograph Series 14. London: Society for the Promotion of Roman Studies. Niblett, R. (2001). Verulamium: The Roman City of St Albans. Stroud: Tempus.

The Development of Towns    765 O’Neil, H., and Grinsell, L. V. (1960). ‘Gloucestershire Barrows’, Transactions of the Bristol and Gloucestershire Archaeological Society, 79: 5–​149. Ottaway, P. (1999). ‘York: The Study of a Late Roman colonia’, in H. Hurst (ed.), The Coloniae of Roman Britain: New Studies and a Review. Papers of the Conference held at Gloucester on 5–​6 July, 1997. Journal of Roman Archaeology Supplementary Series 36. Portsmouth, RI: Journal of Roman Archaeology, 136–​150. Ottaway, P. (2004). Roman York. Stroud: Tempus. Perring, D. (1991). Roman London. London: Seaby. Perring, D. (2011). ‘Two Studies on Roman London’, Journal of Roman Archaeology, 24: 249–​282. Perring, D., and Roskams, S. [with Allen, P.] (1991). Early Development of Roman London West of the Walbrook: The Archaeology of Roman London, Volume 2. CBA Research Report 70. London: Council for British Archaeology. Potter, T. W. (2002). ‘The Transformation of Britain from 55 bc to ad 61’, in P. Salway (ed.), The Roman Era: The British Isles: 55 bc–​ad 410. Oxford: Oxford University Press, 11–​36. Putnam, B. (2007). Roman Dorset. Stroud: Tempus. Revell, L. (2009). Roman Imperialism and Local Identities. Cambridge:  Cambridge University Press. Richmond, I. A. (1963). Roman Britain. London: Jonathan Cape. Rivet, A. L. F., and Smith, C. (1979). The Place-​Names of Roman Britain. London: Batsford. Rogers, A. (2008). ‘Religious Place and its Interaction with Urbanization in the Roman Era’, Journal of Social Archaeology, 8/​1: 37–​62. Rogers, A. (2011a). ‘Reimagining Roman Ports and Harbours: The Port of Roman London and Waterfront Archaeology’, Oxford Journal of Archaeology, 30/​2: 207–​225. Rogers, A. (2011b). Late Roman Towns in Britain:  Rethinking Change and Decline. Cambridge: Cambridge University Press. Rogers, A. (2012). ‘Water and the Urban Fabric: A Study of Towns and Waterscapes in the Roman Period’, International Journal of Nautical Archaeology, 41/​2: 327–​339. Rogers, A. (2013). Water and Roman Urbanism: Towns, Waterscapes, Land Transformation and Experience in Roman Britain. Leiden: Brill. Rykwert, J. (1976). The Idea of a Town: The Anthropology of Urban Form in Rome, Italy and the Ancient World. London: Faber and Faber. Sharples, N. (2010). Social Relations in Later Prehistory: Wessex in the First Millennium bc. Oxford: Oxford University Press. Shepherd, J. (2012). Discovery of the Roman Fort at Cripplegate, City of London: Excavations by W. F. Grimes 1947–​68. London: Museum of London Archaeology. Steane, K. [with Darling, M. J., Mann, J. Vince, A., and Young, J.] (2001). The Archaeology of Wigford and the Brayford Pool. Oxford: Oxbow Books. Steane, K. (2006). The Archaeology of the Upper City and Adjacent Suburbs. Lincoln Archaeology Studies 3. Oxford: Oxbow Books. Todd, M. (2004). ‘The Claudian Conquest and its Consequences’, in M. Todd (ed.), A Companion to Roman Britain. Oxford: Blackwell Publishing, 42–​59. Tomlin, R. S. O. (2006). ‘Was Roman London ever a colonia? The Written Evidence’, in R. J. A. Wilson (ed.), Romanitas: Essays on Roman Archaeology in Honour of Sheppard Frere on the Occasion of his Ninetieth Birthday. Oxford: Oxbow Books, 49–​64. Wacher, J. S. (1969). Excavations at Brough-​on-​Humber, 1958–​1961. Reports of the Research Committee of the Society of Antiquaries of London 25. London: The Society of Antiquaries of London.

766

766   Adam Rogers Wacher, J. S. (1975). The Towns of Roman Britain. London: Batsford. Wacher, J. S. (1995). The Towns of Roman Britain. 2nd edn. London: Batsford. Wacher, J. S., and McWhirr, A. (1982). Early Roman Occupation at Cirencester. Cirencester Excavations 1. Cirencester: Cirencester Excavations Committee. Webster, G. (1966). ‘Fort and Town in Early Roman Britain’, in J. S. Wacher (ed.), The Civitas Capitals of Roman Britain. Leicester: Leicester University Press, 31–​45. Webster, J. (2001). ‘Creolizing the Roman Provinces’, American Journal of Archaeology, 105: 209–​225. White, R., and Barker, P. (1998). Wroxeter: Life and Death of a Roman City. Stroud: Tempus. Williams, J. (2005). ‘“The Newer Rite Is Here”: Vinous Symbolism on British Iron Age Coins’, in C. Haselgrove and D. Wigg-​Wolf (eds), Iron Age Coinage and Ritual Practices. Mainz am Rhein: Van Zabern, 25–​42. Woodward, P., and Woodward, A. (2004). ‘Dedicating the Town: Urban Foundation Deposits in Roman Britain’, World Archaeology, 36: 68–​86.

Chapter 37

Urban Monum e nta l i t y i n Roman Bri ta i n Louise Revell

Introduction One of the most conspicuous changes to the landscape of Britain after the Roman conquest was the appearance of towns. Constructed from stone, standing two or more storeys high, the public buildings of these towns bore witness to a new form of monumentality. More than this, they announced a new ideology of power, not only of the distant authority of Rome, but also of the British elite. The construction of towns began shortly after the conquest, and, with evidence for rebuilding through to the fourth century, these towns proved a long-​term landmark in the lives of the provincial populations. In recent years, work on Roman urbanism has moved away from an aesthetic approach to a more spatial enquiry, the so-​called spatial turn. This has not been confined to the Roman period; drawing on geography, anthropology, and sociology, among others, the archaeology of monuments and buildings has focused more on the concept of inhabited spaces. Furthermore, it is this act of dwelling and the social practices which these spaces allow, which form part of the short-​term and long-​term processes of social reproduction. Urban buildings are not only symbols of power, but part of the process of the creation and maintenance of varying relationships of power. These ideas have forced us to rethink the Roman townscape as a whole, and how people inhabited it or moved through it (MacDonald 1986; Zanker 1998; Kaiser 2000; Revell 2009; Laurence and Newsome 2011). Concurrently there has been a recognition of the variability in the urban experience within the empire, such as Gros’ monumental survey (1996) of public architecture, or Laurence, Esmonde C ​ leary, and Sears’ examination (2011) of the development of the urban landscape within Italy and the western provinces. One strand of these new approaches has been the role towns played in the maintenance of an idea of Roman-​ness, linking the peoples of the provinces into wider ideologies at play within the empire, and creating a sense of belonging in the empire. A central idea

768

768   Louise Revell in the work of Laurence, Esmonde ​Cleary, and Sears is that one role of the Roman town was the production of citizens. While the idea of Roman citizenship as such is less relevant within the western provinces, and Roman Britain in particular, I have argued elsewhere that the daily experience of inhabiting the Roman town created a particular form of Roman ideology of urbanism among the provincial communities which tied them into a wider Roman identity (Revell 1999, 2009). The analysis of the towns of Roman Britain has played a limited role in these new approaches to the Roman townscape. Often viewed as a failed experiment, the British towns have been seen as a poor relation of their continental cousins. Debates have focused on their origins (military or civilian) and how early they can be said to have declined (if they ever had a high point). The character of the mature Romano-​British town has been of little interest. In spite of the numbers of books focusing on individual towns (e.g. Niblett 2001; Jones 2002; White 2002), syntheses are few and far between. Wacher’s monumental survey (1995) of the towns of Roman Britain has not been reproduced, although there have been certain works which have focused on particular questions such as the origins of towns (Creighton 2006) or the significance of water in the town (Rogers 2013). This chapter goes some way to addressing the lack of application of these new strands of investigation to Roman towns. In particular, it concentrates on the monumental buildings of the town, the kinds of activities they fostered, and how far they were implicated in the re-​creation of a particular Roman ideology of Roman urbanism. It focuses on the larger chartered towns rather than including the so-​called small towns, as the latter generally lack the range of monumental architecture symbolic of Roman urbanism and Roman power.

Space for Politics and Administration The first form of activity to consider is political activity and the role of the forum (this material is discussed in greater detail in Revell 2007a). Within the Roman system of imperialism from the time of Augustus until late antiquity, the town was the means of organizing the provinces and the focus for provincial administration through activities such as the collection of taxes, the administration of justice, and the taking of the census. Through the institutions of the town council and local magistrates, the town was central to the more localized administration of the empire. The forum was the symbolic heart of the Roman town, and the stage for the enactment of the political rituals which held the town together. This was the case in ancient Rome, and all our evidence suggests that this was similarly the case for the provinces. The importance of the forum to the urban experience can be demonstrated through its location at the heart of the street grid. At the newly founded town of Caerwent, which arguably most conforms to the stereotype of the Roman street grid, the forum lay at the point where the two major roads through the town intersected. This centrality is also evident in towns which arose out of Iron Age settlements, such as Silchester. At

Urban Monumentality in Roman Britain    769 Verulamium, where there was a more complex street grid, the forum was given added significance by being constructed on the site of the Central Enclosure, which had been either a royal enclosure, or some form of ceremonial or religious precinct (Niblett 2001: 42–​43). The size of the fora would have meant they dominated the townscape at key locations within passages of movement. The Romano-​British forum generally took the form of a large open courtyard, enclosed by porticoes on three sides and a basilica running the full length of the fourth. The layout and decoration of the basilica emphasized its association with political activities, as opposed to other possible functions. The raised platform of the tribunal was the site of magisterial authority and the application of the law throughout the empire, and these were located within the basilica, at one or both ends of the nave. Architectural features distinguished them from the rest of the nave, such as pillars to the sides, or different styles of decoration. Within the second-​century basilica at London, there seems to have been a single tribunal located in a semi-​circular apse at the eastern end of the nave, set apart from the main part by some form of antechamber (Milne 1992). The area of the antechamber was decorated with at least four phases of multi-​coloured wall-​paintings, including one with floral and possibly figural motifs. The apse itself may have been lined with marble wall plaques. At Silchester, there was a tribunal at either end of the basilica, and, following the second-​century rebuilding of the basilica, the tribunals were set within square apses, the northern one faced with thin slabs of Purbeck marble (Fulford and Timby 2000). At Cirencester, a tribunal has been uncovered in the west end of the basilica, set within a large semi-​circular apse, and decorated with imported marbles and architectural decoration (Martin 1898). At Caerwent, the tribunal was accentuated by engaged columns and more than one phase of wall-​painting (Brewer 1993). The second area for political activity was the curia or council chamber; however, these are difficult to identify from the archaeological evidence alone. Balty (1991) uses as criteria the identification of benches, architectural form, position, and internal decoration. From these criteria, the curia at Caerwent can be plausibly identified (Ashby 1906; Frere 1989: 274). Here, one room showed evidence of benches along two opposite walls, a small platform between them, elaborate wall-​painting, and a later mosaic. Others have proved more controversial (Figure 37.1). At Verulamium, it is likely that one of the three adjoining buildings was a curia, but various commentators have argued for each one of the three, all with plausible supporting arguments. At Caistor-​by-​Norwich, Frere (1971) argued that the curia lay to the south end of the basilica nave, based on its position and size (9.1m × 9.14m). The other alternative is room 2, which lay behind the basilica, with an anteroom separating it from the bustle, as at Caerwent, and a hypocaust adding to its comfort and status. Other political requirements, such as an archive room and offices for the town clerks, were met through the other rooms within the rear range of the basilica, or within the porticoes surrounding the forum courtyard. Elsewhere in the empire, the forum was used as a venue for the display of inscriptions and statues of both leading local magistrates and also the emperor and imperial family (Lefebvre 2004; Revell 2013). These are usually argued to have been absent from the fora of Britain, but a review of the evidence suggests this may not be the case.

770

770   Louise Revell

Curia

0

5

10 m

Figure 37.1  Plan of Caerwent forum, with location of probable curia marked. Source: adapted from Brewer (2006: 39).

The forum at Silchester was decorated with a series of statues and inscriptions (Figure 37.2). In the northern apse of the basilica, traces of statue bases were found, and nearby a fragment of a larger-​than-​life statue of a military commander, almost certainly an emperor (Joyce 1873; Fox and Hope 1893; Boon 1973; Cunliffe and Fulford 1982; Isserlin 1998; Fulford and Timby 2000). A bronze eagle was found in the room adjoining the southern tribunal, a stone head, argued to represent the guardian deity of the town, was found in the central apsed room, and there were another four fragments of human figures without firm location, at least one of which may also be of an emperor. Further statue bases stood in the northern range and eastern ambulatory of the forum. A number of inscriptions probably stood in the basilica and the ambulatories. The majority

Urban Monumentality in Roman Britain    771

Approximate locations of: Statues Possible statues Inscriptions Possible inscriptions

0

25 m

Figure  37.2  Plan of Silchester forum with known and approximate locations of inscriptions and statues indicated. Source: P. Copeland; adapted from Isserlin (1998: figure 9.1) with additions.

are fragmentary, but one contains the phrase ob honorem (RIB 67), suggesting it was in honour of a magistrate or other leading figure. Two more were monumental in size, forming part of the fabric of the building. No other forum has quite the wealth of evidence, but most have the remains of the podia for statues, or fragments of bronze statues likely to be of emperors. At Cirencester, the eye from a bronze statue was found within the tribunal apse (Martin 1898), and fragments of an inscription found unstratified on the site refer to the Res Publica (RIB 114). At Gloucester, a statue base approximately 4.1m × 3m was found in front of the basilica with fragments from a bronze statue around it (Hurst 1972; Henig 1993: no. 177). From Verulamium and Wroxeter, we have the inscriptions for the fora, dedicating them to the reigning emperor (RIB 3123, RIB 288, respectively). At the majority of the fora in Britain which have seen systematic excavation, there is evidence, however fragmentary, for their decoration with statues and inscriptions.

772

772   Louise Revell While the forum–​basilica in Britain clearly provided an appropriate space for carrying out specific political activities, the lack of temples within them is a notable feature. The exclusion of a temple from the forum is not unique to Britain, and is also seen in the North African provinces (Euzennat and Hallier 1986). One problem with understanding how the British examples fit into the wider history of the development of the forum is that we lack some of the earliest structures built, in particular that at Cirencester, while the forum at Colchester is somewhat problematic and will be discussed later in the chapter. Nevertheless, the earliest fora do show more evidence for incorporating temples. The forum–​basilica at Verulamium, dedicated in ad 79, in its first phase had a single building opposite the basilica, possibly a temple (RIB 3123; Niblett and Thompson 2005: 78–​83). In the second phase, probably mid-​second century, this was flanked by two further buildings, potentially although not definitely temples. Less problematic is the Flavian forum from London, which was constructed with a small classical temple, 20m × 10m, immediately to the west (Milne 1992). Most tentative is the argument for a temple as part of the Flavian timber basilica at Silchester (Fulford and Timby 2000; Millett 2001). However, in the second-​century rebuild, the shrine was moved into the basilica. Nevertheless, even without the evidence for free-​standing temples, it would be erroneous to see the fora of Britain as completely devoid of religious activity. In reality, most examples seem to have an internal shrine opening onto the nave of the basilica, in some cases as part of the rear range of rooms. Their identification as shrines generally rests on the fact that they were open to the rest of the basilica, and lay on the central axis of the forum–​basilica. This is the case at the rebuilt forum at Silchester, where the central room of the rear range was open, and raised up above the level of the rest of the nave (Fulford and Timby 2000). It was decorated with thin slabs of white marble, and the original excavator commented that it appeared more richly decorated than the rest of the building. This was the room in which the tutela head was discovered, adding weight to the identification of the room as a shrine. A raised, open room on the dominant axis is also present in the fora at Caistor-​by-​Norwich and Caerwent, again suggesting that these were shrines. The dedication of such shrines is uncertain, although they are often assumed to be associated with the imperial cult through comparison with others elsewhere in the empire. However, this obscures the fact that very few shrines or temples to the imperial cult have been unambiguously identified in the west, and creates an element of circularity to their identification (Alföldy 1996; Wallace-​Hadrill 2011). The forum had many purposes: economic, religious, social, but arguably the most important was the political. It was the product of a new political organization, in which the pre-​Roman system of social organization and elite status was replaced by one based upon the system of magistracies, with power distributed among a wider group, and supreme power held for a restricted period. This was the ideal of Roman politics, and it was the ideal used to formulate Roman administration systems. How far it is applicable in reality is unclear, and, instead of competition, it is possible that power was concentrated in the hands of a few families (Blagg 1990). Nevertheless, the wider community

Urban Monumentality in Roman Britain    773 was brought into this through the idea of the citizen community, and practices such as voting, and the taking of the census, provided a formal, if infrequent, opportunity to engage with the town as a political centre. While certain activities were applicable only to Roman citizens, voting was extended to the urban citizenship. Applying the picture to Britain is not uncontentious. The four colonia of Gloucester, Colchester, Lincoln, and York and the one municipium of Verulamium may well have had similar charters to those from elsewhere in the western provinces (González 1986; Crawford 1996), but we have no evidence for the form of charters given to the civitas capitals, which comprised the majority of the towns. Therefore we have to assume that the political activities carried out there would be similar, although not necessarily identical, to those we know about from other provinces. Nevertheless, the evidence for their ongoing alterations to suit better the needs to the community demonstrates their significance, and that these were not white elephants in the urban skyline.

Space for Worship As discussed in the previous section, one notable feature of the Romano-​British forum was the lack of a freestanding temple, with, instead, a shrine leading from the basilica. Nevertheless, religion formed an important part of the urban experience within the province. Most towns provide evidence for one or more temples of varying monumentality and size. There were few classical-​style temples, with Romano-​Celtic temples being the main type, both in the centre of the towns and on the periphery (Goodman 2013). Nevertheless, this should not detract from the function of temples as the focus for religious activity, and there is little evidence for them being used for non-​Roman or resistant religious practices. Instead of focusing on the building type, it is more fruitful to explore how they were used in the articulation of urban space. In certain towns, temples formed part of a monumental urban centre, associated with the forum and/​or a theatre. This is best expressed in Verulamium, with the pairing of the temple and theatre, linked to the forum by a road out of alignment with the main street grid. Much has been written about the location of the temple in insula 16 in Verulamium and its relationship with the theatre and the Folly Lane burial (Haselgrove and Millett 1997; Niblett 2001; Creighton 2006). The temple was constructed in the late first century ad on what seems to have been a pre-​conquest sacred site (Niblett and Thompson 2005). It was set within a large porticoed temenos or sacred precinct, 100m × 50m, and took the form of a square Romano-​Celtic temple. In the later second century ad, the shrine was further elaborated with side wings, and a double portico was added to the wall surrounding the temenos, along with a monumentalized porch. This porch faced south, towards the theatre, which had been constructed ad 140–​160, and was also altered in the late second century (Kenyon 1935). There is some indication of an open area between the theatre and the temple, and this may have had some form of ritual purpose. This temple–​theatre complex was linked to the forum by a street leading

774

774   Louise Revell to the side entrance of the forum, creating a focus for public religious and political activity within the town. It seems likely that a similar monumental centre can be identified at Canterbury, although here the picture is not straightforward (Figure 37.3). A series of excavations have uncovered a large portico enclosing a large central court, constructed c. ad 100–​120 (Bennett and Nebiker 1989). There has been less excavation of the area within this court, but in one corner stood a timber Romano-​Celtic temple, built when the enclosure itself was first constructed, and demolished at some point in the third century ad (Bennett 1981). It is usually assumed that this was one of a series of temples within the temenos, possibly the large classical temple suggested by Blagg (1984) based on the architectural fragments. A theatre stood to the east of the temenos complex. It was constructed in the late first century ad, earlier than the temple, and was altered in the early third century (Frere and Simpson 1970). To the north of the theatre, a substantial baths building was constructed in the early second century (Blockley et al. 1995). The insula directly to the north of the temenos has produced evidence for monumental structures, and this has led to the hypothesis that it was the site of the forum. A large quantity of marble was excavated in this area, approximately two-​thirds of which was imported, predominantly from Carrera, but also from other eastern Mediterranean quarries (Blagg 1984: 66–​7 1). Fragments of Corinthian columns were also among these remains. Thus, by the early second century, the heart of Roman Canterbury consisted of four insula forming a monumental core of sacred shrine, adjoining theatre, baths, and possibly the forum, linked by a wide road out of alignment with the main grid. This association has led Millett to argue that the town could be seen as a religious sanctuary (Millett 2007). However, in terms of monumentality, there is an interesting contrast between the overall appearance of the theatre, baths, and the exterior of the temenos, with the Roman-​style architecture and use of marble, and the only known temple inside the temenos, which was built of timber and not reconstructed in masonry before its demolition in the third century. Even if there was the large temple suggested by Blagg, this deliberate curation of an increasingly ‘archaic’ building material can only be seen as creating a particular symbolism for the shrine. At Colchester, we also have the situation of a monumental temple complex which has been identified as the Temple to Deified Claudius, mentioned in Tacitus’ account of the Boudiccan Revolt (Tac. Ann. 14.31). There is evidence of an associated theatre, but, again, no certain location for the forum. One possibility is the insula opposite the entrance to the temenos. This irregular shaped insula (insula 30) has produced remains for a substantial masonry building, possibly relating to some form of basilica. There are also traces of a foundation wall crossing the street between them. This could have been a monumental arch, echoing that at the Balkerne Gate, or, alternatively, it could indicate that the two insulae formed a single complex. It has been argued that insula 30 was the forum, and, if linked, the two together could have formed a Gallic-​style double forum, with the main road from the Balkerne Gate leading straight to it. This complex formed a monumental focus for the city. The remains of the podium survive below the Norman castle, and it was 23.5m wide and 32m long (Drury et al. 1984). Nothing remains of the

Figure 37.3  Plan of Canterbury. Source: Millett (2007: figure 5.15). Reproduced by kind permission of the Kent History Project, Kent County Council, from plans provided by the Canterbury Archaeological Trust.

776

776   Louise Revell superstructure, so it is impossible to reconstruct its full appearance, but it was presumably a classical-​style temple. Raised up on a 4m-​high podium, it must have been a dominant feature on the skyline. The large podium temple was set within an architecturally elaborate complex. In front of the temple stood an altar, set within a paved area. This was approached through a monumental gateway, set within an arched screen wall fronting onto the main road. The complex seems to have undergone a major refurbishment in the late second century, and it may be that at this point it was linked with a newly constructed theatre in the adjoining insula (Crummy 1982), which theatre has been dated to the second or third century, with no evidence for earlier structures. While these three examples provide the clearest evidence of temples as part of a monumental centre, we see hints of this in other towns. At Caistor-​by-​Norwich, two Romano-​Celtic temples stood in the insula to the north of the forum (Atkinson 1931), and at Caerwent, when a temple was constructed within the town walls in the fourth century, it was in the insula adjoining the forum (Brewer 1993). At Gloucester and Wroxeter, it has been suggested that large colonnaded insulae adjacent to or opposite the forum represent the temenos of a temple (Hurst 1999; White et al. 2013: 92–​93). At Cirencester, the relationship is more problematic. It has been suggested that insula 6 was a temple from the evidence of a large open courtyard surrounded by porticoes on at least three sides, constructed in the mid-​second century (Holbrook 1998). However, while it was in the insula adjoining the forum and so looks related on a plan, in reality there was no immediate link between them at street level, as this complex was placed behind the basilica, and so was at the opposite end to the forum entrance. These ritual centres formed one way temples featured prominently within the urban experience. A second way was to mark key points and junctions within the town grid. One was the entrance to the town, moving through the peripheral zone of outside the boundary, with its cemeteries and industrial working, and moving into the town itself (Goodman 2007). As the traveller approached the western Balkerne Gate into Colchester, they would see a square Romano-​Celtic temple separated from the road by a gravelled court containing the plinth for a statue or an altar (Crummy 2006). This was constructed in the first quarter of the second century, although there is a suggestion of an earlier shrine, and it was probably dedicated to Mercury. There may have been a second, smaller temple further outside the town, although the evidence for this is problematic. At Silchester, a temenos stood just inside the eastern entrance to the town, which contained at least two temples and associated structures of unknown date. An alternative location was on key roads through the towns, as can be seen at Wroxeter (Figure 37.4), where three of the four known temples run along the main road through the town (White et al. 2013). At Verulamium, the triangular temple marked the splitting of the main road into the town on the eastern side, and a temple stood slightly back from the road in insula 21 (Niblett and Thompson 2005). Other complexes stood further away from the town, potentially connected through processions, but still a distinct extra-​ mural ritual focus. This is most clearly seen at the Gosbecks complex at Colchester, which formed a sanctuary complex of temple and theatre (Hull 1958: 259–​269). A similar phenomenon can be seen at Caistor-by-Norwich, where a temple stood to the north-​east

Urban Monumentality in Roman Britain    777

Figure 37.4  Plan of Wroxeter. Source: White et al. (2013: figure 4.21).

of the town, reached from the town by a street diagonal to the main street grid, linking this temple with the pair of temples in the town centre (Bowden 2013). One characteristic of some of these religious areas is the clustering of Romano-​Celtic temples in groups of two or more. We have seen this already at Silchester, where the

778

778   Louise Revell insula 30 complex comprised two or possibly three shrines within a single temenos wall, and possibly at the Balkerne Gate at Colchester. This is repeated at Caistor-​by-​Norwich, where two Romano-​Celtic temples were found in insula 9, probably dating to the late second century (Atkinson 1931). Recent excavations at Tabard Square in London, on the south banks of the Thames, have revealed a pair of temples constructed in the late first or early second century (Durrani 2004). An open space with a mortar floor stood between them, with evidence for stone plinths decorating it. An inscription from the site recorded a dedication to the Numina of the emperors and Mars Camulos (RIB 3014). At Verulamium, two Romano-​Celtic temples were located in insula 17, although these are only known from survey, and so it is difficult to understand the sequence of their construction. However, in both cases, architectural differences demonstrate that neither was envisaged as an identical pair. This is clearest at Caistor, where the two temples were of different dimensions and different construction methods. Temple B was approximately 13.7m × 12.8m, with walls constructed of irregularly shaped stones, with more regular pieces used for quoins at the corners. Temple A was smaller, at approximately 6.1m × 5.5m, and was constructed of rubble walls faced with flints, and bonding courses of tile. It is possible that temple A was constructed first, although the dating evidence is imprecise, and could equally indicate simultaneous construction. This difference in size is also evident at Silchester; however, the London temples have been interpreted as a deliberate pair, both seeming to be 11 metres × 11 metres and sharing the same alignment. The variability in religious experience is reinforced by the presence of temples to the so-​called oriental deities within the some towns. Perhaps the most well-​known evidence for this comes from London, with the temple of Mithras on the eastern bank of the Walbrook, although this was not constructed until the mid-​third century. A temple to Isis is hinted at within the epigraphic evidence, with a graffito on a pot reading londoni ad fanum isidis (at the temple of Isis, London; RIB 2.8.2503.127; discussed in Haynes 2000), and an inscription recording the rebuilding of the temple to Isis by the imperial governor in the third century (RIB 3001). A further urban Mithraeum may have been built in Leicester during the second century, although this has not been fully published, and its identification has been questioned (Crompton 1971: 75; Sauer 2004). Nevertheless, the preliminary report of a basilica-​style structure with the aisles raised slightly higher than the nave does echo the form of that found in London (Shepherd 1998). There is also a temple to Serapis at York (Ottaway 1993:  88). Also associated with the eastern deities is the cult of Cybele, which has been argued to be present in Verulamium. Here, an unusual shaped triangular temple was excavated by Wheeler, and consisted of an internal portico, with a series of shrines or rooms first constructed in the early second century. While temples within towns are usually overlooked as a group (although see Esmonde Cleary 2005; Creighton 2006), they form an important feature in the monumentality of the townscape. The rituals carried out in the urban temples, such as animal sacrifice and the burning of incense, would have added to the sensory experience of the town: the noises of the animal victims, and the smell of the burning meat, or the wafting of incense. Processions, and spectacles in the theatres (whatever form they took), would

Urban Monumentality in Roman Britain    779 have punctuated the daily grind, adding a sense of spectacle to the urban experience. One element which is clear is the variability in religious experience. Within the town, there were the monumental centres, single or double shrines, potentially set at strategic points within the towns, and the shrines of the so-​called oriental cults. Taken together, they provide a picture of a vibrant and varied religious experience within the towns.

Space for Spectacles and Bathing The relationship between the temples and both baths and theatres in the previous section raises the question of a third form of monumental architecture within the town—​ buildings, which might be loosely grouped together as those for entertainment. These might be considered somewhat trivial compared to fora and theatres, but they were an important feature of Roman urbanism. Buildings for spectacles (theatrical games, gladiatorial games, and chariot-​racing) and the large public baths were spaces where core tenets of Roman-​ness were enacted. Whether they encouraged the ideals of cleanliness or the competition between gladiators or charioteers, these buildings allowed their users to participate in an ideal of urbanism which spread throughout the empire. The feature all such buildings shared was the idea of public living. Beginning with buildings for spectacles, amphitheatres dominate, with fewer theatres and only one known stadium. However, the unexpected discovery of the Colchester circus in 2004 allows the possibility that other theatres and amphitheatres may yet be revealed. This is reinforced by the recent geophysical survey of Caistor-​by-​Norwich, which revealed a possible theatre in the insula opposite the two Romano-​Celtic temples (Bowden and Bescoby 2008). However, some structures thought to have been amphitheatres or theatres are unlikely. The so-​called amphitheatre at Caerwent illustrates some of these problems: this was an antiquarian excavation, poorly reported. The walls do not fit well into the surrounding buildings if this was an amphitheatre, and, if it was indeed one, it seems to have been a late construction and never completed (Ashby et al. 1905). The association between theatres and temples has already been discussed, but how do these work as buildings in their own right? There is evidence for seven urban theatres from Britain, from a total of six towns (this includes two at Colchester). However, the theatre at Caistor-​by-​Norwich is known only from the recent geophysics, and that at Brough-​on-​Humber is assumed from an inscription commemorating its construction (RIB 707). Only Canterbury and Verulamium have produced any form of stratigraphy, but both show the same sequence of development. The former was constructed in the first century ad, and Wilmott (2007) has argued that this initially followed the plan of the Gallo-​Roman hybrid theatre–​amphitheatre, where the seating extends beyond the usual semi-circle, creating a larger orchestra that could then be used as the arena for gladiatorial games. However, in reality it had an unusually elongated cavea, which does not suggest its use as an amphitheatre (Frere and Simpson 1970). It was then substantially

780

780   Louise Revell altered in the third century ad to create a more classical-​style theatre. This pattern is also seen with the theatre at Verulamium (Kenyon 1935), which was constructed in the mid-​ second century ad, again as a Gallo-​Roman theatre. It was rebuilt within two or three decades, and again the new layout resembled a more classical-​style theatre. In contrast, the theatre at Gosbecks, three miles west of Colchester, was constructed in a classical style from its earliest phase, although there is some uncertainty over the arrangement of the stage (Dunnett 1971). The theatre within Colchester itself seems to have been of classical style, although the stratigraphy is unclear and the remains may not represent the earliest phase (Crummy 1982). The Gallo-​Roman theatres have been the subject of much debate (summarized in Sear 2006: 98–​101), and they have variously been argued to be hybrid theatre–​amphitheatres because of the shape of the seating area, or not theatres at all because of the lack of a substantial stage-​setting. The rebuilding of two of the British examples as more classical-​style theatres suggests that they were considered theatres of some type, but that, over time, different architectural arrangements were desired by the townspeople. Amphitheatres are more common than theatres throughout the province overall, but in towns the proportions are more equal, with seven amphitheatres (excluding Caerwent but including Aldborough) as opposed to seven theatres. Those that can be dated were early constructions, with the earliest at Silchester constructed in timber during the period c. ad 55–​75. This was followed by a group dated to the Flavian period: the first timber phase at London, and the masonry amphitheatres at Chichester and Dorchester, while the amphitheatre at Cirencester dates to the early second century ad. Unlike theatres, these never formed part of the monumental centres of the town, and were positioned either at the outer edge of the town or just outside. Nevertheless, their positions near major routes allowed them to provide a dominant landmark within the cityscape. The amphitheatre at Cirencester, for example, stood just outside the western bath gate, on the main west–​east road through the town, which ran past the front of the forum. The amphitheatre at Dorchester reused a Neolithic henge monument. There is an obvious functional explanation, as this reuse would have reduced the effort needed for its construction. However, this may also have served as a memorial to the past occupation of the area, creating an idea of continuity which we find in the choice of location of buildings in other towns. This reflects an unexpected pattern: Welch (1994, 2007) has argued that the growth in amphitheatres during the mid to late Republic was due to their association with the military and with veteran settlements. However, although we do see non-​urban amphitheatres associated with fortresses in Britain (Wilmott 2007), the urban examples do not show this connection. Only London has definite military connections, with the Cripplegate fort and some form of military installation at Southwark, while no colonia has produced evidence for an amphitheatre. However, other towns with amphitheatres, such as Chichester, Cirencester, and Silchester, do not seem to have had a military origin. This is not to deny the military association as a general phenomenon, but it does suggest that, in Britain, there were multiple routes to their adoption. A type of building which can be included in this group is the monumental bathing complexes. Discussing these in any depth is difficult due to the problematic nature of the

Urban Monumentality in Roman Britain    781 evidence. In some towns, such as Colchester, there is no evidence for large public baths, but this is attributable more to continuous occupation on the site than to certainty of their absence. Furthermore, in most cases where a large set of public baths has been excavated, this is only a partial excavation, in many cases revealing only part of the main suite, or a series of snapshots difficult to piece together. However, in terms of their contribution to the monumentality of the town, there are three important points to be made. The first is that, while they were not necessarily the first public buildings the town possessed, they were an early part of their monumentalization, and by the mid-second century all those which we can date either were completed or were underway. The first phase of the baths at Caerwent date to the late first century ad, at about the same time as the construction of the forum (Nash-​Williams 1930). The baths at Chichester (Down 1978) and at Exeter (Bidwell 1979) also date to the first century ad, while those at Canterbury (Blockley et al. 1995), Wroxeter (Ellis 2000), and Lincoln (Jones et al. 2003: 79–​80) date to the first half of the second century. Others may have had similarly early dates, but lack published dating evidence. The second feature is that most of these bath buildings provide evidence for their rebuilding or rearrangement. At Caerwent, the entire structure was completely rebuilt seemingly within fifty years of its original construction, and then additional rooms and a new wing added, apparently during the third and fourth centuries (Nash-​ Williams 1930). The baths at Wollaston House in Dorchester also seem to have been rebuilt, which would have doubled the size of the complex, with further rooms added at a later date. Other baths show less dramatic extensions: at Silchester the frigidarium was extended and a second caldarium added in period 3, and the palaestra was widened in period 4 (Boon 1974). At Canterbury and Chichester, rooms went through changes of function, usually changing between heated and unheated. The third point is that these were large buildings, which seem to have gone beyond the functional row of cold, warm, and hot rooms. In many of the British urban bathhouses there were associated steam rooms, plunge pools, palaestrae or basilicas, and swimming pools (usually outdoor). Most of the bathhouses seem to show some evidence for an entrance through either a peristyle/​palaestra or a basilica. Both Caerwent and Wroxeter were entered through a large basilica, while at Caistor-​by-​Norwich and Silchester there was evidence for a porticoed peristyle court. At Canterbury and Exeter these seem to have contained open-​air swimming pools, and this also seems to have been the case in the insula 19 baths at Verulamium. Similarly, we see multiple heated rooms at Lincoln, Verulamium, and Leicester going beyond the single tepidarium and caldarium. Many contain pools or plunge baths, whether heated or cold, and at Silchester there seems to have been an additional sweat room. DeLaine (1992) has argued for a distinction between bathing complexes primarily for hygiene in the Republic and the imperial baths with an emphasis on pleasure and enjoyment, and the idea of variability in the social role of bathing can be seen in the difference between the facilities and decoration evident in the legionary and auxiliary bathhouses of Britain (Revell 2007b). These urban bathhouses all adhere to the idea of bathing as a social experience, with the basilicas and palaestrae providing space for exercise and chatting, while the multiple bathing facilities allowed for a more leisurely visit.

782

782   Louise Revell Often ignored or seen as frivolous, the baths and the buildings for games all share two important features. The first is that they were usually constructed early in the history of each town, either at its inception, or during the first fifty years. The second characteristic is that they were subjected to continual rebuilding through to the fourth century. Romano-​British urbanism has been categorized as a failed experiment, and the lack of public facilities has been used to support this. However, it is an error to view a Roman town as consisting of an essential suite of buildings, and there are examples of towns in Gaul and Iberia which do not have baths, theatre, and amphitheatre. In Britain, we see that most of the towns have a large public bathhouse with a range of bathing facilities. Similarly, most also have a theatre or an amphitheatre, with Cirencester and Aldborough possibly having both, and Colchester suffering an embarrassment of riches, with two theatres and a circus. This was not true for all towns, and the recent publication of the geophysics plan for Wroxeter demonstrates that, while it had a monumental bathhouse, it lacked both theatre and amphitheatre (White et al. 2013). Similarly, there is no evidence for buildings for spectacles at Lincoln, Exeter, Leicester, and York, although these towns lack full excavation due to continued occupation. Nevertheless, the late second century and early third century were a time of renovation for many of these buildings, showing the ongoing reformulation of the idea of desired building type, and the development of a discourse of architectural design within the province.

Space for Living and Working A Roman town was more than a collection of public buildings; it was also a place for living and working. This is an overlooked element of Romano-​British urbanism, and contributes to the picture of towns as underpopulated spaces. While we have an image of the elite (in spite of their lack of visibility epigraphically in contrast to elsewhere in the empire), and can set them within the towns, there is less consideration of the non-​elite within the towns. Urban populations are usually assumed to be low, although in reality there has not been any systematic attempt to calculate the demography of the towns. However, the archaeology suggests a more nuanced picture. If we take the example of Wroxeter, the geophysics results show the centre of the town dominated by public buildings, and immediately around this a dense area of generally elite houses (White et al. 2013). The rest of the insulae do not necessarily show evidence for specific houses, but the geophysical survey suggests industrial activity, and possibly timber housing. This raises the possibility that the towns were more densely occupied than we have acknowledged in the past. Studies investigating the economic activities in individual towns have reinforced this picture. Hall’s study (2005) of crafts in London represents a somewhat typical picture of multiple, small-​scale trades. There is evidence for all stages of metalworking from the smelting of ores through to the production of metal objects. Crucible fragments were found on numerous sites, suggesting this was widespread throughout the city. Dumps

Urban Monumentality in Roman Britain    783 of glass-​working debris indicate glass production, and similarly wooden offcuts indicate carpentry and dumped leather shoe-​making and other leatherworking. There is also the evidence for the production of foodstuffs. This picture is repeated at Verulamium, where again there is evidence for varied forms of small-​scale industrial activity throughout the town (Niblett and Thompson 2005). Metalworking and pottery production flourished from the earliest phase of the town through to the mid-​third century. Those living in the towns may also have been involved in low-​level agricultural production, with allotments suggested from aerial photographs, and the possibility of back gardens. Some made their living from the processing of agricultural produce for food, with remains of corn-​drying ovens, bakeries, and brewing beer. As at London, there is evidence for leather-​working and possibly tanning. There may also have been wool working or weaving, with the possible remains of large wool carding combs, and possibly cloth preparation. Furthermore, at Colchester, ceramic, glass, and bone-​working industries have all been identified, and, as well as general metalworking, there is both material and epigraphic evidence for a specialist coppersmith (Gascoyne 2013; RIB 194). Those engaged in many of these industries would be living and working in the same premises. A typical example is the timber-​built insula 14 buildings in Verulamium, excavated and published by Frere (1972). Although there has been some debate over the significance of the building technique (contrast Frere 1972 with Millett 1990), the interpretation of the buildings’ function has not been debated. These combined living quarters, workshop, and retail premises of metalsmiths, who worked mainly in bronze but also in iron and gold, making and repairing metal items. Similar structures have also been seen in Caerwent, although the sequence is difficult to interpret, and at No. 1 Poultry in London (Brewer 2006; Hill and Rowsome 2011). These buildings lack the investment we see in the public buildings, being largely built of timber, not stone or brick, with few rooms in addition to the workshop, and lacking decoration. It is only in the late second century and the early third century that we see the replacement of timber buildings with masonry with new internal arrangements and more elaborate decoration. Faulkner’s analysis (2002) of the development of domestic architecture in Britain shows an increase in the number of rooms occupied in town houses throughout the second and early third centuries until c. 250, after which it remains broadly constant until a decrease begins from c. 325, which continues throughout the fourth century. This long sequence of construction and reconstruction can be seen at Dorchester. At the Greyhound Yard site the initial first-​century sequence of buildings comprised low-​density individual houses occupying street frontages, possibly with enclosures to the rear (Woodward et al. 1993). Towards the end of the second century, buildings with continuous stone footings began to be constructed, and by the early third century many of the timber houses had been replaced, but on a roughly similar scale. In the late third century these were then replaced with larger buildings, either through extensions, or through demolition and rebuilding. Similarly at the County Hospital site, the earliest buildings were of timber, dating possibly as early as ad 60–​65, and were replaced by a new series of town houses, probably masonry built, with a layout of multiple rooms and a portico, and decorated with mosaic floors. In the fourth century a large town house

784

784   Louise Revell was built, set back from the street, and with evidence for elaborate polychrome mosaics. This trajectory can also be seen in Silchester in the ongoing excavations in insula 9, where the early timber houses were replaced with two masonry houses during the second century, and then, in the third century, these were amalgamated into a single masonry house. This new house was multi-​roomed, arranged along a veranda (Fulford and Clarke 2011). Wallace’s study (2014) of early Roman London reinforces this picture of lack of investment in domestic structures, arguing for a lack of mosaic, marbles, large receptions rooms, and other accoutrements associated with social display. The development of non-​public structures stands in contrast to that of the public buildings. Initially low key and timber built, they lack the evidence of investment seen in the fora, temples, baths, theatres, and amphitheatres. Nevertheless, it is important to recognize two key characteristics. The first is that, in spite of the comparative lack of investment, these buildings are early constructions, and illustrate that Romano-​British urbanism was more than the acquisition of a series of public buildings, but were places where people lived and worked from the outset. Secondly, while these were initially small scale, the townspeople were committed enough to the towns to continue investing in urbanism, redeveloping as finance and opportunity arose.

Urban Variability Thus far, I have argued that, in the first two centuries after the Claudian conquest, we can see a pattern of urban monumentalization which focused on the public above the domestic, and on spaces for public gatherings, whether political, religious, or for entertainment, all of which broadly followed the ideal of urbanism set out in Rome itself. Nevertheless, the one issue I have yet to raise is that of the variability between towns. While all the chartered towns of Britain accord with this ideal to some extent, this does not mean we see uniform patterns of urbanism, and it is this variability I will address in this section. While we lack complete plans for all towns, there are enough to demonstrate that there are substantial differences between the towns. The first element of this is in the size of the towns. The rank-​size graph of the area enclosed by defences for the towns of Britain demonstrate that the public towns have a steep drop-​off, from the largest, London, at over 130 hectares (Millett 1990: figure 62). Only a further three are over 60 hectares, with most clustering at 30–​50 hectares in size. The smallest public town was Caistor-​by-​Norwich, at 14 hectares. There are obvious problems with this means of quantification, as argued by Hurst (2005), because of the exclusion of suburbs from these figures. Nevertheless, the symbolism of the city walls and their role in creating the idea of inside and outside make this a factor in the creation of difference. A second element is the distribution of public buildings, with some containing multiple examples, while others are more meanly furnished. Verulamium, for example, has a monumental centre of forum, temple, and theatre, while elsewhere in the town we see

Urban Monumentality in Roman Britain    785 further temples and the suggestion of public baths. The public buildings in Silchester are more dispersed, but still include a forum, multiple temples, an amphitheatre, and large town baths. Further west, Cirencester also seems to develop a monumental centre, with a theatre and amphitheatre in addition to the forum, although no temples have been securely identified. In contrast to these, there are towns with fewer public buildings, such as Wroxeter, which survey evidence demonstrates to have been a densely occupied town of 76.8 hectares, but which does not show the same level of monumentality (see Figure 37.3): there was a large forum–​basilica with a substantial baths and macellum occupying the whole of the opposite insula (White et al. 2013). However, we do not see a theatre or amphitheatre forming part of the monumental core. While it is tempting to relate this to size, Caistor-​by-​Norwich, much smaller than Wroxeter, was furnished with a forum–​basilica, twin temples, a public baths, a theatre, and an amphitheatre (Frere 1971; Bowden and Bescoby 2008). It is this variability in size and monumentality which has led to the categorization of towns as more or less successful, as though there was an expectation of Roman towns to be homogenous. However, various studies have pointed to the inherent variability of urbanism within any single province of the empire (Keay 1995; Woolf 1997), and that, in reality, urbanism was the creation of a hierarchy of towns. This was in part political, but other factors came into play, such as economics, religion, pilgrimage, and so on. The lack of textual evidence (including epigraphic) makes such factors difficult to understand, but we can identify their influence on urbanism in Britain. In terms of political hierarchies, the role of London as an administrative centre in the post-​Boudiccan period creates a different form of monumentality from that seen in the first two decades of its development (Wallace 2013). Cirencester and York may have similarly developed as provincial capitals as Britain was divided into two and then four provinces. Onto this, we can add the possibility of a religious hierarchy, most notably Canterbury, with its ritual centre of temple, baths, and theatre, and also Colchester as the centre for the provincial imperial cult with the Temple to Deified Claudius. These different hierarchies and networks would have intersected, creating a complexity to urbanism within the province which goes beyond a single criterion for their assessment.

Conclusion There has been a tendency to view Romano-​British urbanism as something of a failed experiment, reflected (and distorted) through a prism of the idealized Roman town. In contrast, when towns are studied on their own terms, we see a different picture. The emphasis on monumentality and public spectacle ties the towns of Britain into the empire-​wide discourse of urbanism. The ongoing maintenance and reconstruction of the buildings, as well as the gradual investment in commercial and domestic buildings, demonstrate that this idea continued through to the late third and fourth centuries, by which time a new concept of urbanism was emerging. At the same time, the differences

786

786   Louise Revell between Britain and other provinces, as well as the differences between towns within Britain, all point to an inherent variability. As Hurst has argued, the differences between Gloucester and Cirencester were the result of ‘two communities expressing themselves, through material culture we know as Romano-​British, in ways which convey their identity and aspirations. They used that expression partly to be alike … but also … to mark their differences’ (Hurst 2005: 303; see also Revell 2009). None of this is unique to Britain, but in many ways echoes the picture in other provinces, warning against viewing the province in a vacuum. For example, urbanism in Britain is often argued to have been a delayed development (Mattingly 2006), but Woolf (1997) has shown that the construction of towns in Gaul could take up to two generations. Similarly, the majority of towns in Iberia and Gaul do not differ substantially in size or monumentality from those in Britain (Bedon et al. 1988; Bendala Galán 1993). Viewed within their wider context, Romano-​British towns represent an important change in the lives of the inhabitants of the provinces, and a key part of the political, social, economic, and cultural changes in the province following the Roman conquest.

Abbreviation RIB R. G. Collingwood and R. P. Wright, The Roman Inscriptions of Britain. Volume I: Inscriptions on Stone. Oxford: Clarendon Press, 1965–​to date.

References Primary Sources Tacitus, Annals, trans. J. Jackson (London: Heinemann, 1925–37).

Secondary Soiurces Alföldy, G. (1996). ‘Subject and Ruler: Subjects and Methods: An Attempt at a Conclusion’, in A. Small (ed.), Subject and Ruler: The Cult of the Ruling Power in Classical Antiquity. Journal of Roman Archaeology Supplementary Series 17. Ann Arbor, Michigan, 237–​259. Ashby, T. (1906). ‘Excavations at Caerwent, Monmouthshire, on the Site of the Romano-​British City of Venta Silurum, in the Year 1905’, Archaeologia, 60/​1: 111–​130. Ashby, T., Hudd, A. E., and Martin, A. T. (1905). ‘Excavations on the Site of the Romano-​British city of Venta Silurum’, Archaeologia, 50: 88–​115. Atkinson, D. (1931). ‘Caistor Excavations, 1929’, Norfolk Archaeology, 26: 93–​139. Balty, J. C. (1991). Curia ordinis: Recherches d’architecture et d’urbanisme antiques sur les curies provinciales du monde romain. Brussels: Palais des Academies. Bedon, R., Chevallier, R., and Pinon, P. (1988). Architecture et urbanisme en Gaule romaine. Paris: Editions Errance. Bendala Galán, M. (1993). The Hispano-​Roman Town. Barcelona, Ministerio of Cultura: Àmbit Servicios Editoriales.

Urban Monumentality in Roman Britain    787 Bennett, P. (1981). ‘68–​69A Stour Street’, Archaeologia Cantiaca, 97: 279–​281. Bennett, P., and Nebiker, D. (1989). ‘No. 76 Castle Street’, Archaeologia Cantiaca, 107: 283–​286. Bidwell, P. T. (1979). The Legionary Bath-​House and Basilica and Forum at Exeter. Exeter: Exeter City Council, University of Exeter. Blagg, T. F.  C. (1984). ‘Roman Architectural Ornament in Kent’, Archaeologia Cantiaca, 100: 65–​80. Blagg, T. F. C. (1990). ‘Architectural Munificence on Britain: The Evidence of the Inscriptions’, Britannia, 21: 13–​31. Blockley, K., Blockley, M., Blockley, P., Frere, S., and Stow, S. (1995). The Archaeology of Canterbury:  Excavations in the Marlowe Car Park and Surrounding Areas. Canterbury: Canterbury Archaeological Trust. Boon, G. C. (1973). ‘Serapis and Tutela: A Silchester Coincidence’, Britannia, 4: 107–​114. Boon, G. C. (1974). Silchester, the Roman Town of Calleva. Newton Abott: David & Charles. Bowden, W. (2013). ‘The Urban Plan of Venta Icenorum and its Relationship with the Boudican Revolt’, Britannia, 44: 145–​169. Bowden, W., and Bescoby, D. (2008). ‘The Plan of Venta Icenorum (Caistor-​by-​ Norwich): Interpreting a New Geophysical Survey’, Journal of Roman Archaeology, 21/​1: 324–​335. Brewer, R. J. (1993). ‘Caerwent—​Venta Silurum: A Civitas Capital’, in S. J. Greep (ed.), Roman Towns:  The Wheeler Inheritance. A  Review of Fifty Years Research. London:  Council for British Archaeology Research Report 93: 56–​65. Brewer, R. J. (2006). Caerwent Roman Town. Cardiff: Cadw. Crawford, M. (1996). Roman Statutes. London: Bulletin of the Institute of Classical Studies 64. Creighton, J. (2006). Britannia: The Creation of a Roman Province. London: Routledge. Crompton, J. (1971). ‘Archaeology in Leicestershire and Rutland April 1968–​March 1970’, Transactions of the Leicestershire Archaeological and Historical Society, 45: 74–​78. Crummy, N. (2006). ‘Worshipping Mercury on the Balkerne Hill, Colchester’, in P. Ottaway (ed.), A Victory Celebration: Papers on the Archaeology of Colchester and Late Iron Age Roman Britain. Colchester: Friends of Colchester Archaeological Trust, 55–​68. Crummy, P. (1982). ‘The Roman Theatre at Colchester’, Britannia, 13: 299–​302. Cunliffe, B. W., and Fulford, M. (1982). Bath and the Rest of Wessex. Oxford:  Oxford University Press. DeLaine, J. (1992). ‘New Models, Old Modes: Continuity and Change in the Design of Public Baths’, in H. J. Schalles and P. Zanker (eds), Der Romische Stadt im 2 Jahrhundert n. Chr. Der Functionswandel des offentlichem Raumes. Cologne: Reinland-​Verlag GmbH, 257–​275. Down, A. (1978). Chichester Excavations III. Chichester: Phillimore and Co. Drury, P. J., Bayley, J., Blagg, T. F.  C., Evans, J., Going, C. J., Hassall, M. W.  C., Niblett, B. R. K., and Wickenden, N. P. (1984). ‘The Temple of Claudius at Colchester Reconsidered’, Britannia, 15: 7–​50. Dunnett, R. (1971). ‘The Excavation of the Roman Theatre at Gosbecks’, Britannia, 2: 27–​47. Durrani, N. (2004). ‘Tabard Square Excavations, Southwark’, Current Archaeology, 192: 540–​547. Ellis, P. (2000). The Roman Baths and Macellum at Wroxeter: Excavations by Graham Webster, 1955–​85. London: English Heritage. Esmonde Cleary, A. S. (2005). ‘Beating the Bounds: Ritual and the Articulation of Urban Space in Roman Britain’, in A. MacMahon and J. Price (eds), Roman Working Lives and Urban Living. Oxford: Oxbow Books, 1–​17.

788

788   Louise Revell Euzennat, M., and Hallier, G. (1986). ‘Les Forums de Tingitane: Observations sur l’influence de l’architecture militaire sur les constructions civiles de l’Occident romain’, Antiquités africaines, 22/​1: 73–​103. Faulkner, N. (2002). ‘The Debate about the End:  A  Review of Evidence and Methods’, Archaeological Journal, 159: 59–​76. Fox, G. E., and Hope, S. J. W. H. (1893). ‘Excavations on the Site of the Roman City at Silchester, Hants, in 1982’, Archaeologia, 53/​2: 539–​573. Frere, S. S. (1971). ‘The Forum and Baths at Caistor by Norwich’, Britannia, 2: 1–​26. Frere, S. S. (1972). Verulamium Excavations. Volume I. London: Society of Antiquaries. Frere, S. S. (1989). ‘Roman Britain in 1988. 1. Sites Explored’, Britannia, 20: 258–​326. Frere, S., and Simpson, G. (1970). ‘The Roman Theatre at Canterbury’, Britannia, 1: 83–​113. Fulford, M. G., and Clarke, A. (2011). Silchester: City in Transition. The Mid-​Roman Occupation of Insula IX C. a.d. 125–​250/​300: A Report on Excavations Undertaken since 1997. London: Society for the Promotion of Roman Studies. Fulford, M., and Timby, J. (2000). Late Iron Age and Roman Silchester: Excavations on the Site of the Forum–​Basilica 1977, 1980–​86. London: Society from the Promotion of Roman Studies. Gascoyne, A. (2013). Colchester, Fortress of the War God:  An Archaeological Assessment. Oxford: Oxbow. González, J. (1986). ‘The Lex Irnitana: A New Copy of the Flavian Municipal Law’, Journal of Roman Studies, 76: 147–​243. Goodman, P. J. (2007). The Roman City and its Periphery:  from Rome to Gaul. London: Routledge. Goodman, P. (2013). ‘Temple Architecture and Urban Boundaries in Gaul and Britain’, in T. Kaizer, A. Leone, E. Thomas, and R. Witcher (eds), Cities and Gods: Religious Space in Transition. Leuven: Peeters, 81–​96. Gros, P. (1996). L’Architecture romaine. Paris: Les Manuels d’Art et d’Archeologie Antiques. Hall, J. (2005). ‘The Shopkeepers and Craft Workers of Roman London’, in A. MacMahon and J. Price (eds), Roman Working Lives and Urban Living. Oxford: Oxbow Books, 167–​190. Haselgrove, C., and Millett, M. (1997). ‘Verlamion Reconsidered’, in A. Gwilt and C. Haselgrove (eds), Reconstructing Iron Age Societies: New Approaches to the British Iron Age. Oxford: Oxbow, 282–​296. Haynes, I. (2000). ‘Religion in Roman London’, in I. Haynes, H. Sheldon, and L. Hannigan, (eds), London Under Ground: The Archaeology of a City. Oxford, Oxbow Books, 85–​101. Henig, M. (1993). Roman Sculpture from the Cotswold Region, with Devon and Cornwall. Oxford: Oxford University Press. Hill, J., and Rowsome, Pl (2011). Roman London and the Walbrook Stream Crossing: Excavations at 1 Poultry and Vicinity, City of London. London: Museum of London Archaeology. Holbrook, N. (1998). Cirencester:  The Roman Town Defences, Public Buildings and Shops. Cirencester: Cotswold Archaeological Trust. Hull, M. R. (1958). Roman Colchester. London: Society of Antiquaries. Hurst, H. (1972). ‘Excavations at Gloucestershire, 1968–​7 1: First Interim Report’, Antiquaries Journal, 52: 24–​69. Hurst, H. (1999). ‘Civic Space at Glevum’, in H. Hurst (ed.), The Coloniae of Roman Britain: New Studies and a Review. Ann Arbor: Journal of Roman Archaeology, 152–​160. Hurst, H. (2005). ‘Roman Cirencester and Gloucester Compared’, Oxford Journal of Archaeology, 24/3: 293–​305.

Urban Monumentality in Roman Britain    789 Isserlin, R. M. J. (1998). ‘A Spirit of Improvement? Marble and the Culture of Roman Britain’, in R. Laurence and J. Berry (eds), Cultural Identity in the Roman Empire. London: Routledge, 125–​155. Jones, M. J. (2002). Roman Lincoln: Conquest, Colony and Capital. Stroud: Tempus. Jones, M. J., Stocker, D., and Vince, A. (2003). The City by the Pool: Assessing the Archaeology of the City of Lincoln. Oxford: Oxbow. Joyce, J. G. (1873). ‘The Excavations at Silchester’, Archaeological Journal, 30: 10–​27. Kaiser, A. (2000). The Urban Dialogue: An Analysis of the Use of Space in the Roman City of Empuries, Spain. British Archaeology Reports International Series 901. Oxford: Archaeopress. Keay, S. J. (1995). ‘Innovation and Adaptation:  The Contribution of Rome to Urbanism in Iberia’, in B. W. Cunliffe and S. J. Keay (eds), Social Complexity and the Development of Towns in Iberia. London: Proceedings of the British Academy, 86: 291–​337. Kenyon, K. M. (1935). ‘The Roman Theatre at Verulamium, St Albans’, Archaeologia, 84: 213–​261. Laurence, R., Esmonde ​Cleary, S., and Sears, G. (2011). The City in the Roman West, c.250 bc–​c. ad 250. Cambridge: Cambridge University Press. Laurence, R., and Newsome, D. J. (2011). Rome, Ostia, Pompeii:  Movement and Space. Oxford: Oxford University Press. Lefebvre, S. (2004). ‘Espace et pouvoir local dans les provinces occidentales:  Quelques remarques’, in C. Auliard and L. Bodiou (eds), Au jardin des Hespérides: Histoire, société et épigraphie des mondes anciens. Mélanges offerts à Alain Tranoy. Rennes:  Presses Universitaires de Rennes, 379–​406. MacDonald, W. L. (1986). The Architecture of the Roman Empire. II. An Urban Appraisal. New Haven: Yale University Press. Martin, A. T. (1898). ‘Notes in the Roman Basilica at Cirencester Lately Delivered by Wilfred J. Cripps Esq., C.B.’, Transactions of the Bristol and Gloucestershire Archaeological Society, 21: 70–​78. Mattingly, D. (2006). Britain: An Imperial Possession. London: Allen Lane. Millett, M. (1990). The Romanization of Britain:  An Essay in Archaeological Interpretation. Cambridge: Cambridge University Press. Millett, M. (2001). ‘Review of Fulfold and Timby’, Archaeological Journal, 158: 394–​395. Millett, M. (2007). ‘Roman Kent’, in J. H. Williams (ed.), The Archeology of Kent to ad 800. Woodbridge: Boydell Press, 135–​184. Milne, G. (1992) (ed.). From Roman Basilica to Medieval Market: Archaeology in Action in the City of London. London: HMSO. Nash-​Williams, V. E. (1930). ‘IX.—​Further Excavations at Caerwent, Monmouthshire, 1923–​5’, Archaeologia (Second Series), 80: 229–​288. Niblett, R. (2001). Verulamium: The Roman city of St Albans. Stroud: Tempus. Niblett, R., and Thompson, I. (2005). Alban’s Buried Towns: An Assessment of St Alban’s Archaeology up to ad 1600. Oxford: Oxbow Books. Ottaway, P. (1993). Roman York. London: Batsford/​English Heritage. Revell, L. (1999). ‘Constructing Romanitas: Roman Public Architecture and the Archaeology of Practice’, in P. Baker, C. Forcey, S. Jundi, and R. Witcher (eds), TRAC98. Proceedings of the Eighth Annual Theoretical Roman Archaeology Conference. Leicester 1998. Oxford: Oxbow, 52–​58. Revell, L. (2007a). ‘Architecture, Power and Politics: The Forum-​Basilica in Roman Britain’, in J. Sofaer (ed.), Projecting Identities. London: Blackwells, 127–​151.

790

790   Louise Revell Revell, L. (2007b). ‘Military Bath-​Houses in Britain: A Comment.’ Britannia, 38: 230–​237. Revell, L. (2009). Roman Imperialism and Local Identities. Cambridge:  Cambridge University Press. Revell, L. (2013). ‘The Written City:  Political Inscriptions from Roman Baetica’, in G. Sears, P. Keegan, and R. Laurence (eds), Written Space in the Latin West: 200 bc–​ad 300. London: Bloomsbury, 231–​245. Rogers, A. (2013). Water and Roman Urbanism: Towns, Waterscapes, Land Transformation and Experience in Roman Britain. Leiden: Brill. Sauer, E. (2004). ‘Not just Small Change: C ​ oins in Mithraea’, in M. Martens and G. de Boe (eds), Roman Mithraism: The Evidence of the Small Finds. Papers of the International Conference, Tienen, 7–​8 November 2001. Brussels, 327–​353. Sear, F. (2006). Roman Theatres: An Architectural Study. Oxford: Oxford University Press. Shepherd, J. D. (1998). The Temple of Mithras, London:  Excavations by W.  F. Grimes and A. Williams at the Walbrook. London: English Heritage. Wacher, J. S. (1995). The Towns of Roman Britain. 2nd edn. London: Batsford. Wallace, L. (2013). ‘The Foundation of Roman London:  Examining the Claudian Fort Hypothesis’, Oxford Journal of Archaeology, 32/​3: 275–​291. Wallace, L. M. a. (2014). The Origin of Roman London. Cambridge: Cambridge University Press. Wallace-​Hadrill, A. (2011). ‘The Monumental Centre of Herculaneum:  In Search of the Identities of the Public Buildings’, Journal of Roman Archaeology, 24: 121–​160. Welch, K. (1994). ‘The Roman Arena in Late-​Republican Italy: A New Interpretation’, Journal of Roman Archaeology, 7: 59–​80. Welch, K. E. (2007). The Roman Amphitheatre: From its Origins to the Colosseum. New York and Cambridge: Cambridge University Press. White, R. H. (2002). Wroxeter: Life & Death of a Roman City. 2nd edn. Stroud: Tempus. White, R. H., Gaffney, C., and Gaffney, V. L. (2013). Wroxeter, the Cornovii and the Urban Process: Final Report on the Wroxeter Hinterland Project 1994–​1997, ii. Characterizing the City. Oxford: Archaeopress. Wilmott, T. (2007). The Roman Amphitheatre in Britain. Stroud: Tempus. Woodward, P. J., Davies, S. M., and Graham, A. H. (1993). Excavations at the Old Methodist Chapel and Greyhound Yard, Dorchester, 1981–​1984. Dorchester: Dorset Natural History & Archaeological Society. Woolf, G. (1997). ‘The Roman Urbanization of the East’, in S. E. Alcock (ed.), The Early Roman Empire in the East. Oxford: Oxbow Books. Zanker, P. (1998). Pompeii: Public and Private Life. Cambridge, MA: Harvard University Press.

Chapter 38

The Expl oitat i on of An im als in Roma n Bri ta i n Mark Maltby

Introduction: The History of Romano-​British Zooarchaeology The study of the exploitation of animals in Roman Britain using archaeological evidence is a relatively recent development. Although classic studies of the province often included sections on animal husbandry or diet, the evidence was based on a very limited number of archaeological investigations and often relied heavily on historical inference drawn from Roman agronomists (e.g. White 1970). The advent of historical zooarchaeology from the 1970s onwards has resulted in the study of assemblages from many hundreds of Romano-​British sites, producing a burgeoning and increasingly unwieldy archive. As the information has increased, several syntheses for the evidence have been produced (e.g. Grant 1989, 2004; King 1991, 1999; van der Veen and O’Connor 1998; Cool 2006; Albarella et al. 2008; Hesse 2011). This chapter will both draw upon these previous surveys and incorporate new evidence.

The Scope of This Review Evidence for the exploitation of animals in Roman Britain is derived from many facets of archaeological and, to a lesser extent, documentary sources. These will be mentioned where relevant, but this chapter will focus primarily on the zooarchaeological evidence. It will largely exclude discussion of the roles of animals in ritual and religion, as this is covered in other chapters in this volume (e.g. Smith and Weekes, this volume). The main

792

792   Mark Maltby theme that will be examined is that of innovation. Put simply, in what ways did human exploitation of animals change in the Romano-​British period? It should be noted from the outset that, although there are some common trends that can be seen with regard to the exploitation of animals in Roman Britain, there are also many variations. It can be shown that bone assemblages can vary significantly chronologically, regionally, and intra-​regionally, between different types of settlement, between settlements of the same type, and within individual settlements themselves. Arguably, it is these differences that tell us more about the complexities of the pastoral economy and the meat diet than the similarities. The review will focus initially on domestic mammals and their exploitation. This will be followed by a shorter summary of the evidence for the exploitation of other mammals, birds, and fish.

Cattle Previous reviews of Romano-​British diet have concluded that beef was comfortably the meat most commonly consumed. Cattle bones form the highest component of many faunal assemblages (King 1984, 1991, 1999; Grant 2004), whereas Iron Age assemblages are more commonly dominated by sheep/​goat (Hambleton 1999; Albarella 2007). Even in cases where minimum number calculations indicate that more sheep were eaten than cattle, the much heavier carcass weight of cattle would mean that beef still formed the bulk of the meat diet (Grant 2000: 428). King’s analyses (1984, 1999) of species representation in a wide range of assemblages demonstrated that, in addition to the general dominance of cattle, there were variations in the percentages of cattle, sheep/​goat, and pig on different types of site. This calculation masks a considerable amount of variation, which has been encountered in assemblages both within and between settlements, but King’s surveys showed that cattle tend to be best represented on military sites and large towns—​settlements that would have contained higher proportions of people not directly involved with food production. The increased demand for food created by urban populations and military personnel is likely to have been a major factor in the change of emphasis in Romano-​British meat production. The consensus view has been that the dominance of cattle tended to increase in the later Roman period following King’s original results (1984) (Grant 2004; Albarella 2007), with cattle forming a higher proportion of the bones on all types of settlement. Although this generalization can still be supported by the evidence, there are a lot of variations and numerous exceptions to this trend. For example, in a recent survey of Romano-​British assemblages from twenty-​eight sites in fourteen large towns where material from different periods could be compared, cattle percentages (in relation to sheep/​goat and pig) increased in the later Roman period in seventeen cases but decreased or remained static in the other eleven (Maltby 2010: 266–​267). In addition to changes in the diet, there are a number of factors that can account for variations in species in assemblages, including

The Exploitation of Animals in Roman Britain    793 differential preservation, butchery, and disposal practices. More detailed and critically evaluative reviews of species representation are required to provide a more comprehensive understanding of the extent to which beef production became more important in the later Roman period. A major change that accompanied the increased emphasis on beef production concerned how cattle carcasses were processed. It has long been recognized that butchery methods involved the greater use of cleavers and heavy blades in carcass-​processing than has been encountered in British Iron Age assemblages (Grant 1989; Maltby 1989). Assemblages from large urban and military sites in particular have evidence for consistent treatment of carcasses utilizing these implements, which have left characteristic butchery marks (Maltby 1989, 2007; Seetah 2006). In particular, the appearance of distinctive indentations (scoops) on the shafts of upper limb bones made by heavy blades during the removal of meat is testament to the processing of carcasses by specialist butchers. While such butchery is found frequently in urban and military assemblages, similar filleting marks have been found much less consistently on small nucleated settlements and villas. They tend to be even less frequent and often absent on other rural settlements (Maltby 2007, 2010). These methods, which may well have had military origins, as they have been evidenced in some early forts such as Exeter (Maltby 1979), would have increased the speed of processing and were thus favoured by butchers dealing with large numbers of carcasses (Seetah 2006). There is abundant evidence for large-​scale processing of cattle in towns, and examples of substantial accumulations of waste emanating from this can be found in all the major towns that have received extensive excavations (Maltby 1984, 2007, 2010). This is one of the major reasons why species representation varies so greatly within them. Areas (both intra-​and extra-​mural) that include large dumps of processing waste have invariably produced high percentages of cattle. Examples include Rack Street, Exeter (Maltby 1979), Chester Street, Cirencester (Maltby 1998), and Balkerne Lane, Colchester (Luff 1993). Other common finds on urban and military sites are cattle scapulae with distinctive blade marks on the edges of the articular surface (glenoid) and neck. These scapulae were disarticulated from the humerus. Holes in the blade of some of the scapulae suggest they were hung probably during preservation by smoking and salting. Well-​known examples have been illustrated from York (O’Connor 1988) and Lincoln (Dobney et al. 1996), and perforated specimens have been recorded in several other major towns (Maltby 2007). In addition, scapulae with similar blade marks on the neck and the glenoid to those commonly found in towns have also been recovered on some villas and other rural settlements such as Owslebury in Hampshire (Maltby 1989) and Marsh Leys Farm in Bedfordshire (Maltby 2011). This raises the possibility that preserved shoulders of beef may have been traded. There is evidence for a substantial increase in salt production during the Romano-​British period in many areas of England (de Brisay and Evans 1975; Bradley 1992; Lane and Morris 2001; Hathaway 2006; Rippon 2008), raising the probability that much of the meat eaten was in the form of cured products.

794

794   Mark Maltby Marrow and grease were other cattle products that were commonly utilized in Roman Britain. In addition to evidence that many bones were routinely broken for pot-​boiling, there is evidence that upper limb bones were sometimes collected for bulk processing. In addition to the characteristic filleting marks already described, these bones were split longitudinally to release marrow. Large accumulations of these split bones have been found mainly, but not exclusively, on major urban sites such as Gloucester and Winchester (Maltby 2007). Such accumulations have often been interpreted as evidence for the production of stock or broth. However, marrow can also be utilized for other products. Dobney et al. (1996: 25), for example, have suggested that marrow was obtained from mandibles in Lincoln and may have been used in oil lamps. Studies of mortality patterns based on tooth ageing evidence have shown that in many settlements a high percentage of cattle were not slaughtered until adulthood (Grant 2004: 373; Maltby 2010: 288). In most cases, however, relatively few of the adult cattle mandibles were from very old animals. Therefore, although cattle were undoubtedly valued for dairy products and traction power, cattle were commonly slaughtered for beef as young or mature adults (perhaps mainly between 4 and 8 years of age). Many major towns have high percentages of cattle of this age. Examples include Winchester, Dorchester (Maltby 1994), Caerwent (Maltby 2010: 288), Lincoln (Dobney et al. 1996), Silchester (Grant 2000; Ingrem 2006), Exeter (Maltby 1979), and Colchester (Luff 1993). Similar results have been obtained in assemblages from small towns such as Alcester, Warwickshire (Maltby 2001), and Heybridge, Essex (Albarella et al. 2008: 1837). Military sites producing assemblages with high percentages of adult cattle include South Shields (Stokes 2000), Loughor (Sadler 1997), and Portchester Castle (Grant 1975). Cattle mortality rates on villas and other rural settlements appear to be more variable. Although many assemblages contain high percentages of adult cattle, the peak of slaughter of animals of this age is often less marked. Examples include Owslebury, Hampshire (Maltby 1994), Odell (Grant 2000), Marsh Leys Farm and Newnham in Bedfordshire (Maltby 2011), and Abingdon and Barton Court Farm villa in Oxfordshire (Grant 2000). Most of these assemblages include fairly high percentages of cattle slaughtered between 1 and 3 years of age. These animals were not required for breeding, working, or sale. This may indicate that the inhabitants at these settlements were sufficiently wealthy to afford to cull or acquire potentially productive immature cattle for their meat. However, this is not necessarily related to status, as the phenomenon seems to occur at both villa and non-​villa sites. Veal was seemingly eaten in relatively small amounts compared to mature beef. Teeth and bones of young calves have been found on many sites but never in very high percentages. Some of these calves may have been natural mortalities, but others, particularly those found in towns such as Dorchester (Maltby 1994), probably reflect deliberate slaughter of animals acquired for their meat. Another aspect of cattle in Roman towns is the bias towards smaller cattle. Most breadth measurements of cattle limb bones in assemblages from Roman towns are positively skewed. Examples can be drawn from Cirencester (Maltby 1998: 362), Winchester (Maltby 2010: 150), and Alcester (Maltby 2001). This probably indicates that the majority

The Exploitation of Animals in Roman Britain    795 of the cattle were females, although this interpretation is complicated by the introduction of larger cattle in some parts of Roman Britain. However, the comparative lengths and breadths of metacarpals display sexual dimorphism, and measurements of complete bones consistently indicate that most of the adult cattle in urban assemblages belonged to females. Good examples can be found in Colchester (Luff 1993: 61), Lincoln (Dobney et al. 1996), and Winchester (Maltby 2010: 148). Unfortunately most metacarpals on Roman sites have been broken to release the marrow, but distal measurements also demonstrate the bias towards smaller (female) cattle in urban assemblages. Examples include York (O’Connor 1988:  55), Exeter (Maltby 1979), and Dorchester (Maltby 1994). If this interpretation is correct, it seems that the procurers of cattle in towns and forts were preferentially acquiring mature but not very elderly female cattle, and/​or farmers were supplying cattle at an age when their productivity in breeding and milking decreased. Assemblages containing a more even distribution of distal metacarpal breadths, which may indicate the presence of higher percentages of adult males, include the rural sites at Owslebury, Hampshire (Maltby 2010:  148), Frocester villa, Gloucestershire (Noddle 2000), and Marsh Leys Farm, Bedfordshire (Maltby 2011). This may indicate that plough oxen were more often likely to have been retained for slaughter on farms after they had ceased working. Evidence for the use of cattle for ploughing is also indicated by the presence of pathological conditions on cattle foot bones. In several samples, including those from Wroxeter (Hammon 2005) and several rural settlements in Bedfordshire, severe pathology tended to be more prevalent in larger specimens. This suggests that more males were affected. Metrical analysis has also shown that some Romano-​British cattle were substantially larger than their Iron Age counterparts. Increases in the average size of cattle from early in the Roman period have been demonstrated in, for example, Hampshire (Maltby 1981) and Essex (Albarella et al. 2008). The latter have also shown that exceptionally large cattle were present at Great Holts Farm, Essex, in the third century ad and have argued convincingly that new stock were introduced to Britain from the continent from the early Roman period. However, there are regional and chronological variations in the presence of larger cattle in Roman Britain. Most of the sites that have produced evidence for substantial numbers of larger cattle are located in the south-​east and the Midlands. Assemblages from south Wales and the south-​west do not show much in the way of size improvements throughout the Roman period (Maltby 1981, 2010; Hammon 2005).

Sheep and Goats Apart from the exceptional case of Uley, Gloucestershire, where goats were selected specifically for slaughter at the temple (Levitan 1993), sheep greatly outnumber goats in assemblages where the two species have been differentiated. In major towns, they have

796

796   Mark Maltby not provided more than 10 per cent of any assemblage, even including horn core counts, and they usually contribute much less than that (Maltby 2010: 268). Sheep are regarded as the mainstay of Iron Age animal husbandry (Albarella 2007), although their percentages in faunal assemblages can vary quite substantially regionally (Hambleton 1999; Albarella 2007). As already noted, they tend to be better represented on non-​villa rural settlements than on other types of Romano-​British settlement (King 1984, 1999) and on early Roman sites, although there is a wide range of variability. Minimum number counts indicate that sheep continued to be the most common animal slaughtered on many sites, possibly including some towns, but, as already discussed, when carcass weights are brought into consideration, lamb and mutton were much less important than beef in the average diet. This is not to say that meat production was not an important consideration in sheep husbandry. Although mortality data show a lot of variation, many faunal assemblages have substantial numbers of sheep killed between 6 months and 3 years of age. Often there is a peak of slaughter of sheep aged between 18 and 36 months, indicating a focus on the culling of sub-​adult and young adult animals once they had reached a good carcass size. Some assemblages, particularly in the later Roman period, have quite high percentages of adult animals, supporting the contention that woollen textile production was a significant consideration in sheep husbandry for some Romano-​British farmers (Wild 2002; Grant 2004: 378; Maltby 2010: 290). Young lambs were slaughtered in quite large numbers on some sites. In some cases, particularly on rural sites, these may simply have been stock surplus to breeding or market requirements. However, as in the case of veal, lamb may have been regarded as a luxury meat. At Colchester, there were high percentages of lambs in several assemblages, particularly from intra-​mural sites (Luff 1993: 73). Relatively high percentages of lambs have also been found in assemblages in the centres of several other towns (Grant 2004: 378; Maltby 2010: 290). Sheep associated with temples and foundation deposits have tended to consist mainly of juvenile and immature animals (King 2005; Maltby 2012). Although sheep as small as the native Iron Age stock continued to be exploited throughout the Roman period, particularly in western and northern areas (Maltby 2010: 294–​295), there is, as in the case of cattle, evidence for significant improvement in the stature of the stock in some areas of the province. The average sizes of sheep were slightly larger in the towns of Gloucester, Cirencester, and Alcester than in Exeter in the south-​ west (Maltby 2001; 2010: 294–​295). Large sheep have been found on several sites in Essex dating from the second century ad onwards (Albarella et al. 2008). Some of these sheep may have derived from stock imported from the continent. Some breeds of sheep have horned skulls, others are hornless. Hornless sheep have been found on a number of Romano-​British sites in areas where only horned sheep were found previously (Maltby 2010: 181). Hornless skulls appear in early Roman features in Winchester and Dorchester (Maltby 1994), indicating that the inhabitants had access to new types of sheep from early in the Roman occupation. Such skulls have not been found on several Iron Age and early Roman rural settlements in their hinterland and appear only in late Roman deposits at Owslebury, where they were associated with large metapodials, implying

The Exploitation of Animals in Roman Britain    797 that hornless sheep were larger than the horned variety (Maltby 1994). Assemblages from later Roman Winchester included a higher proportion of hornless sheep that at Owslebury and they also produced more bones from large sheep (Maltby 2010: 182), suggesting that the hornless type may have been of larger stature. However, it seems unlikely that the two types were kept totally isolated, and the effects of potential hybridization are poorly understood.

Pigs Pigs generally rank third behind cattle and sheep/​goat in most Romano-​British assemblages (King 1999; Grant 2004; Maltby 2010: 264–​265; Hesse 2011). However, there are again some consistent variations in their relative abundance. They tend to be better represented on most military sites, major towns, and some villas than in small towns and other rural settlements (King 1984, 1999). The reasons for this could include dietary preferences, status, and the usefulness of pigs as meat producers for settlements where large amounts of meat were consumed. Taking carcass weights into account, pork products were probably more commonly eaten than lamb and mutton in most military and large urban centres. There are some interesting variations in pig abundance that transcend gross settlement comparisons. For example, pig remains are consistently very well represented in urban deposits in towns in south-​east England, such as Colchester and London, compared with many other urban assemblages elsewhere in England (Maltby 2010: 269). This may reflect that more pigs were kept in the region. It may also reflect a continuation of a late Iron Age phenomenon that saw high levels of pigs on some of the oppida and other trading centres in the region, such as Silchester and Braughing (King 1984; Hambleton 1999; Grant 2000; Albarella 2007). This could be related to well-​ documented continental trading and cultural influences that emerged in that region during the late Iron Age. There is evidence to suggest that pig carcasses were prepared for shipment during that period—​a trade probably facilitated by the increase in salt production (Maltby 2006). Smoked and cured ham and bacon are likely to have been commonly traded products throughout the Roman period. Large-​scale processing of pigs is evidenced, for example, in a dump of pig foot bones discarded at Nazeingbury, Essex (Huggins 1978). The carcasses processed there may well have been destined for the urban market. Another striking pattern is that pigs are often better represented on sites in centrally located sites within towns than in their suburbs. Examples include Caerwent, Exeter, Dorchester, Winchester, and Wroxeter (Maltby 2010: 264–​265). This may again be linked to the greater demand for pork products by those of higher status within the towns. The link between wealth and pig consumption is also evidenced by their high percentages at high-​status sites such as Fishbourne (Grant 2004) and the Winchester Palace site in Southwark (Reilly 2005). In contrast, less prestigious rural settlements generally have

798

798   Mark Maltby fewer pigs than found on urban settlements in their vicinity, as can be demonstrated in the Dorchester and Winchester hinterlands (Maltby 1994, 2010). It would appear that the supply and demand of pork were more heavily focused on urban markets and on some high-​status settlements as well as major military centres. As expected, few Romano-​British sites have produced large percentages of adult animals. Most pigs are killed for their meat while immature. There is again a lot of variability between assemblages, but many include high percentages of pigs killed in their second and third years. Younger piglets have been found less consistently, but form a significant component of some assemblages, again particularly on some (but not all) sites near the centres of some towns such as Caerwent, Silchester, Leicester, and Dorchester (Grant 2004: 379; Maltby 2010: 291). It has been speculated (e.g. Maltby 1994), but never proven, that some pigs were raised within towns. Discussions about the sizes of pigs in Roman Britain are limited. At Heybridge, Essex, Albarella et al. (2008) have demonstrated from measurements of teeth and bones that there were small but significant improvements in the sizes of pigs in the second century ad. Pig distal tibiae breadths in later Roman Winchester were nearly all larger than those from the neighbouring rural settlement of Owslebury (Maltby 2010: 203). Although this may reflect greater numbers of males being consumed in the town, it may also indicate that Winchester had access to pigs (both male and female) of larger stature than those eaten at Owslebury. Similarly large pigs were found in Dorchester (Maltby 1994). MacKinnon (2006) has shown from zooarchaeological, documentary, and pictorial evidence that two types of pig were commonly exploited in Roman Italy: one was hairy and slender, often fattened in woodland; the second tended to be larger, hairless, and raised in sties. Similar variations in husbandry methods and stock types may account for the variations observed in Roman Britain, although much more research is needed on this topic.

Horses and Other Equids Horses of course played an important role within the Roman army (Hyland 1990), and one assumes that many were imported to Britain during the occupation to supplement native stock. At Heybridge, Albarella et al. (2008) have shown a significant increase in the average size of horses from the second century ad onwards. These were comparable in size to a New Forest pony. This is slightly higher than averages obtained from several Roman towns (Maltby 2010: 297). Horses tend to be better represented on rural settlements and suburbs of towns than in their centres, where they rarely form more than 5 per cent of the total cattle and horse assemblage (Maltby 2010: 269–​270). This is because they were much less important as a source of food than cattle, sheep, and pig. They tend to be found in relatively higher frequencies on all types of rural settlement, often forming over 10 per cent of the total cattle and horse assemblage (Maltby 1994), but there is little evidence that they were exploited

The Exploitation of Animals in Roman Britain    799 frequently for food. Although butchery marks have occasionally been found on horse bones on Romano-​British sites, they were exploited for meat much less intensively than cattle. Their bones tend to be less fragmented, and there are many more examples of partial and complete horses being deposited than those of cattle. Horses were primarily valued as a means of transport and as beasts of burden. There are very few examples compared to the other domestic species of immature horse bones on Roman sites, indicating that most could expect to be kept alive for as long as they were considered to be useful unless they died of natural causes. Donkeys and mules have been positively identified on a small number of Roman sites (Johnstone 2010). The potential presence of mules has been overlooked by most zooarchaeologists, although neither mules nor donkeys were likely to have been present in Britain in large numbers.

Dogs and Cats Dogs were commonly kept by the inhabitants of Roman Britain. Their carcasses were occasionally skinned but very rarely eaten. They were kept as pets, as farm animals, as guards, and, judging from the evidence of several mosaics, for hunting. It has long been recognized that the Roman period was one where there was a significant increase in the diversity of dogs. Dogs ranged greatly in stature from animals the size of wolfhounds to those as small as a Yorkshire terrier (Harcourt 1974; Clark 1995). Such diversity indicates the importation of new types. From metrical analysis alone it is impossible to assign the different sizes of dogs to breeds, but there must have been specialist breeding. Dogs, including peri-​natal puppies, have been found commonly as complete or partial skeletons. Sometimes these are isolated burials; in other cases large numbers of dogs were deposited in the same feature (sometimes with other skeletons). Examples of multiple burials have been found in Dorchester, Winchester (Maltby 2010), and Springhead, Kent (Grimm 2008). To what extent these can be regarded as ritual depositions beyond the burial of carcasses not required for processing is the subject of debate that goes beyond the remit of this chapter (see Smith, this volume, for discussion of animals in ritual deposition). Cats are less commonly found than dogs, and, although they have been recorded on many sites, it is only usually in small numbers. They first appeared in Britain in the Iron Age and may have been kept to control vermin as well as for pets in Roman settlements (Kitchener and O’Connor 2010).

Hunted Animals Generally speaking, deer remains occur very infrequently on Romano-​ British sites. Red and roe deer rarely provide over 1 per cent of the food mammal counts on

800

800   Mark Maltby urban sites, and some of these counts include antler collected for working (Maltby 2010: 271). Deer bones appear more frequently (but rarely in high percentages) on some sites of higher status (Allen, forthcoming), supporting the widely held belief that deer-​hunting may have been mainly the prerogative of the upper echelons of Romano-​British society. This is further supported by the introduction of fallow deer to Fishbourne Palace and the subsequent establishment of a herd, presumably for the pleasure of the owners and their guests (Sykes et al. 2011). Confirmed identifications of fallow deer have been restricted to a handful of Roman sites mainly in southern England (Sykes 2010a). Hares have also been found in many Romano-​British faunal assemblages but again usually only in small numbers. They rarely form over 1 per cent of the mammal assemblage, and, where hares are present, it is usually on sites of high status such as Winchester Palace, Southwark (Reilly 2005) or Whitehall villa, Northamptonshire (Sykes 2010b). Authenticated identifications of rabbits in Roman Britain are extremely rare, the most convincing example coming from Beddingham villa, Essex (Sykes and Curl 2010). This was possibly imported from the Mediterranean. Identifications of wild boar are complicated by their similarity to large domestic pigs, particularly if they are killed when immature. They are occasionally recorded, however. One of the best examples comes from the legionary fortress at Caerleon, south Wales. Here a wild boar was eaten with venison and cranes, possibly at a banquet hosted by a high-​ranking officer (Hamilton-​Dyer 1993).

Birds General reviews of birds in Roman Britain can be found in Parker (1988), Yalden and Albarella (2009: 95–​153), and, for large towns, Maltby (2010: 272–​279). The main results of these surveys have shown that domestic fowl (chickens) became significantly more important during the Roman period, although they would not have formed a significant part of the diet. Although introduced to Britain in the Iron Age, their distribution was restricted and their use as a source of food debatable (Sykes 2012). In the Roman period, the evidence suggests that chickens became more ubiquitous but not evenly exploited. Higher percentages of chickens have generally been found in urban and military assemblages than in those from small towns and villas, and they tend to be most poorly represented in non-​villa rural assemblages (Maltby 1997). Again, this would appear to reflect the dietary preferences of different sectors of the Romano-​British population. Chickens were exploited for their eggs as well as their meat and probably used for cockfighting (Sykes 2012). They have also been found in substantial numbers of human graves and associated with other ritual sites such as Uley (Levitan 1993). There is some evidence to suggest that chickens, particularly males, tended to increase in size during the Roman period (Albarella et al. 2008).

The Exploitation of Animals in Roman Britain    801 The domesticated status of ducks and geese in Roman Britain is less clear (Albarella 2005). Ducks and geese of various sizes are present on Roman sites in modest numbers (Yalden and Albarella 2009: 102–​105). The former tends to outnumber the latter, which usually is found more frequently in later Roman sites. Other avian species likely to have been eaten that occur quite frequently on Roman sites include pigeons (some of which may have been domestic), woodcock, and a range of other waders, including crane, plus partridge, and black grouse (especially in northern England). Non-​native pheasants have been recorded on several sites, and a peafowl bone was identified at Portchester Castle (Yalden and Albarella 2009). Other rarities included the butchered wing of a great auk on the Isle of Portland (Maltby and Hamilton-​D yer 2012). Ravens and small corvids are common, but are rarely claimed to have been eaten. The former has often been suggested to be linked with ritual practices (e.g. Fulford 2001).

Fish Locker (2007) has produced a comprehensive survey of fish assemblages found in Roman Britain, and this section largely summarizes those results. The first point to make is that fish remains have been found extremely rarely on late prehistoric sites in southern Britain, despite the increasing number of excavations where sieved sampling has been carried out. This has led to the suggestion that there may have been a taboo on the exploitation of marine resources (Dobney and Ervynck 2007). There is no doubt that fish were more commonly eaten during the Romano-​British period, although the evidence is patchy and handicapped by the inconsistency of sieving. Eel is the species that that has been found most frequently. The distribution of most species tends to reflect whether they were locally available. For example, York has produced relatively high frequencies of freshwater species such as cyprinids (carp family), whereas cod is more common on London sites. Salmon were quite common on sites in the north and the Midlands (Locker 2007). Bream, bass, gadids (cod family), and wrasse caught in the inshore waters off Portland were the species most commonly exploited, as well as eels in Dorchester and neighbouring sites (Maltby and Hamilton-​Dyer 2012). There is little evidence for deep-​sea fishing. Most commentaries on Roman food have highlighted the importance of garum and other fish-​based sauces in the diet. Although they have been found on several Romano-​ British sites (including urban deposits in London, Lincoln, York, and Dorchester) processing sources of small fish locally available (Locker 2007), the occurrences are rare and the importance of this food resource may have been over-​emphasized. Indeed, although isotopic studies attest to the variable presence of marine foods in the diet of some of the inhabitants, these studies have shown that such sources remained a small component of the diet (Redfern et al. 2010).

802

802   Mark Maltby

Summary This brief review of the zooarchaeological evidence has confirmed that the Romano-​ British meat diet was dominated by the acquisition of beef and, to a lesser extent, pork, lamb, and chicken. Other meats including goat, venison, other game, and fish supplemented some diets but rarely contributed significant portions of them. There is some evidence that people of higher status enjoyed more varied diets and perhaps bestowed largesse by hosting banquets that contained rarer meats. There are some indications that those living in towns had access to, or preferred to eat, a wider range of meats than their contemporaries in some rural farms. There is clear evidence that the beef supply to fortresses and large towns in particular relied heavily on the work of specialist butchers. They mainly acquired, or were supplied with, adult cattle, which they intensively processed for sale as both fresh and preserved meat. They could also provide the raw material for large-​scale processing of marrow, hides, horns, and bone, either by themselves or by other specialists. The acquisition of cattle for the urban and military markets must have had a significant impact on traditional means of slaughtering and the distribution of beef. It is less clear to what extent specialist butchers were involved in processing pigs and sheep, and it is feasible that there was a more diverse system of acquisition and distribution of their products. There were changes in general husbandry practices in the Romano-​British period. More sheep were kept until their second and third years before slaughter for meat, and it seems that increasingly more of them were allowed a longer life to produce wool for the expanding textile market in the later Roman period. Some pigs may have become largely confined to sties. Plough cattle possibly became increasingly important on farms, and milk would certainly have been acquired from cattle, goats, and sheep, although there is little evidence for intensive dairy production. Chickens would have been a more common sight during the Roman period, perhaps particularly in and around towns. Through a combination of importation, more controlled breeding, and better husbandry practices, larger cattle, sheep, pigs, horses, and even chickens have been found in some parts of Roman Britain, allowing improvements in meat production in particular. Conversely, Romano-​British people would have encountered a much greater range of dogs than their ancestors. There remains much further research to be carried out on human and animal relationships in the Roman period. The advent of isotopic analysis (e.g. Sykes et al 2006) and genetic studies is already beginning to have a major impact on improving our knowledge of the movements, diet, and ancestry of both animals and humans in this period. Traditional zooarchaeological analysis can embrace these studies to develop deeper understanding of how animals were exploited and perceived by different sections of the community across Roman Britain.

The Exploitation of Animals in Roman Britain    803

References Albarella, U. (2005). ‘Alternate Fortunes? The Role of Domestic Ducks and Geese from Roman to Medieval Times in Britain’, in G. Grupe and J. Peters (eds), Feathers, Grit and Symbolism: Birds and Humans in the Ancient Old and New Worlds. Rahden: Verlag Marie Leidorf, 249–​258 Albarella, U. (2007). ‘The End of the “Sheep Age”: People and Animals in the Late Iron Age’, in C. Haselgrove and T. Moore (eds), The Later Iron Age in Britain and Beyond. Oxford: Oxbow Books, 389–​402. Albarella, U., Johnstone, C., and Vickers, K. (2008). ‘The Development of Animal Husbandry from the Late Iron Age to the End of the Roman Period: A Case Study from South-​East Britain’, Journal of Archaeological Science, 35: 1828–​1848. Allen, M. (forthcoming). ‘Chasing Sylvia’s Stag: Placing Deer in the Countryside of Roman Britain’, in N. Sykes, K. Baker, R. Carden, and R. Madgwick (eds), Deer and People: Past, Present and Future. Oxford: Windgather Press. Bradley, R. (1992). ‘Roman Salt Production in Chichester Harbour:  Rescue Excavations at Chidham, West Sussex’, Britannia, 23: 27–​44. Clark, K. (1995). ‘The Later Prehistoric and Proto-​Historic Dog: The Emergence of Canine Diversity’, Archaeozoologia, 7: 9–​32. Cool, H. (2006). Eating and Drinking in Roman Britain. Cambridge:  Cambridge University Press. de Brisay, K., and Evans, K. (1975) (eds). Salt:  The Study of an Ancient Industry. Colchester: Colchester Archaeology Group. Dobney, K., and Ervynck, A. (2007). ‘To Fish or not to Fish? Evidence for the Possible Avoidance of Fish Consumption during the Iron Age around the North Sea’, in C. Haselgrove and T. Moore (eds), The Later Iron Age in Britain and Beyond. Oxford: Oxbow Books, 403–​418. Dobney, K., Jacques, D., and Irving, B. (1996). Of Butchers and Breeds: Report on Vertebrate Remains from Various Sites in the City of Lincoln. Lincoln: Lincoln Archaeological Studies. Fulford, M. (2001). ‘Links with the Past:  Pervasive “Ritual” Behaviour in Roman Britain’, Britannia, 32: 199–​218. Grant, A. (1975). ‘The Animal Bones’, in B. Cunliffe, Excavations at Portchester Castle. Volume 1: Roman. Society of Antiquaries Research Report 32. London: The Society of Antiquaries, 378–​408. Grant, A. (1989). ‘Animals in Roman Britain’, in M. Todd (ed.), Research on Roman Britain: 1960–​1989. Britannia Monograph Series 11. London: Society for the Promotion of Roman Studies, 135–​146. Grant, A. (2000). ‘Diet, Economy and Ritual Evidence from the Faunal Remains’, in M. Fulford and J. Timby (eds), Late Iron Age and Roman Silchester: Excavations on the Site of the Forum-​ Basilica 1977, 1980–​86. Britannia Monograph Series 15. London: Society for the Promotion of Roman Studies, 425–​482. Grant, A. (2004). ‘Domestic Animals and their Uses’, in M. Todd (ed.) A Companion to Roman Britain. Oxford: Blackwell, 371–​392. Grimm, J. (2008). ‘A Dog’s Life: Animal Bone from a Romano-​ British Ritual Shaft at Springhead, Kent’, in N. Benecke (ed.), Beiträge zur Archäozoologie und Prähistorischen Anthropologie 6. Langenweißbach: Gesellschaft für Archäozoologie und Prähistorischen Anthropologie, 54–​75.

804

804   Mark Maltby Hambleton, E. (1999). Animal Husbandry Regimes in Iron Age Britain. BAR British Series 282. Oxford: Archaeopress. Hamilton-​Dyer, S. (1993). ‘The Animal Bones’, in V. Zienkiewicz, ‘Excavations in the Scamnum Tribunorum at Caerleon’, Britannia, 24: 132–​136. Hammon, A. (2005). ‘Late Romano-​British–​Early Medieval Socio-​Economic and Cultural Change:  Analysis of the Mammal and Bird Bone Assemblages from the Roman City of Viroconium Cornoviorum, Shropshire’. Unpublished Ph.D. thesis, University of Sheffield. Harcourt, R. (1974). ‘The Dog in Prehistoric and Early Historic Britain’, Journal of Archaeological Science, 1: 151–​176. Hathaway, S.-​J. (2006). ‘Poole Harbour: A Review of Early and More Recent Archaeological Investigations with Evidence for Iron Age and Romano-​British Salt Production’, Proceeding of the Dorset Natural History and Archaeological Society, 127 (for 2005), 53–​58. Hesse, R. (2011). ‘Reconsidering Animal Husbandry and Diet in the North-​West Provinces’, Journal of Roman Archaeology, 24: 215–​248. Huggins, P. (1978). ‘Excavation of Belgic and Romano-​British Farm with Middle Saxon Cemetery and Churches at Nazeingbury, Essex, 1975–​6’, Essex Archaeology and History, 10: 29–​115. Hyland, A, (1990). Equus: The Horse in the Roman World. London: Batsford. Ingrem, C. (2006). ‘The Animal Bone’, in M. Fulford, A. Clarke, and H. Eckardt, Life and Labour in Late Roman Silchester. Excavations in Insula IX since 1997. Britannia Monograph Series 22. London: Society for the Promotion of Roman Studies, 167–​88. Johnstone, C. (2010). ‘Donkeys and Mules’, in T. O’Connor and N. Sykes (eds), Extinctions and Invasions: A Social History of British Fauna. Oxford: Windgather Press, 17–​25. King, A. (1984). ‘Animal Bones and the Dietary Identity of Military and Civilian Groups in Roman Britain, Germany and Gaul’, in T. Blagg and A. King (eds), Military and Civilian in Roman Britain. BAR British Series 137. Oxford: Archaeopress, 187–​217. King, A. (1991). ‘Food Production and Consumption—​Meat’, in R. Jones (ed.), Britain in the Roman Period: Recent Trends. Sheffield: J. R. Collis, 15–​20. King, A. (1999). ‘Meat Diet in the Roman World: A Regional Inter-​Site Comparison of the Mammal Bones’, Journal of Roman Archaeology, 12: 168–​202. King, A. (2005). ‘Animal Remains from Temples in Roman Britain’, Britannia, 36: 329–​369. Kitchener, A., and O’Connor, T. (2010). ‘Wildcats, Domestic and Feral Cats’, in T. O’Connor and N. Sykes (eds), Extinctions and Invasions:  A  Social History of British Fauna. Oxford: Windgather Press, 83–​94. Lane, T., and Morris, E. (2001). A Millennium of Saltmaking:  Prehistoric and Romano-​ British Salt Production in the Fenland. Lincolnshire Archaeology and Heritage Report 4. Sleaford: Lincolnshire Archaeology. Levitan, B. (1993). ‘The Vertebrate Remains’, in A. Woodward and P. Leach, The Uley Shrines. English Heritage Archaeological Report 17. London: English Heritage, 257–​345. Locker, A. (2007). ‘In piscibus diversis: The Bone Evidence for Fish Consumption in Roman Britain’, Britannia, 38: 141–​180. Luff, R. (1993). Animal Bones from Excavations in Colchester 1971–​85. Colchester Archaeological Report 12. Colchester: Colchester Archaeological Trust. MacKinnon, M. (2006). ‘Bones, Text and Art in Roman Italy’, in M. Maltby (ed.), Integrating Zooarchaeology. Oxford: Oxbow Books, 52–​59. Maltby, M. (1979). Faunal Studies on Urban Sites:  The Animal Bones from Exeter. Sheffield: University of Sheffield.

The Exploitation of Animals in Roman Britain    805 Maltby, M. (1981). ‘Iron Age, Romano-​British and Anglo-​Saxon Animal Husbandry: A Review of the Faunal Evidence’, in M. Jones and G. Dimbleby (eds), The Environment of Man: The Iron Age to the Anglo-​Saxon Period. BAR British Series 87. Oxford: Archaeopress, 155–​204. Maltby, M. (1984). ‘Animal Bones and the Romano-​British Economy’, in J. Clutton-​Brock and C. Grigson (eds), Animals and Archaeology. Volume 4:  Husbandry in Europe. BAR International Series 227. Oxford: Archaeopress, 125–​138. Maltby, M. (1989). ‘Urban–​Rural Variation in the Butchering of Cattle in Romano-​British Hampshire’, in D. Serjeantson and T. Waldron (eds), Diets and Crafts in Towns. BAR British Series 199. Oxford: Archaeopress, 75–​106. Maltby, M. (1994). ‘The Meat Supply in Roman Dorchester and Winchester’, in A. Hall and H. Kenward (eds), Urban–​Rural Connexions: Perspectives from Environmental Archaeology. Oxbow Monograph 47/​Symposia of the Association of Environmental Archaeologists 12. Oxford: Oxbow Books, 85–​102. Maltby, M. (1997). ‘Domestic Fowl on Romano-​British Sites:  Inter-​Site Comparisons of Abundance’, International Journal of Osteoarchaeology, 7: 402–​414. Maltby, M. (1998). ‘Animal Bones from Romano-​British Deposits in Cirencester’, in N. Holbrook (ed.), Cirencester Excavations V: The Roman Town Defences, Public Buildings and Shops. Cirencester: Cotswold Archaeological Trust, 352–​370. Maltby, M. (2001). ‘Faunal Remains (AES76–​ 7)’, in P. Booth and J. Evans, Roman Alcester: Northern Extramural Area. CBA Research Report 127. London: Council for British Archaeology, 265–​290. Maltby, M. (2006). ‘Salt and Animal Products: Linking Production and Use in Iron Age Britain’, in M. Maltby (ed.), Integrating Zooarchaeology: Proceedings of the Ninth ICAZ Conference, Durham 2002. Oxford: Oxbow Books, 119–​124. Maltby, M. (2007). ‘Chop and Change: Specialist Cattle Carcass Processing in Roman Britain’, in B. Croxford, N. Ray, R. Roth, and N. White (eds), TRAC 2006: Proceedings of the Sixteenth Annual Theoretical Roman Archaeology Conference, Cambridge 2006. Oxford:  Oxbow Books, 59–​76. Maltby, M. (2010). Feeding a Roman Town:  Environmental Evidence from Excavations in Winchester, 1972–​1985. Winchester: Winchester Museums Service. Maltby, M. (2011). ‘Animal Bone’, in M. Luke and T. Preece, Farm and Forge: Late Iron Age/​ Romano-​ British Farmsteads at Marsh Leys, Kempston, Bedfordshire Bedford:  Albion Archaeology, 123–​128. Maltby, M. (2012). ‘Sheep Foundation Burials in Roman Winchester’, in A. Pluskowski (ed.), The Ritual Killing and Burial of Animals: European Perspectives. Oxford: Oxbow Books, 152–​163. Maltby, M., and Hamilton-​Dyer, S. (2012). ‘Big Fish and Great Auks: Exploitation of Birds and Fish on the Isle of Portland, Dorset, during the Romano-​British Period’, Environmental Archaeology, 17: 168–​176. Noddle, B. (2000). ‘Large Vertebrate Remains’, in E. Price, Frocester:  A  Romano-​British Settlement, its Antecedents and Successors. Volume 2: The Finds. Stonehouse: Gloucester and District Archaeological Research Group, 217–​244. O’Connor, T. (1988). Bones from the General Accident Site, Tanner Row. The Archaeology of York: Volume 15, Fascicule 2. York: Council for British Archaeology. Parker, A. (1988). ‘The Birds of Roman Britain’, Oxford Journal of Archaeology, 7: 197–​226. Redfern, R., Hamlin, C., and Athfield, N. (2010). ‘Temporal Changes in Diet: A Stable Isotope Analysis of Late Iron Age and Roman Dorset, Britain’, Journal of Archaeological Science, 37: 1149–​1160.

806

806   Mark Maltby Reilly, K. (2005). ‘Animal Bones’, in B. Yule, A Prestigious Roman Building Complex on the Southwark Waterfront: Excavations at Winchester Palace, London, 1983–​90. London: MoLAS Monograph 23. London: Museum of London Archaeology Service, 158–​166. Rippon, S. (2008). ‘Coastal Trade in Roman Britain:  The Investigation of Crandon Bridge, Somerset, a Romano-​British Trans-​Shipment Port beside the Severn Estuary’, Britannia, 39: 85–​144. Sadler, P. (1997). ‘Faunal Remains’, in A. Marvell and H. Owen-​John, Leucarum: Excavations at the Roman Auxiliary Fort at Loughor, West Glamorgan 1982–​4 and 1987–​8. Britannia Monograph Series 12. London: Society for the Promotion of Roman Studies, 396–​409. Seetah, K. (2006). ‘Multidisciplinary Approach to Romano-​British Cattle Butchery’, in M. Maltby (ed.), Integrating Zooarchaeology. Oxford: Oxbow Books, 109–​116. Stokes, P., (2000). ‘A Cut above the Rest? Officers and Men at South Shields Roman Fort’, in P. Rowley-​Conwy (ed.), Animal Bones, Human Societies. Oxford: Oxbow Books, 145–​151. Sykes, N. (2010a). ‘European Fallow Deer’, in T. O’Connor and N. Sykes (eds), Extinctions and Invasions: A Social History of British Fauna. Oxford: Windgather Press, 51–​58. Sykes, N. (2010b) ‘Worldviews in Transition: The Impact of Exotic Plants and Animals on Iron Age/​Romano-​British Landscapes’, Landscapes, 10: 19–​36. Sykes, N. (2012). ‘A Social Perspective on the Introduction of Exotic Animals: The Case of the Chicken’, World Archaeology, 44: 158–​169. Sykes, N., and Curl, J. (2010). ‘The Rabbit’, in T. O’Connor and N. Sykes (eds), Extinctions and Invasions: A Social History of British Fauna. Oxford: Windgather Press, 116–​126. Sykes, N., White, J., Hayes, T., and Palmer, M. (2006). ‘Tracking Animals Using Strontium Isotopes in Teeth:  The Role of Fallow Deer (Dama dama) in Roman Britain’, Antiquity, 80: 948–​959. Sykes, N., Baker, K., Carden, R., Higham, T., Hoelzel, R., and Stevens, R. (2011). ‘New Evidence for the Establishment and Management of European Fallow Deer (Dama dama) in Roman Britain’, Journal of Archaeological Science, 38: 156–​165. Van der Veen, M., and O’Connor, T. (1998). ‘The Expansion of Agricultural Production in Later Iron Age and Roman Britain’, in J. Bayley (ed.), Science in Archaeology: An Agenda for the Future. London: English Heritage, 127–​143. White, K. (1970). Roman Farming. London: Thames and Hudson. Wild, J. (2002). ‘The Textile Industries of Roman Britain’, Britannia, 33: 1–​42. Yalden, D., and Albarella, U. (2009). The History of British Birds. Oxford:  Oxford University Press.

Chapter 39

A rable Fa rmi ng , Horticu ltu re, a nd  Fo od Expansion, Innovation, and Diversity Marijke Van der Veen

Introduction The production and acquisition of enough food were central to most people’s existence in late Iron Age and Roman Britain, and a consideration of continuity and change in agriculture and food following the Roman conquest thus offers an interesting angle to studying Rome’s impact on life in Britain. Not all aspects of food and farming can be covered in this short chapter, and the reader is referred to the previous reviews, mentioned in the next paragraph, for a full treatment of the topic; here some of the latest evidence will be reviewed. Specifically, this chapter will consider three aspects—​arable production, the spread of horticulture, and the diversification of consumption patterns—​followed by a consideration of the impact of these changes in terms of growing regionality, diversity, and new social realities. Evidence for changes in agricultural production and food consumption can, of course, be observed through a wide variety of sources. Here the focus lies on the archaeobotanical evidence—​that is, on the botanical remains of crops and other foods retrieved from archaeological excavations—​but other evidence will be briefly considered where appropriate (for animal husbandry, see Maltby, this volume). Any discussion of these issues is only as good as the data on which it is based. While a huge amount of archaeobotanical work has been carried out on late Iron Age and Roman period sites in Britain, there are marked regional differences in availability of such data, with western and northern parts of the country particularly poorly covered (reflecting the emphasis of archaeological work carried out in Britain since the 1960s

808

808   Marijke van der Veen which, in turn, is a reflection of modern settlement density and development (Van der Veen et al. 2007). Moreover, most data are found in individual site reports, with few regional syntheses available. Previous reviews of Roman and Iron Age agriculture have been published by Jones (1981, 1989, 1991) and Van der Veen and O’Connor (1998); while Morris (1979), Dark and Dark (1997), Fowler (2002), and Taylor (2007) provide additional discussions of landscape, rural settlement, and farming more generally. The food supply to the army is considered in Hall and Huntley (2007) and Stallibrass and Thomas (2008). Finally, a particular group of foods and crops—​that is, those brought to Britain by the Romans—​is reviewed in Van der Veen (2008) and Van der Veen et al. (2008). Throughout, the term ‘Roman’ or ‘Mediterranean’ is used to refer to foods and foodways that were introduced during the Roman period or brought from the Mediterranean region and/​or Gaul just prior to the Roman conquest. It does not suggest that these foods originated from that region (after all, black pepper is native to India, and peaches originally come from China and South Asia). Even so, the term remains problematic, as it implies some degree of uniformity in such food ways, while it is clear that the empire incorporated a mix of different food traditions, though some communality can be observed. A number of trends can be identified within the three aspects considered here, but these trends are largely relevant to central-​southern and eastern Britain. Variability in practice and response is evident at all levels of analysis: countrywide, within regions, within types of settlement, within particular groups of sites, and within sites, and future studies may usefully focus on this variability, to reveal more clearly the immense complexity of life in Roman Britain.

Arable Farming: A Change of Scale Agricultural responses to the Roman conquest may include changes to the crop repertoire, to the balance between animals and crops and land use, to the scale and mode of production, and regional variation in these. Concentrating here on arable farming and field crops, there is remarkably little change in the range of crops cultivated. Arable farming during the late Iron Age and Roman period was dominated by the cultivation of two crops: spelt wheat (Triticum spelta) and barley (six-​row hulled barley, Hordeum vulgare). Four further cereal crops are known from this period: emmer wheat (Triticum dicoccum), bread wheat (Triticum aestivum), rye (Secale cereale), and cultivated oats (Avena sativa/​strigosa), but these all appear to be minor crops, found on a few sites and in small quantities, though there are regional differences (see later in this section). In some parts of Britain there is additional evidence for the cultivation of peas (Pisum sativum), field/​broad bean (Vicia faba), opium poppy (Papaver somniferum), brassicas (cabbage, rape, turnip, black mustard; Brassica spp.) and flax/​linseed (Linum usitatissimum).

Arable Farming, Horticulture, and Food    809 The real change in arable farming noticeable in this period concerns not so much crop choice, but the scale of production, and this is evident from a range of different sources. In archaeobotanical assemblages, changes in scale can be observed by monitoring regional and/​or temporal variations in the bulk deposition of charred grain or chaff (Van der Veen 2007; Van der Veen and Jones 2007). Such remains are generally found in charred form (see later in this section), and, with the exception of deliberate offerings or destruction owing to conflict, the burning of botanical remains usually occurs in one of three circumstances: (1) when the by-​products of grain dehusking and cleaning are deliberately burnt as either fuel or waste; (2) when an accident occurs during some process involving fire (for example, during parching, drying, or cooking) and (3) when a building containing stored produce catches fire. Most archaeobotanical material originates from (1), the routine practice of preparing grain for consumption (dehusking, cleaning), carried out on a day-​to-​day basis. This practice tends to result in the regular and frequent deposition of small amounts (that is, low densities) of chaff and weed seeds, with only little grain (the grain being consumed). In contrast, the accidents described under (2) and (3) occur much more rarely, but result in sporadic deposition of large amounts of grain (that is, high densities). A rise in the frequency of such high-​density deposits over time suggests a change in the scale of production—​that is, an increase in the bulk handling of grain (processing, storing), which has resulted in an increased likelihood of accidents. Likewise, an increase in the density of chaff remains suggests a move from day-​to-​day processing (to feed the family) to bulk processing (in advance of bulk transport or storage). Thus, a marked rise in either grain-​rich or chaff-​ rich samples points towards a move from a farming system primarily focused on subsistence to one focusing on production for exchange or taxation. Examples of both have been identified in Wessex and the east of England. Among the sites of the Danebury Environs project, the frequency of samples representing accidentally burnt grain increases markedly from the Iron Age to the Roman period (from 6 per cent in the early Iron Age, to 13 per cent in the mid/​late Iron Age, 38 per cent in the latest Iron Age, and 47 per cent in the Roman period (Campbell 2008)). Similarly, Parks (2012: sections 4.6 and 4.7) has identified an increase in samples with dense plant remains (containing > 25 items/​litre of deposit) in the east of England, with the highest levels reached in the middle Roman period. An additional indicator of change is the rise in frequency of germinated grain. Germination may be accidental (poor storage conditions) or deliberate (part of the brewing process), and both signal an increase in the scale of cereal production. Parks (2012: section 4.8, fig. 4.34) has calculated the proportion of samples showing significant evidence of germination (that is, ≥ 20 per cent of grain is germinated) and found that, in the east of England, this rose from 1 per cent in the Iron Age, to 6 per cent in the early Roman, 15 per cent in the mid-​Roman period, then dropping again to 8 per cent. Many other authors report the presence of dense deposits of germinated spelt grain—​ for example, at the nucleated settlement of Catsgore, Somerset (Hillman 1982), the roadside settlement of Springhead, Kent (Stevens 2011), and the Roman villa at Northfleet, Kent (W. Smith 2011).

810

810   Marijke van der Veen Starting with storage, there is now convincing evidence that grain spoiled by inadequate storage was a serious problem in Roman Britain (Figure 39.1(a–​b)). Grain pests (Coleoptera), which thrive in poorly ventilated storage buildings and in grain that is not fully dry when put into storage, make their first appearance in Britain during the Roman period (Smith and Kenward 2011). These grain beetles have not been recorded on Iron Age or earlier sites and are not thought to be native to Britain. They appear from the very start of the Roman conquest, probably as adventitious inclusions in grain brought into Britain by the Roman army during its campaigns. For example, a grain weevil (Sitophilus granarius) was found in a ditch of the annexe to the vexillation fortress at Alchester, dating to ad 44–​45, together with other imports, such as millet and coriander. Grain pests were also discovered in deposits dating to between ad 47–​60 at the 1 Poultry site in central London and at York (Kenward and Williams 1979; Booth et al. 2007: 24, 281; D. Smith 2011). The sudden appearance of these grain pests is probably related to the increased use of large, open grain stores containing bulk quantities of grain, which created environments in which these grain pests could thrive, while the large-​scale trade and movement of grain—​both internal and across the Channel—​facilitated their rapid spread (Smith and Kenward 2011). This would appear to be corroborated by the near absence of these pests from early medieval sites and their reappearance with the return of long-​distance grain transport during the later medieval period—​though, interestingly, never to the levels recorded for the Roman period. Germinated grain is also associated with beer-​brewing. Here the grain is deliberately germinated as part of the ‘malting’ process—​an essential step in the production of beer. Such germinated grain, and the associated detached sprouts (coleoptiles), are often found associated with so-​called corn-​driers, as at the Roman villa at Northfleet, Kent (Figure 39.1(c–​d)) (W. Smith 2011). These structures are primarily a mid-​to-​late Roman phenomenon, but start appearing as early as the first century ad (Morris 1979). Early experiments to determine their function suggested that, rather than drying grain, they may have been more suited to produce malt (Reynolds and Langley 1979; Reynolds 1981). A review of the archaeobotanical assemblages found associated with these structures highlights that they were probably multi-​functional, with evidence for both the drying of grain and the production of malt, and that cereal chaff—​especially spelt glume bases—​was frequently used as fuel in these ovens (Van der Veen 1989). Since that publication, a wealth of further information has become available, with finds of germinated grain and large quantities of burnt spelt chaff confirming the earlier interpretations (e.g. Arrow Valley, Warwickshire (Moffett and Ciaraldi 1999); Showell Farm, Chippenham, Wiltshire (Carruthers 2006); Grateley, Hampshire (Campbell 2008); Springhead, Kent (Stevens 2011) and Northfleet, Kent (W. Smith 2011). At Grateley, three pairs of corn-​ driers were discovered, and in each pair the right-​hand oven showed evidence of more intensive burning than the other, suggesting differential usage. The more intensely heated ovens are thought to have been used to dry spelt grain and the more moderately heated ones to germinate grain and produce malt (Cunliffe 2009). As with grain storage, beer-​brewing is not a Roman invention. Beer is likely to have been produced throughout prehistory but on a household scale, using ordinary vessels

(a)

(b)

(c)

Figure 39.1  (a) Charred spelt wheat grain, showing infestation by a grain weevil from a third/​ fourth century corn drier at Grateley South, Hampshire Photo: Gill Campbell; © English Heritage; reproduced by kind permission of English Heritage;

(b) Charred remains of the granary weevil (Sitophilus granarius L.) from a deposit filled with charred ‘rubbings’ of malted spelt in an early/​mid Roman ditch at Northfleet villa, Kent ([D.] Smith 2011) Photo: David Smith;

(c) Charred germinated spelt grains from a late Roman corn drier at Northfleet villa, Kent ([W.] Smith 2011: Plate 9) © High Speed 1 Ltd; image reproduced with the kind permission of High Speed 1 Ltd;

812

812   Marijke van der Veen (e)

(d)

0

5mm

0.5 mm

(f)

Figure 39.1  Continued(d) Charred detached spelt grain sprouts from a late Roman ditch at Northfleet villa, Kent ([W.] Smith 2011: Plate 10) (© High Speed 1 Ltd; image reproduced with the kind permission of High Speed 1 Ltd);

(e) Charred coriander; one of a cache of more than 1,000 coriander fruits found on the floor of a shop in Colchester that was burnt down during the Boudiccan Revolt of ad 60/61. The material was recovered as part of the excavations carried out at 45-​6 High Street (Murphy 1977). (Photo: Gill Campbell; © English Heritage; image reproduced by kind permission of English Heritage);

(f) Waterlogged olives from a latrine block at the rear of a first century tavern at 1 Poultry, London (Davis 2011: figure 39.275).

(Photo: Andy Chopping; image reproduced by kind permission of Museum of London Archaeology).

and ovens, and thus not easily detectable in the archaeological record. The use of specialized structures and the frequency at which these are found in Roman Britain is indicative of a rural industry. Beer may have represented a ‘cash crop’, where a surplus of grain could be turned into a product that had ‘added value’ and thus could be sold at a profit (Jones 1981). Besides, the quantities of spelt chaff present at these sites highlight

Arable Farming, Horticulture, and Food    813 that the spelt grain was no longer dehusked on a piecemeal, day-​to-​day basis but, instead, bulk-​processed in advance of bulk grain storage or shipment of grain, and the chaff used to fuel the drying/​malting ovens. Thus, the very appearance of corn-​driers—​ specialized structures—​associated with germinated grain and dense spelt chaff deposits indicates that routine household activities such as storing grain and beer-​brewing had—​at some sites at least—​been replaced by large-​scale agricultural operations, tied into bulk storage and long-​distance transport of surplus grain and the production of a cash crop. The appearance of large barns and mills also underlines a growing emphasis on the efficient and bulk processing of cereal crops. For example, at Grateley (Hampshire), two large rectangular buildings were clearly used for crop-​processing, with each containing a pair of corn-​driers, as well as evidence for grain storage, milling, and bread baking (similarly at Darenth, Horton Kirby, and Littlehampton). Examples of watermills have been found at Fullerton (Hampshire), Ickham (Kent), Stanwick (Northamptonshire), and Orton Hall Farm (Cambridgeshire), and animal-​or water-​driven mills (in the form of large mill stones) at Chew Park (Somerset), Heronbridge (Cheshire), Littlecote Park villa at Ramsbury (Wiltshire), and Barton Court Farm (Oxfordshire) (Spain 1984; Fowler 2002: 175; Booth et al. 2007: 298–​303; Cunliffe 2009; Taylor 2011). Technological advances include improvements to ploughs and other tools, in the form of iron coulters, shares and share tips, as well as large iron scythes—​the latter pointing to a greater emphasis on hay-​making, though they may also have been used to harvest other crops (Fowler 2002: ch. 8; Booth et al. 2007: 294). This investment in the agricultural infrastructure does not just point to economic change, but also relates to a wider emphasis on display—​in this case the display of agricultural wealth (Taylor 2011). Additional support for agricultural expansion comes from the Thames Valley, where the higher areas of the floodplain were turned into arable fields, while grassland on the second terrace was ploughed up and parts of the floodplain were converted to hay meadows (Booth et al. 2007: 21–​29, 280, 284). A change in the weed flora associated with the crops has been interpreted as a move onto the heavier clay soils during the late Roman period; something first identified by Jones (1981). At Yarnton (Oxfordshire), the seeds of stinking mayweed (Anthemis cotula)—​a plant that can thrive on calcareous clay soils and heavy loams—​increase from 2 per cent to 8 per cent from the early to the late Roman period, while in the east of England this frequency rises from less than 1 per cent to 9 per cent over the same time period (Parks 2012: figure 6.4). This species has also been found at several other Roman sites (Stevens 2006; Booth et al. 2007: 27; Campbell 2008; W. Smith 2011). While Robinson (1981) suggests this species may represent a Roman import, it has been found at a few Iron Age sites (Parks 2012: section 6.4.3), which highlights the need for a full appraisal of the evidence. Here the importance of studying the weeds becomes clear. While the lack of significant change in the principal cereal crops may suggest continuity, the arable weeds could suggest otherwise. Their life form (annual/​perennial) and ecology (preference for nutrient rich/​poor, acid/​neutral and wet/​dry or heavy/​light soils) help identify the conditions in the arable fields, from which changes in cultivation techniques, arable strategies

814

814   Marijke van der Veen (ploughing, manuring, weeding, and so on), and scales of production can be inferred. Besides, an expansion of agriculture onto heavier clay soils will have demanded a greater use of cattle for traction (and manure), and the heavy work may have increased the occurrence of pathologies on cattle foot bones. The greater importance of cattle in the overall farming system, compared to the Iron Age, is visible in the faunal assemblages (Van der Veen and O’Connor 1998; Grant 2004). There is, currently, little evidence that the mouldboard plough—​which inverts the soil rather than scratch or cut the surface as the ard does—​was introduced prior to the tenth century ad (Fowler 2002: ch. 9; Booth et al. 2007: 286–​288). Systematic regional reviews of changes in arable weed floras and cattle pathology are needed before it can be established how widespread the use of clay soils was during the later Roman period. All these developments emphasize a marked increase in arable production during the Roman period, but it would be wrong to assume that all regions of Britain participated in this process, that all settlements within a region participated, or that this process was continuous. Here the lack of detailed, quantitative, regional studies hampers our ability to offer precise regional definition to the trends observed, but most of the trends identified above concern central-​southern and eastern England, rather than western, south-​western, and northern England, Wales, or Scotland. In these latter regions there is a higher degree of continuity in the nature of arable farming. Besides, in northern England and Scotland barley is the more important crop, while emmer rather than spelt takes second place (Van der Veen 1992; Huntley 2000; Hall and Huntley 2007). Additionally, there are a few areas of England where emmer remains important or reappears—​though no clear pattern or reason has yet been identified (Pelling 2000; Booth et al. 2007: 293; Campbell 2008; Stevens 2011). The role of bread wheat needs further study (a review is in preparation). Samples dominated by bread wheat start to occur from the first century ad onwards but may represent imports rather than British produce. There is a slight increase by the late Roman period, but bread wheat tends to be associated with selected—​possibly mainly high-​status—​sites only (Campbell 2008).

Horticulture: An Innovative New Practice A significant innovation in Romano-​British farming concerns the introduction and development of horticulture—​that is, the cultivation of fruits, vegetables, and herbs in small garden plots and/​or orchards, grown largely for nearby markets rather than just personal consumption. Wild fruits, greens, and roots must have been consumed in prehistoric Britain—​they provide essential vitamins and minerals—​but the archaeobotanical evidence for this remains scarce. The actual cultivation of such foods became—​based on evidence available at the time of writing—​established during the Roman period, as part of a wider introduction of some fifty new food plants. The import and introduction

Arable Farming, Horticulture, and Food    815 of Mediterranean foodstuffs and new food ways into Roman Britain started during the late Iron Age (Hill 2007; Lodwick 2013; see also the section ‘Food: Growing variety and diversity’ below), but saw a significant increase in both range and scale after the conquest. Examples of these newly introduced crops are apple, pear, plum, cherry, walnut, cabbage, turnip, leaf beet, carrot, coriander, dill, and celery. While initially imports—​ that is, foodstuffs brought into the country from elsewhere—​these are so-​called introduced crops because they become incorporated into British agriculture, unlike the imports (which are considered in the section ‘Food: Growing variety and diversity’ below). Detailed discussions of the social and regional uptake of these foods and practices have been published in Van der Veen (2008) and Van der Veen et al. (2008). Before I discuss the evidence, it is important to explain some basic principles of preservation. Dead plant tissues (seeds, grains, chaff, fruit stones) normally decay after a number of years, and some special mechanism of preservation is thus needed to ensure their recovery on archaeological sites. The three most common modes of preservation in Britain are carbonization or charring, waterlogging, and mineral replacement. Crucially, different categories of plant foods tend to be associated primarily with one or other mode of preservation (Figure 39.2). Remains of fruits, vegetables, herbs, and spices are usually recovered in waterlogged form, and any analysis of horticulture and the consumption of fruits, herbs, and vegetables thus tends to rely largely on assemblages of

100

Percentage

80 Mineralized (N = 535)

60

Waterlogged (N = 1,523) 40

Carbonized (N = 2,369)

20

0 Cereals and pulses

Fruits, vegetables, and herbs

Nuts and oil/fibre plants

Figure 39.2  Predominant mode of preservation for the three main categories of food and fibre crops, based on the total number of occurrences of each of these plants in archaeobotanical assemblages from Roman Britain. Source: after Van der Veen (2008); Van der Veen et al. (2008).

816

816   Marijke van der Veen waterlogged plant remains. While such preservation is commonly found on urban—​ and some military—​sites, it is rarely encountered on rural ones; r​ esulting in skewed distributions of sites with these foods, both in terms of site type, geographical area of the country, and chronological period. The rural character of pre-​Roman settlement in Britain, combined with the general lack of waterlogged deposits at rural sites, make it particularly difficult to determine exactly when these new foods first start to appear (see also the section ‘Food: Growing variety and diversity’). Consequently, any analysis of the spread of horticulture and consumption of these types of food does need to take these biases into account (Van der Veen et al. 2007). It is difficult to determine with any certainty exactly when the practice of horticulture starts in Britain; the occurrence of seeds of herbs, stones of fruits, and shells of nuts on archaeological sites may signify either long-​distance trade or horticulture. However, their increasing presence on rural, rather than just urban and military, sites is a good indicator that horticulture was practised. This is particularly the case when seeds of vegetables such as leaf beet, leek, lettuce, or carrot are found, as we eat the leaves or roots rather than the seeds. The seeds of these crops tend to be found at locations of cultivation. This distinction is not clear-​cut, of course. Herbs such as celery, dill, and fennel are utilized for their seeds as well as their leaves, leaf petioles, and stem-​bases, and the seeds of cabbage and turnip (the brassicas) may be used as flavourings or to extract oil, while the leaves and roots of these crops are consumed as vegetables. Many of the newly introduced foods concern plants exotic to Britain (for example, pear, plum, walnut, coriander, dill, leek, onion, cucumber, and lettuce), thus indicating that the Romans brought actual cultivars with them. However, some—​such as celery and apple—​do occur wild in Britain, and here it is theoretically possible (though improbable) that British plants were newly domesticated. For example, celery occurs only in some British salt marshes and thus was most likely grown from imported seeds (Booth et al. 2007: 282). Significantly, the DNA of modern apple cultivars suggests that their wild progenitor is Malus sieversii, a native of the mountain region of Kyrgystan and north-​west China, rather than the British/​European crab apple (Malus sylvestris). This reveals that the Romans brought the cultivated apple to Britain and did not use the local wild variety (Harris et al. 2002). Future DNA analysis of other crops will help clarify the situation further. Most of these new crops are found from the early Roman period onwards (see the section ‘Food: Growing variety and diversity’ for those that occur from the late Iron Age). At rural sites we see an increase over time in the frequencies of fruits such as apple/​ pear (archaeobotanically it is often difficult to distinguish between the two), cherries, plum, damson, and walnut, as well as coriander, celery, mint, carrot, and poppy seed. This suggests that their cultivation was rapidly adopted by local farmers, though evidence for gardening is also found at several military sites (Van der Veen 2008). In terms of geographical distribution there are differences between the fruits and herbs on rural sites: while most herbs, vegetables, and fruits are found across central-​southern and eastern England, this is not the case further north and west, where—​to date—​no herbs have been found and only a few fruits and vegetables (e.g. Van der Veen 2008: figure 11),

Arable Farming, Horticulture, and Food    817 although the meagre database of rural sites with waterlogged preservation makes any such generalizations tentative. Many of the newly introduced herbs—​such as coriander, celery, dill, and summer savoury—​seem to mirror the fortunes of the Roman Empire in Britain: early attraction is witnessed by rare late Iron Age imports; essential requirement is seen in their early Roman presence at military and some urban sites; their growing popularity is evident through increased frequency in the middle Roman period (including on rural sites); while the waning of their appeal is seen in their decline during the late Roman period (Figure 39.3(a)). In contrast, several fruits and vegetables (for example, cherry, plum, walnut, carrot, and leaf beet) display an upward frequency curve (Figure 39.3(b)), becoming increasingly more popular during the course of the Roman period, suggesting they begin to be cultivated in Britain (Van der Veen 2008; Van der Veen et al. 2008). A brief mention of viticulture seems appropriate here. There is one convincing example of a vineyard: at Wollaston in the Nene Valley. Here a complex of parallel trenches was discovered, with low but consistent occurrences of grape (Vitis) pollen in these trenches and with stake holes suggestive of a pastinatio-​type vineyard (Brown et al. 2001). Vineyards have been tentatively identified at a further seven locations (Van der Veen 2008: figure 9(b)), but Wollaston is the only one where remains of grape pollen have been found. How successful this and (the possible) other vineyards were, and whether the fruits were used as table grapes or for wine production, remains uncertain. The prevalence of these crops in central-​southern and eastern England can be partly attributed to the greater volume of archaeological work carried out here and the greater number of sites with waterlogged preservation (Van der Veen et al. 2007: figure 11), but an additional, if not the main, explanation may relate more to the degree of socio-​economic development in this region. Both fruits and vegetables are seasonal, perishable, and not easily stored for long periods of time or transported over large distances (unlike cereal grain and pulses). Large-​scale horticulture or market gardening is, consequently, often found concentrated in or near large centres of population and on important transport routes. The current concentration of evidence for horticulture in central-​southern and eastern parts of the country matches the distribution of towns and evidence for rural development and wealth (villas, corn-​driers, large barns, and so on), suggesting that the logistics of this type of food production was a key factor in its geographical distribution.

Food: Growing Variety and Diversity Food consumption is another area of life where we see significant changes in parts of Roman Britain. A dramatic increase in the choice and variety of available plant foods has been identified, and this is accompanied by marked changes in the range of ceramics; in particular, we see an increase in imported ceramics and the introduction of table wares (e.g. Cool 2006; Pitts 2008). Both are indicative of quite significant changes in

818

(a)

45

Coriander

% of all waterlogged records

40

Celery

35

Dill

30 25 20 15 10 5 0

% of all waterlogged records

(b)

20

Plum

18 16

Walnut

14

Leaf beet

12 10 8 6 4 2 0 Early Roman (N = 71)

(c)

45

% of all waterlogged records

40 35

Middle Roman (N = 64)

Late Roman (N = 63)

Fig Grape Mulberry

30 25 20 15 10 5 0

Figure 39.3  Frequency patterns of selected foods, waterlogged records only. (a) foods that initially increase but then decline; (b) foods that increase over time; (c) foods that decline over time. N = number of records with waterlogged remains. Source: after Van der Veen (2008); Van der Veen et al. (2008).

Arable Farming, Horticulture, and Food    819 local food ways, although, as with the other changes identified in the previous sections, there are marked regional and social differences in their uptake. Prior to the conquest, the diet of the majority of people in Britain consisted of cereal grains (wheat, barley), some pulses (peas, beans), a few oil seeds (flax, poppy), and a restricted number of wild foods (for example, hazelnut, blackberries, and probably some green leaves), as well as animal protein. By the early Roman period, this repertoire had been increased by some fifty new food plants, many of them imports and scarce, but others—​the horticultural crops discussed above—​rapidly becoming incorporated into British agriculture. On current evidence, the majority of these foods are first encountered in the Roman period, but some food imports start to arrive during the late Iron Age (the general lack of waterlogged deposits at rural Iron Age sites makes exact monitoring difficult). For example, wine, olive oil, and fish sauce are manifested by Mediterranean amphorae found in south-​east England, primarily in the graves of wealthy individuals (Hill 2007), while a fig seed is known from the trading centre at Hengistbury Head (Cunliffe 2000: 191–​192). Significantly, recent archaeobotanical research at the late Iron Age oppidum at Silchester has produced olive, celery, coriander, and dill seeds from well deposits dating to c. ad 25–​ 50 (Lodwick 2013). This new evidence mirrors the evidence from Germany, France, and Switzerland for the importation of ‘processed’ (wine, oil) and ‘raw’ (fruits, seeds) foods, prior to the incorporation of the region into the Roman Empire (e.g. Kreuz 2004; Zech-​Matterne et al. 2009). Current evidence suggests that these late Iron Age imports are going to elite locations, in particular to the oppida; highlighting the desire of the late Iron Age elite—​or at least some of them—​to participate in Roman-​style dining practices (Lodwick 2013). Thus, the desire for and acquisition of some exotic ‘Roman-​style’ foods started during the late Iron Age, but a marked increase in both scale and range of imports is noticeable immediately after the conquest; as can be seen in other parts of north-​west Europe. That these foods were brought in from the very start of the conquest is evident from several early Roman sites: millet, coriander, and celery have been recovered at the vexillation fortress at Alchester, dating to ad 44–​45 (Booth et al. 2007: 24, 281); coriander, fig, and walnut, at 1 Poultry, London, dating to ad 45–​53 (Davis 2011); fig, mulberry, and olive at Southwark, London, dating to ad 50–​70 (Giorgi 2009); coriander (Figure 39.1(e)), fig, pine nut, aniseed, lentil, dill, and celery at Colchester dating to ad 60–​61 (Murphy 1977), and fig, grape, walnut, dill, and coriander in deposits at Carlisle dating to ad 72–​105 (Huckerby and Graham 2009). Moreover, many of these foods were not restricted to use by the army; they were quickly available for sale, as is evident from the presence of coriander, anise, fennel, dill, mustard, and poppy seeds in shops at Colchester and London (both burnt down during the Boudiccan revolt of ad 60/​61) and olives (Figure 39.1(f)) and grapes in a latrine at the back of a first-​century ad tavern in London (Murphy 1977; Davis 2011). Fig is particularly common in Roman Britain, being present at 39 per cent of all early Roman sites where waterlogged plants were recovered (Figure 39.3(c)). In addition to these exotics, staples such as grain were also imported. The finds of grain weevils at

820

820   Marijke van der Veen mid-​first-​century Alchester (see section: ‘Arable Farming: A change of scale’ above) and the first-​century legionary fortress at Exeter (Straker et al. 1984), among others, point to the importation of grain from the continent; while the occurrence of some seeds of non-​British species such as einkorn, lentils, and bitter vetch, in a batch of first-​century charred grain in London, also suggests the presence of imported grain, this time probably from the Mediterranean region (Straker 1984). As mentioned in the section ‘Horticulture: An innovative new practice’, the temporal uptake of these exotics varies significantly. Some imports—​for example, black pepper, pomegranate, peach, sesame, lettuce, and black cumin—​remain rare throughout the Roman period and are found almost exclusively in London and at military sites. Those that are more common but cannot easily be grown in Britain—​ for example, fig, grape, mulberry, lentil—​show a downward trend—​that is, their frequencies decline, suggesting that many were popular with the invading army, but their importation did not increase over time (Figure 39.3(c)). Yet, one of these foods, fig, is the most common import found—​possibly because of its ease of transport (in dried form) and the numerous seeds (1,000+ per fruit)—​and, although down from a peak of 39 per cent, it was still present at about 25 per cent of sites by the late Roman period. Other foods increase in popularity after their initial introduction but drop back to early Roman levels by the late Roman period. Examples are many of the herbs (coriander, celery, dill, mint, summer savory (Figure 39.3(a)), as well as the brassicas and parsnip. A final group of foods—​mostly those that can easily be grown in Britain and indeed became incorporated into British farming, such as cherry, plum, walnut, leaf beet, black mustard—​increase steadily throughout the period, though their frequencies remain low (below 15 per cent (Figure 39.3(b)). See Van der Veen (2008) and Van der Veen et al. (2008) for a more detailed discussion of the patterns found. There are regional and social variations as well. To date, most foods have been found on settlements in central-​southern and eastern England, but imports and herbs are absent from rural sites further north, though they are present at the northern military sites. Only certain foods—​apple, plum, cherry, walnut, brassicas, carrot, and parsnip—​ that is, all foods that became incorporated into British farming practice—​have been found on rural sites in northern Britain (Van der Veen 2008: figures 11c–​d). Of note is the social uptake of these foods. Several different consumer groups can be identified. London stands out as a community with access to the widest range of new foods and the greatest variability—​not surprisingly given its special position as a trading centre and its metropolitan character, with a large immigrant population and social mix of people. The other major towns show similar consumption patterns, although with more limited access to the rare imports (which may be a sampling issue). The small towns are very different, however: they resemble the rural sites rather than the major towns as far as food consumption is concerned (Van der Veen 2008: table 5)—​a conclusion also reached by studies of other material culture, such as nail-​cleaners and samian ware (Eckardt 2005; Willis 2005). The military sites form a separate group, with access to most new foods and, in particular, a high proportion of herbs (especially coriander,

Arable Farming, Horticulture, and Food    821 dill, and celery). The abundance of herbs is especially suggestive of the adoption of new practices (see later in this section). The rural inhabitants of Britain can be divided into at least three separate consumer groups. First, there are those people living in south-​east England with access to many of the new foods, including ‘exotics’ such as figs, grapes, pine nuts, and lentils. Strikingly, this group does not consist exclusively of rural elite sites (villas) but includes some nucleated rural sites and small farmsteads, as well as a few minor towns (Van der Veen et al. 2008: figure 7). A second group consists of rural sites, mostly located in the south-​east and again including elite and non-​elite sites, that had access to a selection of the new foods, especially those that later became part of British farming practice (namely, apple, pear, plum, cherry, turnip, leaf beet, parsnip, and carrot). Finally, the third rural group is made up of settlements where no evidence of the new foods has been found, and this group concerns most of those living in northern and western Britain, and many living in central-​southern and eastern Britain. Yet, this group nevertheless made some changes in their diet in that the frequency of native wild foods (for example, hazelnut, blackberry, raspberry, elderberry, sloe) on rural sites increases from the Iron Age to the Roman period, suggesting that they too started eating more fruits and nuts, though in this case local wild ones, rather than introduced cultivated ones (Van der Veen 2008: table 6). The formation of different consumer groups is also apparent from ceramic assemblages and kitchen/​dining utensils more generally (e.g. Cool 2006; Hill 2007; Pitts 2008; Cramp et al. 2011), and from the stable isotopes surviving in human bone (e.g. Richards and Hedges 1998; Redfern et al. 2010; Müldner 2013). This latter evidence supports the conclusions reached above for a significant change in diet, with an increase in dietary breadth, greater diversity, and more status-​related differences in the Roman compared to the Iron Age period. There is some evidence that marine foods (for example, fish sauce, fish, oysters) may represent high-​status foods. This, together with the evidence of selective access to exotic plant foods, brings to light the rise of dietary inequality during the Roman period. Some degree of differentiation between the sexes has also been identified, as has the presence of greater nutritional variety in the towns; the latter also evident from the botanical data (Van der Veen 2008; Müldner 2013). This growing diversity emanates not just from the introduction of new foods, but also from the arrival of migrants from other parts of the Roman Empire (see Ivleva and Eckardt, this volume, for discussions on mobility within Britain and between provinces). At York, for example, the atypical diets of some individuals suggests they originated from North Africa; while one or more of the so-​called Headless Romans at York may have originated from central Europe, as indicated by their high carbon ratio, which suggests the consumption of C4 plants such as millet (Leach et al. 2009; Müldner et al. 2011; Müldner 2013). Millet has never been part of the cereal repertoire in Britain, but a few grains have been found at early Roman deposits at military sites such as Alchester and Carlisle, and in London, always associated with other food imports (Van der Veen et al. 2008; Müldner et al. 2011). Thus some migrants can be identified as such by evidence for a different childhood diet, whereas others brought familiar foods such as millet with them, before switching to foods more widely available in Britain.

822

822   Marijke van der Veen Not all foodstuffs exclusively represent food, of course. Several, especially the herbs and spices, were in antiquity (as today) regularly used in medicine, as herbal remedies and food supplements. Importantly, in antiquity the distinction between food, remedy, and ritual was much less marked than today, if it existed at all. Unfortunately, it is virtually impossible to determine from the archaeobotanical evidence whether or not these plants were used as part of treatments, though there is one instance where the archaeological context hints at a medicinal use. A number of medicinal herbs, including dill (Anethum graveolens), henbane (Hyoscyamus niger), vervain (Verbena officinalis), St John’s wort (Hypericum perforatum), and fenugreek (Trigonella foenum-​graecum), were found in an early first-​century ad context in the hospital of the Roman fort at Neuss, Germany. Interestingly, they were found together with charred grains of rice (Oryza sativa) (Knörzer 1967). Rice is frequently mentioned in ancient sources as having beneficial properties for patients—​being easily digested—​and it was recommended for upset stomachs and intestinal problems, and doctors prescribed a medicinal gruel containing rice to their wealthier patients (Konen 1999; Dalby 2000: 197). Many of these herbs and spices will have offered benefits to the consumer, regardless of whether they were ingested as a remedy or as part of the flavouring of dishes (see Baker, this volume, for a further discussion of medicine). The use of plants in ritual and ceremonial activities is another important facet of their occurrence and dispersal in Britain and north-​west Europe more widely, and botanical remains have at present been recovered from at least twelve cemetery sites and eight shrine/​temple sites in Britain (e.g. Davis 2000; Giorgi 2000; Van der Veen et  al. 2008). The choice of food consumed during burial rituals, donated to the deceased, and/​or burnt as offerings in shrines and temples, concerns not solely imported or newly introduced foodstuffs but also local staples; complex ties between foods and identities are in evidence here (e.g. Bouby and Marinval 2004). For example, pine nuts have often been found associated with ceremonial sites, but are equally and possibly more frequently found in domestic contexts; while, in contrast, date stones are almost exclusively found in high-​status ceremonial contexts and linked to particular mystic cults (Kislev 1988; Livarda 2013). Here, even more clearly than elsewhere, food can be seen to represent a form of material culture, rather than simply essential fuel for the body. Nutrition is important too, of course, and the influx of new foods broadened the nutritional base of the British diet. While wheat and barley provided important carbohydrates and thus energy, the new fruits, vegetables, and herbs contributed vitamin C, beta-​carotene, folic acid, iron, as well as other essential minerals and antioxidants—​all vital to health and nutrition. This supplemented the wild fruits, nuts, and leaves already consumed during the Iron Age, and, intriguingly, there is some evidence that the use of these wild plants increased during the Roman period (Van der Veen 2008). Additionally, the herbs and other foods offered new ways to flavour basic staples such as gruel, stews, and bread. More work is needed to determine in what way all these foods were consumed and how this may have varied across Britain and between and within different social groups; here integrating the plants with the faunal remains and ceramics will be beneficial (see Cool 2006).

Arable Farming, Horticulture, and Food    823 What an analysis of these foods in Roman Britain emphasizes is that simple distinctions between Roman and native, elite and non-​elite, or urban and rural are not helpful in understanding changing food patterns: the situation was far more complex than that. But what is clear is that, broadly speaking, selected groups of people in central-​southern and eastern Britain started adopting new food ways, while there is currently little evidence of this in rural parts of northern and western Britain (see discussion in the next section below). It is also worth stressing here that different degrees of archaeological intervention and different preservation conditions strongly influence these patterns, and that future work may reveal more complex patterns and/​or alter our current view. A full discussion of the complexities is given in Van der Veen (2008) and Van der Veen et al. (2008), while the European evidence is discussed in Bakels and Jacomet (2003), Kreuz (2004), Livarda and Van der Veen (2008), and Livarda (2011, 2013).

Regionality, Diversity, and Materiality This review of the botanical evidence for arable agriculture, horticulture, and food consumption has made it clear that there were many and diverse responses to the Roman occupation and that, in part, these responses have a regional basis, but are additionally linked to social and economic realities within each area. On a very basic level we can see two broad regions, central-​southern and eastern England, where significant changes can be observed, and western and northern Britain, where no such changes are in evidence (though the much reduced level of archaeological intervention here is likely to give a false impression of continuity). It would be wrong, however, to assume uniformity within these regions. In northern and western regions the data set is too small for reliable observations of small-​scale differences in response. In contrast, central-​southern and eastern England has rich archaeological data sets that can highlight complex patterns in both the temporality and social participation of change. For example, the evidence for an increase in the scale of arable production and the creation of market gardening is not restricted to elite rural sites, but neither do all elite sites show evidence for such changes. Similarly, access to the newly imported and introduced foods is not limited to elite sites. In fact, exotic foods and introduced foods are found on some small rural settlements, as well as on nucleated settlements, villas, and small towns. Notably, the villas do not stand out as a distinct consumer group, while most small towns appear more similar to rural sites than major towns in terms of food consumption. Differential sampling and preservation (few/​ many samples, presence/​absence of deposits with waterlogged/​mineral-​replaced preservation) is obviously a major concern here, and future work may change the patterns found; nevertheless, the diversity of practices observed will undoubtedly stand. Explanations for the patterns found are equally complex and diverse. Some relate to the wider economic and political realities of the empire, while others concern more

824

824   Marijke van der Veen local issues. For example, structural changes in the empire’s economy, interruptions in supply, and a reduction in the size of the British garrison during the third century ad, are thought to have affected the volume of trade (Millett 1992: 162–​173; Fulford 2004; Mattingly 2006: 501, 514). Imports (as reflected by amphorae) dropped in this period, as did the occurrence of imported foods (for example, reduced frequencies of figs, grapes, lentils, mulberries, pine nuts, and fennel). The supply did not cease, however; as is clear from the finds of fig, mulberry, pine nut, and olive in late Roman London, fig and lentil in fourth/​early fifth-​century ad Silchester (Robinson 2006), and a near-​complete pine cone (Figure 39.4) from a late-​third/​early fourth-​century ad ditch fill at Clatterford, Isle of Wight (McPhillips 2001). The number of people using the new foods to express a new cultural identity, or aspiration to such an identity, is likely to have changed over time, which may possibly explain the frequency pattern of some of the herbs. These foods are strongly associated with military sites, but their frequencies drop off after the mid-​Roman period. Many of the herbs were probably grown in Britain by that time, and their frequency is thus not affected by changes in trade. While the general pattern is that herbs decline by the late

0

1

2

3

4

5 cm

Figure 39.4  A near complete waterlogged imported pine cone (Pinus pinea) from a late third/​ early fourth century ditch fill containing possible other votive objects at Clatterford Roman villa (McPhillips 2001). © English Heritage; image reproduced by kind permission of English Heritage.

Arable Farming, Horticulture, and Food    825 Roman period, they actually increase at rural sites by that time, but this may perhaps be explained by the move of the elite from the towns to the countryside and the associated rise in rural villas. Thus, here a combination of empire-​wide strategies and local situations influences the pattern we see. Geographical location of individual settlements and general prosperity of a region also influenced people’s chances of participating in the new developments. It is generally agreed that the heavy military presence in northern Britain hindered this region’s economic and social development (Mattingly 2006: 174), which may, partially, explain why we see little evidence of significant agricultural change or adoption of new food ways. Even in central-​southern and western England—​the area with the highest degree of prosperity and economic development—​not all settlements show evidence of significant change. Those that do tend to be located in proximity to towns and/​or major transport routes, which offered easier access to the new foods and awarded economic opportunity (ready contact with individuals requiring or desiring grain, vegetables, and fruits). What is not currently clear is the extent to which these changes were shaped by state-​driven supply networks or market forces (Pitts 2008). As far as the actual foods are concerned, these changes accentuated, in greater measure than before, the role foods play in the creation and maintenance of social identity. The availability of foods strongly associated with the new Roman elite offered new ways of displaying one’s wealth and status. But the wide range of foods now available (true exotics as well as foods quickly incorporated into local agriculture) also meant that a much larger section of British society was able to participate in some degree, and we see this, for example, in the presence of some of these foods on small farmsteads. The rapid spread of certain foods may, in part, be explained by emulation—​the adoption of aspects of Roman culture by the local elite in order to belong to the new political order. But other explanations include ones more focused on economic and social opportunity and an ability to take advantage of the new situation. For example, not everyone will have been able—​financially or psychologically—​to make long-​term investments such as the construction of corn-​driers, storage barns, watermills or the planting of orchards (apple, cherry, plum, and walnut trees, for example, typically take five or more years before they start bearing fruit) or divert labour away from the staple cereals to the production of fruits, herbs, and vegetables. The factors involved in the successful adoption of agricultural innovation are complex (see Van der Veen 2010 for a fuller discussion). The chance of innovations being fruitful will have been higher in the more developed and prosperous south and east of the country than in the north and west, owing to differential access to markets, transport infrastructure, and demand. Thus, increasing levels of diversity and inequality are seen in Roman Britain from a variety of archaeological evidence, and plant foods are no exception in this respect. It is more difficult to establish how the new foods were utilized in cuisine. Here the analysis of the actual foods needs to be combined with studies of kitchen ware, fine wares, and other material culture (e.g. Cool 2006). Were the new foods consumed because people enjoyed the taste or primarily because their consumption signalled

826

826   Marijke van der Veen membership of a particular group? And were these foods incorporated into local cuisine and consumed from local table wares—​that is, a form of creolization—​or were they consumed as part of a new cuisine and using new, Roman fine wares? The varying ways in which Roman ceramics are utilized in different parts of Britain suggests that we should expect not a uniform adoption of ‘Roman’ styles of eating, but rather variable adaptations to suit local circumstances (e.g. Evans 1993; Pitts 2005, 2008; Willis 2011). The very presence of these new foods must have affected how people thought about their food. Were these rejected by some, but perceived as ‘must-​have’ ingredients, regardless of their taste, by others? The arrival of these foods demanded a response and, just as any other type of material culture (fine wares, glass, personal adornments, and so on), created mechanisms through which new social groups were created: by either adopting the new culture or incorporating certain aspects into existing practices. Studying the trajectories of these foods over a longer time span—​that is, into the medieval period—​may help narrow down the specific roles particular foods served in different social agendas and different cultural settings (Livarda and Van der Veen 2008; Livarda 2011; Van der Veen, forthcoming). All these changes will have had an impact on the social fabric of Roman Britain, and social changes may have been located at the household and local community level. The switch from subsistence farming to that of growing a surplus for market exchange (through the sale of surplus grain or beer) impacted on time commitments of individual farmers, as more time was needed (as well as more fields, tools, manure, traction animals, ploughs, barns, corn-​driers, and so on) to plough the fields, manure the soils, sow and weed the crops, harvest and process them (reaping, threshing, dehusking, sieving) and bring them to market, or store them and convert them to beer later in the year. These new day-​to-​day realities, in terms of both time commitments and material settings, will have prompted new social and labour relationships. It meant developing more community-​oriented approaches to the agricultural process, or enlisting more children (larger families?), or taking on farm labourers. In each of these scenarios the social relations of the farmers and their families changed significantly; for example, the latter case meant the adoption of unequal relationships within the farm itself. New social relationships were formed at the point of sale as well, with farmers gaining contacts and contracts with new social groups—​that is, with merchants shipping the grain and other foods or with townspeople if they brought the produce to market themselves. In contrast, the social fabric of society may have changed much less markedly in the north and west of the country. Gender relations are likely to have changed too, especially with the introduction of market gardening. Traditionally, gardens have been located closer to the homestead than arable fields—​to enable more intensive cultivation—​and this proximity to the house may help explain why gardening is perceived by some as a ‘domestic’ activity, carried out foremost by women (Leach 1997). Several authors have identified apparent gender differences between plough agriculture and hoe-​based cultivation (see Van der Veen 2005 for references and discussion of these different practices), and the move towards

Arable Farming, Horticulture, and Food    827 horticulture may, therefore, have meant women becoming involved with the production of fruits, vegetables, and herbs, and thus acquiring new roles—​roles that may also have brought them into contact with the consumers of their produce—​that is, traders passing by the farm, nearby townspeople, or military personnel. The impact of these changes to the native economy was complex and varied, and this may have either benefited or disadvantaged women, depending on the local situation (e.g. Devens 1991). A challenge for the future will be how to start identifying these new relationships in the archaeological record.

Future Work Despite all the work carried out over the last few decades, much remains to be done. The sheer volume of data now available makes new regional surveys essential, and quantification by site type and region of archaeobotanical data, corn-​driers, mills, large barns, tools, ceramic types, and so on, will undoubtedly reveal more complex patterns of change than those visible at the time of writing this chapter. Future work on the ecology of the weeds associated with staple crops—​especially when combined with stable isotope evidence (carbon and nitrogen ratios in the grains themselves)—​will reveal how farmers treated individual crops (degree of manuring, ploughing, weeding, and so on) and how this changed over time in each region. More radio-​carbon dates are needed on ‘late’ occurrences of emmer and ‘early’ ones of bread wheat, rye, and oats. DNA analysis on selected crops will help determine from where these originated and whether any were, instead, developed from local varieties. Once these more detailed studies are integrated with other lines of evidence (field systems, settlement types, transport routes) and other forms of material culture (for example, ceramics and faunal remains), possibly initially on a site-​by-​site basis, we will be able to gain a deeper understanding of Rome’s impact on life in Britain.

Conclusion To conclude, much continuity of practice can be seen in the range of staple crops grown, and in the actual diet of large parts of the population, though this diet was augmented with a greater consumption of wild fruits and nuts. Significant change can be observed in the scale of cereal production, though only in parts of central-​southern and eastern England. Some innovation took place within this field too; in the form of structures and tools that facilitated the increased scale of production and helped display agricultural wealth (better ploughs, scythes, corn-​driers, barns, watermills, hay production, and so on). Significant innovation can be seen in the introduction of horticulture, the cultivation of fruits, nuts, vegetables, and herbs for market. This was made possible by the

828

828   Marijke van der Veen introduction of many new cultivars, plus the know-​how to cultivate and reproduce these crops. Again, this practice is found largely, though not exclusively, in central-​southern and eastern England—​the area where demand, transport facilities, and prosperity coincided. The fact that some fruits and vegetables are found in northern Britain, and the fact that these concern crops that continue to be cultivated in Britain after the withdrawal of the Roman army, brings into focus the fact that these foods were not exclusively desired for their association with Roman elite culture. Innovation is also seen in the diet of some new consumer groups, with the introduction of a vast range of new fruits, nuts, vegetables, and herbs, which brought significant nutritional and health benefits, but also greater variety of taste and cuisine, as well as new social relations. Again we see marked regional and social differences, and at all levels (farming, market-​gardening, and food consumption) we see greater regionality, diversity, and inequality in the Roman period compared to the Iron Age. In material terms this meant greater social and economic diversity, new social relationships, and differentiation between those who took advantage of the new realities and those who did not. Whether the degree of uptake was caused entirely by differences in locally available social and economic opportunities or related more to different degrees of aspiration to ‘become Roman’ remains a matter for future research.

Acknowledgements I would like to thank the Leverhulme Trust for a Major Research Fellowship and the Netherlands Institute for Advanced Study for a NIAS Fellowship during which this chapter was researched. I am grateful to Julie-​Anne Bouchard Perron, Martin Millett, Alison Moore, and Jeremy Taylor for helpful discussions and/​or comments on an earlier draft. Edward Biddulph, Gill Campbell, Andy Chopping, Anne Davis, Rob Goller, David Smith, and Wendy Smith generously offered images for Figure 39.1 and/​or Figure 39.4.

References Bakels, C., and Jacomet, S. (2003). ‘Access to Luxury Foods in Central Europe during the Roman Period: The Archaeobotanical Evidence’, World Archaeology, 34/​1: 542–​557. Booth, P., Dodd, A., Robinson, M., and Smith, A. (2007). Thames through Time: The Archaeology of the Gravel Terraces of the Upper and Middle Thames. The Early Historical Period: ad 1–​1000. Thames Valley Landscapes Monograph 27. Oxford: Oxford Archaeology. Bouby, L., and Marinval, P. (2004). ‘Fruits and Seeds from Roman Cremations in Limagne (Massif Central) and the Spatial Variability of Plant Offerings in France’, Journal of Archaeological Science, 31: 77–​86. Brown, A. G., Meadows, I., Turner, S. D., and Mattingly, D. J. (2001). ‘Roman Vineyards in Britain: Stratigraphic and Palynological Data from Wollaston in the Nene Valley, England’, Antiquity, 75: 275–​287.

Arable Farming, Horticulture, and Food    829 Campbell, G. (2008). ‘Plant Utilization in the Countryside around Danebury:  A  Roman Perspective’, in B. Cunliffe (ed.), The Danebury Environs Programme. A  Wessex Landscape during the Roman Era. Volume 1:  Overview. Oxford University School of Archaeology Monograph 70. Oxford:  English Heritage and Oxford University School of Archaeology, 53–​74. Carruthers, W. (2006). ‘Charred Plant Remains’, in R. Young and A. Hancocks, ‘Early Bronze Age Ring Ditches and Romano-​ British Agriculture at Showell Farm, Chippenham. Excavations in 1999’, Wiltshire Archaeological and Natural History Magazine, 99: 38–​43. Cool, H. E.  M. (2006). Eating and Drinking in Roman Britain. Cambridge:  Cambridge University Press. Cramp, L. J.  E., Evershed, R. P., and Eckardt, H. (2011). ‘What was a Mortarium Used for? Organic Residues and Cultural Change in Iron Age and Roman Britain’, Antiquity, 85: 1339–​1352. Cunliffe, B. (2000). The Danebury Environs Programme. The Prehistory of a Wessex Landscape. Volume I:  Introduction. Oxford University Committee for Archaeology Monograph 48. London: English Heritage and Oxford University Committee for Archaeology. Cunliffe, B. (2009). ‘Continuity and Change in a Wessex Landscape’, Proceedings of the British Academy, 162: 161–​210. Dalby, A. (2000). Empire of Pleasures:  Luxury and Indulgence in the Roman World. London: Routledge. Dark, K., and Dark, P. (1997). The Landscape of Roman Britain. Stroud: Sutton. Davis, A. (2000). ‘Plant Remains’, in B. Barber and D. Bowsher, The Eastern Cemetery of Roman London:  Excavations 1983–​1990. MoLAS Monograph 4.  London:  Museum of London Archaeology Service, 368–​378. Davis, A. (2011). ‘Plant Foods’, in J. Hill and P. Rowsome, Roman London and the Walbrook Stream Crossing: Exacavations at 1 Poultry and Vicinity, City of London. Volume 2. MoLAS Monograph 37. London: Museum of London Archaeology Service, 290–​304. Devens, C. (1991). ‘Gender and Colonization in Native Canadian Communities: Examining the Historical Record’, in D. Walde (ed.), The Archaeology of Gender. Proceedings of the Twenty-​ Second Annual Conference of the Archaeological Association of the University of Calgary. Calgary: The University of Calgary Archaeological Association, 510–​515. Eckardt, H. (2005). ‘The Social Distribution of Roman Artefacts: The Case of Nail-​Cleaners and Brooches in Britain’, Journal of Roman Archaeology, 18: 139–​160. Evans, J. (1993). ‘Pottery Function and Fine Wares in the Roman North’, Journal of Roman Pottery Studies, 6: 95–​118. Fowler, P. (2002). Farming in the First Millennium ad. Cambridge: Cambridge University Press. Fulford, M. (2004). ‘Economic Structures’, in M. Todd (ed.), A Companion to Roman Britain. Oxford: Blackwell, 309–​326. Giorgi, J. (2000). ‘The Plant Remains’, in A. Mackinder, A Romano-​British Cemetery on Watling Street: Excavations at 156 Great Dover Street, Southwark, London. MoLAS Archaeological Study Series 4. London: Museum of London Archaeology Service, 65–​66. Giorgi, J. (2009). ‘Plant Remains’, in C. Cowan, F. Seeley, A. Wardle, A. Westman, and L. Wheeler (eds), Roman Southwark, Settlement and Economy:  Excavations in Southwark, 1973–​91. MoLAS Monograph 42. London: Museum of London Archaeology Service, 101–​ 104, 112–​115, 242–​243 + CD. Grant, A. (2004). ‘Animals and the Economy and Ideology of Roman Britain’, in M. Todd (ed.), A Companion to Roman Britain. Oxford: Blackwell, 371–​392.

830

830   Marijke van der Veen Hall, A. R., and Huntley, J. P. (2007). A Review of the Evidence for Macrofossil Plant Remains from Archaeological Deposits in Northern England. English Heritage Research Department Report Series 87/​2007. London: English Heritage. Harris, S. A., Robinson, J. P., and Juniper, B. E. (2002). ‘Genetic Clues to the Origin of the Apple’, Trends in Genetics, 18: 426–​430. Hill, J. D. (2007). ‘The Dynamics of Social Change in Late Iron Age Eastern and South-​Eastern England c.300 bc–​ad 43’, in C. Haselgrove and T. Moore (eds), The Later Iron Age in Britain and Beyond. Oxford: Oxbow Books, 16–​40. Hillman, G. (1982). ‘Evidence for Spelt Malt’, in R. Leech (ed.), Excavations at Catsgore 1970–​ 73. A  Romano-​ British Village. Western Archaeological Trust Excavation Monograph 2. Bristol: Western Archaeological Trust, 137–​141. Huckerby, E., and Graham, F. (2009). ‘Waterlogged and Charred Plant Remains’, in C. Howard-​ Davis (ed.), The Carlisle Millennium Project, Excavations in Carlisle, 1998–​2001. Volume 2: The Finds. Lancaster: Oxford Archaeology North, 926–​936 + appendix 14. Huntley, J. (2000). ‘The Charred and Waterlogged Remains’, in C. Haselgrove and R. McCullagh (eds), An Iron Age Coastal Community in East Lothian: The Excavation of Two Later Prehistoric Enclosure Complexes at Fisher Road, Port Seton, 1994–​95. Scottish Trust for Archaeological Research. Edinburgh:  Scottish Trust for Archaeological Research, 157–​169. Jones, M. (1981). ‘The Development of Crop Husbandry’, in M. Jones and G. Dimbleby (eds), The Environment of Man: The Iron Age to the Anglo-​Saxon Period. BAR British Series 87. Oxford: Archaeopress, 95–​127. Jones, M. (1989). ‘Agriculture in Roman Britain: The Dynamics of Change’, in M. Todd (ed.), Research on Roman Britain: 1960–​1989. Britannia Monograph Series 11. London: Britannia, 117–​134. Jones, M. (1991). ‘Food Production and Consumption—​Plants’, in R. F. J. Jones (ed.), Britain in the Roman Period: Recent Trends. Sheffield: J. R. Collis Publications, 21–​27. Kenward, H. K., and Williams, D. (1979). ‘Biological Evidence from the Roman Warehouses at Coney Street’, The Archaeology of York, The Past Environment of York, 14/​2: 44–​100. Kislev, M. (1988). ‘Pinus pinea in Agriculture, Culture and Cult’, in H. J. Küster (ed.), Der Prähistorische Mensch und seine Umwelt:  Festschrift für Udelgard Körber-​Grohne. Stuttgart: Konrad Theiss, 73–​79. Knörzer, K.-​ H. (1967). ‘Der Römerzeitliche Heilkräuterfund aus Neuss/​ RH’, Bonner Jahrbücher, 23/​2: 65–​75. Konen, H. (1999). ‘Reis im Imperium Romanum: Bemerkungen zu seinem Anbau und seiner Stellung als Bedarfs–​und Handelsartikel in der Römischer Kaiserzeit’, Münstersche Beiträge zur antiken Handelsgeschichte, 18: 23–​47. Kreuz, A. (2004). ‘Landwirtschaft im Umbruch? Archäobotanische Untersuchungen zu den Jahrhunderten um Christi Geburt in Hessen und Mainfranken’, Bericht der Römisch -​ Germanischen Kommission, 85, 97–​104. Leach, H. M. (1997). ‘The Terminology of Agricultural Origins and Food Production Systems: A Horticultural Perspective’, Antiquity, 71/​271: 135–​148. Leach, S., Lewis, M., Chenery, C. Müldner, G., and Eckardt, H. (2009). ‘Migration and Diversity in Roman Britain:  A  Multidisciplinary Approach to the Identification of Immigrants in Roman York, England’, American Journal of Physical Anthropology, 140: 546–​561. Livarda, A. (2011). ‘Spicing up Life in North-​Western Europe: Exotic Food Plant Imports in the Roman and Medieval World’, Vegetation History and Archaeobotany, 20: 143–​164.

Arable Farming, Horticulture, and Food    831 Livarda, A. (2013). ‘Date, Rituals and Socio-​Cultural Identity in the North-​Western Roman Provinces’, Oxford Journal of Archaeology, 32/​1: 101–​117. Livarda, A., and Van der Veen, M. (2008). ‘Social Access and Dispersal of Condiments in North-​West Europe from the Roman to the Medieval Period’, Vegetation History and Archaeobotany, 17 (Supplement 1), 201–​209. Lodwick, L. (2013). ‘Condiments before Claudius: New Plant Foods at the Late Iron Age oppidum at Silchester, UK’, Vegetation History and Archaeobotany. Published on-​line 9 June 2013. DOI 10.1007/​s00334-​013-​0407-​1. McPhillips, S. (2001). ‘The Objects’, in P. Busby, D. de Moulins, M. Lyne, S. McPhillips, and R. Scaife, ‘Excavations at Clatterford Roman Villa, Isle of Wight’, Proceedings of the Hampshire Field Club and Archaeological Society, 56: 95–​128. Mattingly, D. (2006). An Imperial Possession: Britain in the Roman Empire. London: Allen Lane. Millett, M. (1992). The Romanization of Britain:  An Essay in Archaeological Interpretation. Cambridge: Cambridge University Press. Moffett, L., and Ciaraldi M. (1999). ‘Charred Plant Remains’, in S. C. Palmer, ‘Archaeological Excavations in the Arrow Valley, Warwickshire’, Birmingham Warwickshire Archaeological Society Transactions, 103: 207–​208. Morris, P. (1979). Agricultural Buildings in Roman Britain. BAR British Series 70. Oxford: Archaeopress. Müldner, G. (2013). ‘Stable Isotopes and Diet: Their Contribution to Romano-​British Research’, Antiquity, 87: 137–​149. Müldner, G., Chenery, C., and Eckardt, H. (2011). ‘The “Headless Romans”:  Multi-​isotope Investigations of an Unusual Burial Ground from Roman Britain’, Journal of Archaeological Science, 38: 280–​290. Murphy, P. (1977). ‘45–​46 High Street, Colchester: Carbonised Fruits and Seeds’, AML Reports (Old Series) 4018, English Heritage Research Department Reports (accessed May 2013). Parks, K. (2012). ‘Arable Practice in the Iron Age and Roman East of England’. Unpublished Ph.D. thesis, University of Leicester (accessed May 2013). Pelling, R. (2000). ‘Charred Plant Remains’, in P. Booth and C. Hayden, ‘A Roman Settlement at Mansfield College, Oxford’, Oxoniensia, 65: 324–​328. Pitts, M. (2005). ‘Pots and Pits: Drinking and Deposition in Late Iron Age South-​East Britain’, Oxford Journal of Archaeology, 24/​2: 143–​161. Pitts, M. (2008). ‘Globalizing the Local in Roman Britain: An Anthropological Approach to Social Change’, Journal of Anthropological Archaeology, 27: 493–​506. Redfern, R. C., Hamlin, C., and Beavan-​ Athfield, N. (2010). ‘Temporal Changes in Diet:  A  Stable Isotope Analysis of Late Iron Age and Roman Dorset, Britain’, Journal of Archaeological Science, 37/​6: 1149–​1160. Reynolds, P. (1981). ‘New Approaches to Familiar Problems’, in M. Jones and G. Dimbleby (eds), The Environment of Man: The Iron Age to the Anglo-​Saxon Period. BAR British Series 87. Oxford: Archaeopress, 19–​49. Reynolds, P., and Langley, J. K. (1979). ‘Romano-​British Corn-​Drying Oven: An Experiment’, Archaeological Journal, 136: 27–​42. Richards, M. P., and Hedges, R. E. M. (1998). ‘Stable Isotope Analysis Reveals Variations in Human Diet at the Poundbury Camp Cemetery Site’, Journal of Archaeological Science, 25: 1247–​1252.

832

832   Marijke van der Veen Robinson, M. (1981). ‘The Iron Age to Early Saxon Environment of the Upper Thames Terraces’, in M. Jones and G. Dimbleby (eds), The Environment of Man: The Iron Age to the Anglo-​ Saxon Period. BAR British Series 87. Oxford: Archaeopress, 251–​277. Robinson, M. [with N. Fulford and K. Tootell] (2006). ‘Chapter 5:  The Macroscopic Plant Remains’, in M. Fulford, A. Clarke, and H. Eckardt (eds), Life and Labour in Roman Silchester:  Excavations in Insula IX since 1997. Britannia Monograph Series 22. London: Britannia, 206–​218, 374–​379. Smith, D. (2011). ‘Insects from Northfleet’, in C. Barnett, J. I. McKinley, E. Stafford, J. M. Grimm, and C. J. Stevens (eds), Settling the Ebbsfleet Valley: High Speed I Excavations at Springhead and Northfleet, Kent. The Late Iron Age, Roman, Saxon and Medieval Landscape. Volume 3: Late Iron Age to Roman Human Remains and Environmental Reports. Oxford/​ Salisbury: Oxford/​Wessex Archaeology, 88–​91. Smith, D., and Kenward, H. (2011). ‘Roman Grain Pests in Britain:  Implications for Grain Supply and Agricultural Production’, Britannia, 42: 243–​262. Smith, W. (2011). ‘Charred Plant Remains from Northfleet’, in C. Barnett, J. I. McKinley, E. Stafford, J. M. Grimm, and C. J. Stevens (eds), Settling the Ebbsfleet Valley: High Speed I Excavations at Springhead and Northfleet, Kent. The Late Iron Age, Roman, Saxon and Medieval Landscape. Volume 3: Late Iron Age to Roman Human Remains and Environmental Reports. Oxford/​Salisbury: Oxford/​Wessex Archaeology, 105–​113. Spain, R. J. (1984). ‘Romano-​British Watermills’, Archaeologia Cantiana, 100: 101–​128. Stallibrass, S., and Thomas, R. (2008). Feeding the Roman Army: The Archaeology of Production and Supply in North-​west Europe. Oxford: Oxbow Books. Stevens, C. J. (2006). ‘Charred Plant Remains’, in M. G. Fulford, A. B. Powell, R. Entwistle, and F. Raymond, Iron Age and Romano-​British Settlements and Landscapes of the Salisbury Plain. Wessex Archaeological Reports 20. Salisbury: Wessex Archaeology, 152–​158. Stevens, C. J. (2011). ‘Charred Plant Remains from Springhead’, in C. Barnett, J. I. McKinley, E. Stafford, J. M. Grimm, and C. J. Stevens, Settling the Ebbsfleet Valley:  High Speed I  Excavations at Springhead and Northfleet, Kent. The Late Iron Age, Roman, Saxon and Medieval Landscape. Volume 3: Late Iron Age to Roman Human Remains and Environmental Reports. Oxford/​Salisbury: Oxford/​Wessex Archaeology, 95–​105. Straker, V. (1984). ‘First and Second Century Carbonized Cereal Grain from Roman London’, in W. van Zeist and W. A. Casparie (eds), Plants and Ancient Man. Rotterdam: Balkema, 323–​329. Straker, V., Robinson, M., and Robinson, E. (1984). ‘Biological Investigations of Waterlogged Deposits in the Roman Fortress Ditch at Exeter’, Proceedings of the Devon Archaeological Society, 42: 59–​69. Taylor, J. (2007). An Atlas of Roman Rural Settlement in England. CBA Research Report 151. London: Council for British Archaeology. Taylor, J. (2011). ‘The Idea of the Villa: Reassessing Villa Development in South-​East Britain’, in N. Roymans and T. Derks (eds), Villa Landscapes in the Roman North: Economy, Culture and Lifestyles. Amsterdam: Amsterdam University Press, 179–​194. Van der Veen, M. (1989). ‘Charred Grain Assemblages from Roman-​Period Corn Driers in Britain’, Archaeological Journal, 146: 302–​319. Van der Veen, M. (1992). Crop Husbandry Regimes: An Archaeobotanical Study of Farming in Northern England, 1000 bc–​ad 500. Sheffield: J. R. Collis. Van der Veen, M. (2005). ‘Gardens and Fields: The Intensity and Scale of Food Production’, World Archaeology, 37/​2: 157–​163.

Arable Farming, Horticulture, and Food    833 Van der Veen, M. (2007). ‘Formation Processes of Desiccated and Carbonised Plant Remains:  The Identification of Routine Practice’, Journal of Archaeological Science, 34: 968–​990. Van der Veen, M. (2008). ‘Food as Embodied Material Culture: Diversity and Change in Plant Food Consumption in Roman Britain’, Journal of Roman Archaeology, 21: 83–​110. Van der Veen, M. (2010). ‘Agricultural Innovation: Invention and Adoption or Change and Adaptation?’, World Archaeology 42/​1: 1–​12. Van der Veen, M. (forthcoming). ‘The Dispersal of Cultivated Fruits, Vegetables and Herbs in Roman and Medieval Britain’. Van der Veen, M., and Jones, G. (2007). ‘The Production and Consumption of Cereals: A Question of Scale’, in C. Haselgrove and T. Moore (eds), The Later Iron Age of Britain and Beyond. Oxford: Oxbow Books, 419–​429. Van der Veen, M., and O’Connor, T. (1998). ‘The Expansion of Agricultural Production in Later Iron Age and Roman Britain’, in J. Bayley (ed.), Science in Archaeology: An Agenda for the Future. London: English Heritage, 127–​143. Van der Veen, M. Livarda, A., and Hill, A. (2007). ‘The Archaeobotany of Roman Britain: Current State and Identification of Research Priorities’, Britannia, 38: 181–​210. Van der Veen, M., Livarda A., and Hill A. (2008). ‘New Food Plants in Roman Britain: Dispersal and Social Access’, Environmental Archaeology, 13/​1: 11–​36. Willis, S. (2005). ‘Samian Pottery, a Resource for the Study of Roman Britain and Beyond: The Results of the English Heritage-​Funded Samian Project. An e-​monograph. Internet Archaeology 17 (accessed May 2013). Willis, S. (2011). ‘Samian Ware and Society in Roman Britain and Beyond’, Britannia, 42: 167–​242. Zech-​Matterne, V., Bouby, L., Bouchette, A., Cabanis, M., Derreumaux, M., Durand, F., Marinval, P., Pradat, B., Sellami, M.-​F. and Wiethold, J. (2009). ‘L’Agriculture du VIe au Ier siècle avant J.-​C. en France: État des recherches carpologiques sur les établissements ruraux’, in I. Bertrand, A. Duval, J. Gomez de Soto, and P. Maguer (eds), Habitats et paysages ruraux en Gaule et regards sur d’autres régions du monde celtique. Tome II. Chauvigny: Association des Publications Chauvinoises (Mémoire XXXV), 383–​416.

834

Chapter 40

C oinag e a nd the Ec onomy Philippa Walton and Sam Moorhead

Introduction Scholars of Roman coinage in Britain are blessed with a numismatic record that is unlikely to be rivalled by any other province of the Roman Empire. Over the past 200 years, more than 2,570 coin hoards have been found, and the numbers continue to rise, particularly since the introduction of Treasure legislation in 1997. These hoards are complemented by records of more than 180,000 stray losses and site finds collected throughout England and Wales by the Portable Antiquities Scheme (www.finds.org.uk) and the Iron Age and Roman Coin Finds from Wales Project (Guest and Wells 2007) (see Figure 40.1) as well as more than 450 assemblages found during research excavations and development control projects published both individually and as summaries (cf. Reece 1991; Walton 2012). Most research undertaken with these numismatic data has focused on the quantification, classification, and publication of individual hoards and assemblages. This has led to the compilation of a range of invaluable reference works for the identification of Roman coins; most notably the Cunetio and Normanby hoard catalogues (Besly and Bland 1983; Bland and Burnett 1988) and Robertson’s Inventory of Romano-​British Coin Hoards (2000). In recent years, there has also been a growing interest in the numerical and statistical analyses of coin assemblages from archaeological sites, known collectively as ‘applied numismatics’ (Casey and Reece 1974; Lockyear 2000; Walton 2012). These analyses have led to a recognition that there are variations in distribution patterns of coin loss and have engendered a new appreciation of the potential of numismatic data for exploring the socio-​economic development of the province (Davies and Gregory 1991; Moorhead 2001; Guest 2008b; Walton 2012). However, the focus on the publication of hoards and site assemblages has led to the critical exploration of the use of coinage in Roman Britain being somewhat neglected.

Figure 40.1  Roman coins recorded by the Portable Antiquities Scheme. Source: © Philippa Walton.

836

836    Philippa Walton and Sam Moorhead Although there have been some detailed discussions regarding the role of money in Romano-​British society (cf. Walker 1988: 301 and passim; Reece 2002: 107 and passim), there has been a tendency to concentrate on the development of a ‘monetary economy’ rather than on a consideration of the non-​monetary roles it might acquire or represent. An ‘evolutionist’ model continues to prevail (Aarts 2005: 41), situating monetization as one aspect of a wider process of acculturation or ‘Romanization’. As such, native societies throughout the province moved from pre-​monetary systems of barter and social exchange to a single monetized economy, as a direct result of integration within the Roman Empire (Aarts 2005: 8). Hence, at the beginning of the Roman period, the use of coinage was intimately linked to the incoming army and provincial administration (Greene 1986: 61; Boon 1988: 118; Esmonde Cleary 1989: 8; Guest 2008a: 139), with military installations and the nascent urban centres representing ‘coin using islands… in an overwhelming sea of virtually coinless peasants’ (Abdy 2002b: 14). However, by the mid-​second century other sectors of society had begun to employ coinage and it was embraced even in rural areas (Reece 1988c: 37; Walker 1988: 288 ff.; Moorhead 2010: 157). By the late third century ad, the transition from pre-​monetary to monetary economy was complete and, until the collapse of Roman Britain in the early fifth century ad, functioned in a manner similar to a modern capitalist economy (Reece 1988c: 102; Esmonde Cleary 1989: 96; Millett 1990: 169; Mattingly 2006: 497). The apparent ubiquity of coinage throughout the province has also led to the assumption that processes of monetization and coin use were uniform and should therefore be analysed and interpreted at this level. Hence, scholars have worked to produce average profiles for coin loss in Britain such as ‘The British Mean’ and ‘The PAS Mean’ (Reece 1995; Walton 2012), and inter-​provincial comparisons have been made (Reece 1987: 91). Although undoubtedly useful in some respects, this approach obscures the sheer variation in patterns of coin loss throughout the province, which some regional studies have highlighted (Davies and Gregory 1991; Lockyear 2000; Moorhead 2001; Guest 2008b; Walton 2012). More emphasis needs to be placed on an investigation of patterns of coin loss at a regional level, as they are likely to indicate a new layer of complexity. This chapter will attempt to convey some of this complexity. Although employing a chronological framework, it is not intended to rival the exhaustive summaries of all aspects of coinage used in Roman Britain provided by Reece (1987) and Moorhead (2013). Instead, it will outline some of the prevailing theories concerning the monetization of the province and will use new data recorded by the Portable Antiquities Scheme to re-​evaluate these theories and to explore the role of Roman coinage in Romano-​British society more broadly. Furthermore, it will explore the applicability of the prevailing ‘evolutionist’ model and the relevance of studying coinage from Britannia at a provincial level.

Roman Coin Use in the Late Iron Age The earliest Roman coins found in Britain are Republican silver denarii, of which a large number have been recorded, both in hoards and as site finds (Robertson 2000; Walton

Coinage and the Economy    837 2012). Owing to their longevity of circulation and the tendency to overplay the impact of the Roman invasion in ad 43, these denarii have invariably been interpreted as representing military pay. Their distribution patterns are used to trace the manœuvres of the army throughout the mid to late first century ad, even contributing to discussions regarding the route taken by the invading force (cf. Orna-​Ornstein 1997; Frere and Fulford 2001). While it is certainly true that a large proportion of Republican coinage arrived and was lost after ad 43, there is some evidence that Iron Age societies, particularly in southern Britain, had familiarity with, and access to, these coins during the first and possibly even the second century bc. This evidence is archaeological, iconographic, and metallurgical in nature. Excavations of the temple complex on Hayling Island, Hampshire (Haselgrove 2005: 386 ff.), and Elm’s Farm, Humberstone, Leicester (Haselgrove 2009), offer examples of Republican coins probably deposited before ad 43, while PAS data provide tantalizing hints that Republican issues may have been deposited within the prehistoric ritual landscape of the Witham Valley, Lincolnshire (Walton 2012: 68), throughout the Isle of Wight (Walton 2012: 128) and in Berkshire (Portable Antiquities Scheme 2014, BERK-​65D307) in the late pre-​Roman Iron Age (Figure 40.2). Furthermore, numerous issues of the late first century bc struck in the south-​east, including those of Cunobelin and Tincommius, incorporate iconographic themes of contemporary Roman coinage (Scheers 1992: 34), and Republican denarii appear to have been the primary source of metal used in the manufacture of the silver denominations of several regions, including those attributed to the Atrebates (Northover 1992: 257; Fulford 1989: 178) and the Corieltavi (Julia Farley, pers. comm.). Although it is possible to establish that some Republican denarii reached late Iron Age Britain, it is more difficult to ascertain the mechanism(s) by which these coins arrived and how they were subsequently employed within Iron Age societies. Strabo

Figure 40.2 BERK-​65D307—​a Roman Republican denarius issued in c. 207 bc found in Berkshire. Source: © Portable Antiquities Scheme.

838

838    Philippa Walton and Sam Moorhead states that Britain was known for its exports of corn, cattle, hides, slaves, gold, silver, iron, and hunting dogs (Strabo IV.5.3), and so it is possible that these coins represent payment for exported commodities. However, recent research has also stressed the importance of gift-​giving in the creation of political alliances between Rome and native elites (Creighton 2006: 42; Hunter 2007: 218), and therefore their presence may represent not only trade and exchange, but also the articulation of Roman diplomatic policy. Once they were in ‘circulation’, it is possible that these coins served a similar function to contemporary Iron Age coinage, which may have been used in ‘non-​commercial’ payments, to emphasize ties of clientage and other social obligations, while also acting as convenient stores of wealth (Haselgrove 1987; Hunter 2007: 221).

Roman Coin Use in the First Century ad Although the invasion of ad 43 did not mark the beginning of Roman coin use in Britain, the arrival of the Roman army en masse and the establishment of the provincial administration significantly increased the quantity of Roman coinage in the province during the first century ad. It is estimated that the campaigning army required 8,500,000 asses (or 531,250 denarii) every four months to pay its soldiers (Boon 1988: 118), and, therefore, it comes as no surprise that so much first-​and early second-​century coinage has been found within military installations, associated urban foundations, and their immediate hinterlands (Lockyear 2000: 403, 413; Guest 2008a: 139; Walton 2012: 62 and passim). Military sites as remote from one another as Swanton Morley, in Norfolk, and Dodderhill, in Herefordshire, possess comparable coin profiles (Davies and Gregory 1991: 71), while the wide geographical spread of Republican to Flavian coinage in Wales (Guest 2008b) and the west Midlands (Walton 2012: 62 and passim) are likely to reflect the militarization of the landscape during first-​century campaigning. Once these coins had been distributed as military pay, establishing their ‘biographies’ until they were lost or deposited is no simple task. The concentrations of finds within military installations suggest that soldiers spent at least some of their income within a network of relatively closed monetary economies provided by forts—​presumably using bronze denominations in lower-​value transactions for goods and services such as food, alcohol, and sex. Indeed, spatial analysis of the distribution of coins from second-​century forts on the Antonine Wall has demonstrated that low-​value bronze denominations were lost by soldiers within the bathhouse, presumably while getting changed, eating, drinking, and gambling, while silver denarii were lost around the headquarters’ building where military pay and savings were kept (Abdy 2002a). However, as early Roman coinage is not recovered solely from military contexts (Walton 2012: 71), some monetary interactions with the native population obviously took place, particularly after the initial phase of invasion. Whether these interactions included the use of coinage as a medium for exchange in everyday transactions at the level of the individual or army unit, the retention of denarii as stores of precious metal, or the payment of taxes within the new

Coinage and the Economy    839 Roman taxation system is open to interpretation. It is not clear, for example, how often resources were requisitioned, compulsorily purchased, or just taken by the army in the immediate post-​conquest period. Imitations of Claudian bronze dupondii and asses, known as ‘Claudian copies’, are frequently cited as evidence of early native adoption of Roman coinage and, by implication, monetary exchange with the army. These were presumably minted in the mid-​first century, and the relatively crude execution of the copies is used, along with the recitation of punitive legislation against the counterfeiting of coin, to suggest that they were of native, rather than Roman, manufacture (Sutherland 1935: 25; Boon 1988: 118). While further research is needed to ascertain the exact function of these copies, this interpretation is a clear example of the assumption of a colonial perspective, based on aesthetic judgement rather than on distribution patterns. Indeed, huge numbers of Claudian copies are found at first-​century forts and fortresses such as Richborough, Colchester, and Usk (Kenyon 1987, 1991; Boon 1988: 118 and passim; Harper 2010; Walton 2012: 79 and passim). Coupled with a clear justification for military need following the cessation of production of bronze coinage at the mint of Lyons in ad 42 (Boon 1988: 119), all this suggests military manufacture. More compelling are the epigraphic data provided by the Vindolanda tablets. These documents, recovered from excavations of a fort in Northumberland, detail cash payments for foodstuffs and commodities purchased by military personnel on the northern frontier in the late first and early second century ad (Bowman 2003). As a result, they are frequently cited as evidence of a swift native transition from barter to monetary exchange throughout the province, even in a frontier region (Bowman 2003: 66). However, it is important both to emphasize the dangers of using information from a single site to generalize at a provincial level and to consider other ways in which this epigraphic material could be interpreted. For example, it is striking that the Vindolanda tablets use the denarius (and fractions thereof) as the main unit of accounting, whereas elsewhere in the empire it appears that the sestertius was commonly employed (Reece 1987: 32; van Heesch 2007: 80). This quirk in accounting practice may be related to the fact that military salaries were paid and deductions taken from their salaries in denarii (cf. Speidel 1992: 90) and that this was simply what was available. Alternatively, it might reflect the demands of a local population who viewed denarii not as one denomination in a monetary system but as convenient stores of precious metal. Tacitus implies that Roman coinage was used in much this way on the Germanic frontier, where the native population was reputed to pick out older, silver Roman coinage for use in trade (Tacitus, Germania 5, 3–​5), and this is amply confirmed by evidence from the area outside the empire. For example, finds from native sites in Scotland show a strong preference for silver denarii over bronze small change (Hunter 2007: 218). Denarii were regarded as one of many prestige goods by the native population in Scotland (Hunter 2007: 221) and acted as tokens for specific specialized transactions, for storing wealth, and for displaying status (Hunter 2007: 218). Although the situation remains ambiguous, what is clear is that any generalization is unwise and that patterns of coin use are likely to have varied ‘from place to place and

840

840    Philippa Walton and Sam Moorhead from time to time’ (Reece 1988b: 58). Indeed, there are marked regional variations in both the quantity of first-​century coins found and the denominations represented. Perhaps most striking is the dichotomy between north and south. In the north, the presence of coinage appears inextricably linked with the presence of the Roman army. Very few coins are recorded at any distance from forts, and the record is dominated by the silver denarius (Walton 2012: 50). In the south, the quantity of coins recorded is far larger, with the full range of denominations represented. Do these differences reflect the separation of the province into ‘military’ and ‘civilian’ zones and the implementation of a monetary taxation system in the ‘civilian’ zone or merely indicate a pre-​existing familiarity with, and willingness to use, coinage in southern Britain? It is impossible to be certain. However, similar denominational patterning in coin assemblages from ‘military’ and ‘civilian’ zones has been noted elsewhere. For example, long-​established ‘civilian’ provinces such as Gaul, Belgica, and Italy received more bronze coinage than ‘military’ ones, such as Upper and Lower Germany, Raetia, and Pannonia (Hobley 1998: 128).

Roman Coin Use in the Second and Early Third Century ad The second century ad is often regarded as a period of monetary consolidation, in which further sectors of Romano-​British society were gradually welcomed into the monetary fold. The increase in the size of the coinage circulation pool is argued to have brought coinage within the reach of more people, while the appearance of Antonine bronze coinage on rural sites is interpreted as reflecting a tentative expansion of the monetary economy into the countryside (Moorhead 2010: 157). However, the stability of Roman coinage in the mid-​Roman period is a major impediment to any investigation of the processes of monetization at this time. Hoards demonstrate the longevity of circulation of all denominations, with denarii of Mark Antony and bronze coins of Vespasian available for deposition well into the second and as late as the third century ad (Reece 2002: 43). For example, a purse hoard from Birdoswald with a terminus post quem of ad 121 is dominated by Republican, rather than contemporary, issues (Robertson 2000: 24). As a result, any observations regarding where and why coinage was being used in the second century ad must be approached with a degree of caution. Furthermore, there is some indication that the apparent ubiquity of coinage in the second century ad is illusory. There was certainly a greater quantity of coinage in Britain than in the previous century; but, even so, it has been calculated, using the assemblage from the Sacred Spring at Bath, that by the ad 150s each person in the province might have had access only to approximately two sestertii (Walker 1988). Richard Reece estimated a range of between 0 and 30 sestertii (Reece 2002: 115). Either way, these totals do not represent anything resembling monetization or everyday coin use. As Reece (2002: 115) commented: ‘Coins and coin use were not universal… the coins are not there

Coinage and the Economy    841 and their values are skewed way above the needs of the farm labourer wanting basic food… ’. Bearing this in mind, the large quantities of coins recovered from Romano-​British temple sites and from votive deposits have a huge significance for understanding the function of coinage. Indeed, at sites such as the Sacred Spring, Bath (Walker 1988), Coventina’s Well, Northumberland (Allason-​Jones and Mackay 1985), and Piercebridge, County Durham (Walton 2008), offerings of coinage represent a significant proportion of the objects recovered and appear in far greater numbers than at contemporary settlement sites. This may suggest that the act of votive deposition was restricted to certain sectors of the population with access to coins, or that coinage was regarded as having the same specialized religious functions it appears to have done in the late Iron Age. Although this function is often described as being ‘non-​economic’ in nature, this is certainly not the case, as votive offerings represent economic transactions with the gods, rather than with men. The composition of these assemblages may also give us an insight into the nature of these transactions. At Bath, large numbers of lower-​value bronze denominations were chosen for deposition, as well as a number of contemporary copies. These copies, which imitate Neronian to Hadrianic asses, form a coherent group thought to have been made by a ‘forging mint’ (Walker 1988: 291). Given their context, it is possible that these bronze pieces were made specifically for deposition as votive offerings and may have been exchanged with pilgrims for genuine issues of a higher value. This theory is supported by evidence from the riverine votive deposit at Piercebridge, County Durham, where numerous lead–​alloy copies of Severan denarii have been found, of which several have been folded and squeezed with tongs, so that they resemble miniature curse tablets. It is possible that these copies also represent a specific votive product (Walton 2008). If this is the case, a major reinterpretation of the function of other contemporary copies in all periods is needed. Rather than always representing local demand for coinage, copies may have served specialized religious functions associated with votive deposition.

Roman Coin Use in the Late Third Century ad While the mid-​Roman period is generally regarded as a period of monetary consolidation, the late third century ad is almost unanimously seen as the beginning of an explosion of coin use that continued until the end of the Roman period (Walker 1988: 306). With the demise of the Augustan coinage system and the introduction of the radiate coinage, there was a massive increase in the volume of coins produced. As a result, debased coinage of the period ad 260–​74 is extremely common in Britain, both in hoards such as Cunetio, Normanby, and Frome (Figure 40.3) (Besly and Bland 1983; Bland and Burnett 1988; Moorhead et al. 2010) and as site-​finds. Indeed, 14.4 per cent of

842

842    Philippa Walton and Sam Moorhead

Figure 40.3  The Frome hoard being excavated. Source: © Somerset County Museums Service.

all coins recorded in Reece (1991) and 12.2 per cent of all coins recorded by Walton (2012) dated to the period ad 260–​75. This plentiful supply is argued to have brought coinage within the reach of all rural populations. Instead of functioning as a unit of taxation and a mechanism for paying the army and administration, coinage was at last embraced by the rural marketplace and used in everyday exchange (Reece 1988b: 102; Esmonde Cleary 1989: 96; Millett 1990: 169; Mattingly 2006: 497). However, although radiates are found in the countryside, they are proportionally more common on urban sites than rural ones (Reece 1987: 91 ff.; 1988b: 103; Davies and Gregory 1991: 76). This does suggest that their primary function was associated with some facet of urban life, whether that be related to the collection of taxes or to the processing of commercial transactions in the marketplace, or some combination of the two. Indeed, the demand for coinage for use in commercial transactions, coupled with a sudden reduction in supply after the demise of the Gallic Empire in ad 274, is usually deemed to have been responsible for the enormous production of locally produced copies, which are universally referred to as ‘barbarous radiates’. These coins are uniform neither in design nor in size, and, as a result, they have tended to be interpreted as wholly unofficial issues produced locally by opportunist forgers. However, as with Claudian copies, do subjective judgements regarding their aesthetic value, exemplified by the use of the label ‘barbarous’, cloud the ability to see radiate copies as official or quasi-​official

Coinage and the Economy    843 products? If ‘barbarous radiates’ were an unofficial response to the lack of coin supply from the continent, how did the army and provincial administration respond? By what mechanisms did soldiers and civil servants continue to be paid efficiently and effectively? The phenomenon of hoarding in the late third century ad may also provide further comment on how coinage was being used at this time. More than 650 hoards of radiates are known from Britain, more than for the rest of the empire put together. Hoards such as Cunetio with 54,951 coins (Besly and Bland 1983) and Frome with 52,503 coins (Moorhead et al. 2010) illustrate the enormous number of radiates circulating in the province. Primarily because of the size of these hoards, scholars have interpreted them as representing ‘pay-​chests’ or savings hoards (Johns 1996: 6 and passim). Conversely, the willingness to deposit such large quantities of coins may indicate that they had little or no monetary value (at least outside of the religious sphere) and should be compared with earlier votive deposits of coins such as those from the Sacred Spring, Bath, Coventina’s Well, and Piercebridge. Indeed, there are some strong indications that the Frome hoard was votive in nature, buried in a single episode near an ancient water source (Moorhead et al. 2010). Therefore, it is possible that, as radiates had little value in the rural marketplace, they were offered to the gods as part of numinous transactions (Walton, forthcoming).

Coin Use in the Fourth Century ad Following an explosion in the production of coinage in the late third century, the full integration of the Romano-​British countryside into the ‘monetary economy’ is generally agreed to have occurred in the fourth century ad (Reece 1972, 1988c, 1995; Walker 1988: 304; Lockyear 2000: 403). Nearly every rural site with coinage produces a nummus dating to the period ad 330 to 348. However, an increase in coin production need not necessarily equate with an increase in the number of users. Indeed, it is becoming increasingly clear that, even in the fourth century, coin use was not a province-​wide phenomenon. As in earlier centuries, sites where coinage was used and lost were situated, for the most part, in southern and eastern Britain (Walton, forthcoming). Nor was coin use a long-​lived phenomenon. As the fourth century progressed, although people may have continued to use earlier coinage, the area that maintained its coin supply shrank to a narrow band of land stretching from the West Country through Northamptonshire to the western edge of East Anglia, with a few outliers to the north (Walton 2012: 105; forthcoming). This pattern certainly does not accord with an evolutionist model of coin use in Roman Britain. While the appearance of large quantities of coinage in rural areas in the early fourth century ad may reflect the embrace of a ‘monetary economy’, it may also be a reflection of the de-​centralization of military and administrative structures. Towns appear to have been in decline, and officials and military units may instead have been stationed

844

844    Philippa Walton and Sam Moorhead throughout the countryside to extract taxes, such as the annona militaris (Moorhead 2001: 94 ff; Moorhead and Stuttard 2012: 226 and passim). Although the archaeological evidence for instability on the continent as the result of barbarian invasion is ambiguous, contemporary literary sources strongly indicate an increased interest in the export of Britain’s agricultural resources (Moorhead and Stuttard 2012: 171 ff.). With this in mind, it is worth noting the striking similarity in the distribution of rural sites using coinage in the late Roman period and the find-​spots of late Roman belt fittings (Laycock 2008: 115; Worrell 2012: 390). These fittings have been interpreted as the insignia of civilian administrators and individual soldiers (Leahy 2007).

Coinage and the End of Roman Britain The study of coinage in the late fourth and early fifth century ad is inextricably linked with an exploration of the fate of the Diocese—​the group of late Roman provinces into which Britannia was divided at this time. As there is so little material culture specifically dated to the period, coinage has often been relied upon to provide reliable dating evidence for events or developments, with a particular emphasis on the chronology of the collapse of Roman Britain. Despite the brief expansion of coinage into rural areas in the fourth century ad, by the Theodosian period (ad 388–​402) coin-​using activity had overwhelmingly contracted back to urban centres, military installations, and settlements situated at nodal points on major arteries (Walton 2012: 106)—​in fact, the very types of site that had used coinage at the beginning of the Roman period. This might suggest that the experiment in de-​centralization had failed and that safety was sought in accessible, yet defensible locations. In a further echo of the early Roman period, variation in the denominations of coins being used at sites with different functions becomes apparent. While late Roman coin assemblages from urban and military sites such as Canterbury, Caerwent, and Richborough are dominated by bronze nummi, silver siliquae are proportionally more common on rural settlement sites (Bland et al. 2013; Moorhead and Walton 2014). It is possible that, although something approaching a tri-​metallic ‘monetary economy’ existed within towns and forts, the rural population continued to select and retain precious metal coinage for its intrinsic value, as they had several centuries before, rather than to use as money. Bronze coinage stopped arriving in Britain soon after ad 395 following the closure of the western mints, although occasional later bronze nummi have been found in Britain, including pieces from Hadrian’s Wall (Collins 2008), St Albans, Wroxeter, Dunstable, and Richborough (Abdy and Williams 2006: 31 and passim). In ad 402 and ad 406 respectively, the silver and gold supply diminished to a trickle (Bland and Loriot 2010; Moorhead and Walton 2014). The muted response to the cessation of coin supply is illuminating. Although some copies of nummi were made, such as those found in the Bishop’s Canning hoard (Guest et al. 1997), there was not the spate of copying seen in

Coinage and the Economy    845

Figure 40.4  The Coleraine Hoard (County Antrim) deposited in the first half of the fifth century ad. Source: © The Trustees of the British Museum.

earlier periods, which may suggest there was little demand for or use of bronze denominations. However, the widespread practice of clipping the circumference of silver siliquae does suggest an attempt to make the best out of the dwindling supply of silver, with the clippings being used to strike siliqua copies, or melted down to form ingots. Who was responsible for clipping these coins and for how long is a matter of some debate (Burnett 1984; Guest 2005; Abdy 2009; Walton 2012; Bland et al. 2013), although again it is likely that there was a significant official element in the process. By ad 430, it is likely that any surviving vestige of a monetary economy in Roman Britain, such as there was, had collapsed. It is likely that any Roman coinage remaining in Britain was regarded as just another form of bullion, in much the same way as the siliquae in contemporary hoards from Barbaricum, such as Coleraine (Figure 40.4) and Traprain Law, appear to have been (Kent and Painter 1977: 123; Robertson 2000: 405 and passim; Hunter and Painter 2013).

Conclusion Many questions regarding the history of coin use in Roman Britain remain unanswered. However, it is clear that the Roman occupation did not provide the impetus for

846

846    Philippa Walton and Sam Moorhead a province-​wide conversion from systems of barter and social exchange to a single monetary economy. Regional and site-​specific variation in patterns of coin loss clearly illustrate that the ‘evolutionist’ model of monetization should not hold sway. Instead, there appears to be a strong case for the military, administrative, and religious functions of Roman coinage in Britain remaining essentially unchanged and unchallenged for more than 400 years of Roman rule (Walton 2012: 169).

Acknowledgements The authors’ thanks are extended to Richard Abdy, Roger Bland, Andrew Burnett, Kris Lockyear, Ian Leins, Eleanor Ghey, Julia Farley, and Julian Bowsher for their information and advice.

References Primary Sources Strabo, Geography, trans. H. L. Jones. Loeb Classical Library. London: Heinemann, 1923. Tacitus, Agricola, Germania and Dialogus de Oratoribus, trans. W. Peterson and M. Hutton. Loeb Classical Library. London: Heinemann, 1914.

Secondary Sources Aarts, I. (2005). ‘Coins, Money and Exchange in the Roman World: A Cultural–​Economic Perspective’, Archaeological Dialogues, 12/​1: 1–​28. Abdy, R. (2002a). ‘A Survey of the Coin Finds from the Antonine Wall’, Britannia, 33: 189–​217. Abdy, R. (2002b). Romano-​British Coin Hoards. Princes Risborough: Shire Publications. Abdy, R. (2009). ‘Stanchester, Wiltshire’, in R. Abdy, E. Ghey, C. Hughes, and I. Leins (eds), Coin Hoards from Roman Britain XII. Wetteren: Moneta, 366–​374. Abdy, R. A., and Williams, G. (2006). ‘A Catalogue of Hoards and Single Finds from the British Isles, ad 410–​675’, in B. Cook and G. Williams (eds), Coinage and History in the North Sea World, c.500–​1250: Essays in Honour of Marion Archibald. Northern World 19. Leiden: Koninklijke Brill, 11–​73. Abdy, R., and Minnit, S. (2002). ‘Shapwick Villa, Somerset’, in R. Abdy, I. Leins, and J. Williams (eds), Coin Hoards from Roman Britain XI. London: Royal Numismatic Society Special Publication 36, 169–​234. Allason-​Jones, L., and Mackay, B. (1985). Coventina’s Well:  A  Shrine on Hadrian’s Wall. Hexham: Trustees of the Clayton Collection. Besly, E., and Bland, R. (1983). The Cunetio Treasure. London: British Museum Press. Birley, A. R. (2005). The Roman Government of Britain. Oxford and New  York:  Oxford University Press. Bland, R., and Burnett, A. (1988). The Normanby Hoard and other Coin Hoards: Coin Hoards from Roman Britain VIII. London: British Museum Press.

Coinage and the Economy    847 Bland, R., and Loriot, X. (2010). Roman and Early Byzantine Gold Coins Found in Britain and Ireland with an Appendix of New Finds from Gaul. Royal Numismatic Society Special Publication 46. London: Royal Numismatic Society. Bland R., Moorhead, T. S. N., and Walton, P. J. (2013). ‘Finds of Late Roman Silver Coins from Britain: The Contribution of the Portable Antiquities Scheme’, in F. Hunter and K. Painter (eds), Late Roman Silver and the End of Empire:  The Traprain Law Treasure in Context. Edinburgh: Society of Antiquaries of Scotland, 117–​166. Boon, G. C. (1988). ‘Counterfeit Coins in Roman Britain’, in J. Casey and R. Reece (eds), Coins and the Archaeologist. London: Seaby, 102–​188. Bowman, A. K. (2003). Life and Letters on the Roman Frontier: Vindolanda and its People. 3rd edn. London: British Museum Press. Burnett, A. M. (1984). ‘Clipped siliquae and the End of Roman Britain’, Britannia, 15: 163–​168. Casey, J., and Reece, R. (1974). Coins and the Archaeologist. BAR British Series 4. Oxford: Archaeopress. Collins, R. (2008). ‘The Latest Roman Coin from Hadrian’s Wall: A Small Fifth-​Century Purse Group’, Britannia, 39: 356–​361. Creighton, J. (2006) Britannia: The Creation of a Roman Province. Abingdon: Routledge. Davies, J. A., and Gregory, T. (1991). ‘Coinage from a “civitas”: A Survey of the Roman Coins found in Norfolk and their Contribution to the Archaeology of the ‘Civitas Icenorum’, Britannia, 22: 65–​101. Esmonde Cleary, S. (1989). The Ending of Roman Britain. London: B. T. Batsford. Frere, S., and Fulford, M. (2001). ‘The Roman Invasion of ad 43’, Britannia, 32: 45–​55. Fulford. M. (1989). ‘The Economy of Roman Britain’, in M. Todd (ed.), Research on Roman Britain 1960–​89. Britannia Monograph Series 11. London: Society for Promotion of Roman Studies, 195–​211. Greene, K. (1986). The Archaeology of the Roman Economy. London: B. T. Batsford. Guest, P. (2005). The Late Roman Gold and Silver Coins from the Hoxne Treasure. London: British Museum Press. Guest, P. (2008a). ‘Coins’, in P. Booth, A-​M. Bingham, and S. Lawrence (eds), The Roman Roadside Settlement at Westhawk Farm, Ashford, Kent, Excavations 1998–​ 9. Oxford Archaeology Monograph Series 2. Oxford: Oxford Archaeology, 35–​139. Guest, P. (2008b). ‘The Early Monetary History of Roman Wales:  Identity, Conquest and Acculturation on The Imperial Fringe’, Britannia, 39: 33–​58. Guest, P., and Wells, N. (2007). Iron Age and Roman Coins from Wales. Collection Moneta 66. Wetteren: Moneta. Guest, P., Bland, R., Orna-​Ornstein, J., and Robinson, P. (1997). ‘Bishops Cannings (Blagan Hill), Wiltshire’, Coin Hoards from Roman Britain X. London: British Museum Press, 426–​462. Harper, P. (2010). ‘Stylistic Links among Copies of Bronze Coins of Claudius I Found in Britain and Spain’, Numismatic Chronicle, 170: 105–​114. Haselgrove, C. (1987). Iron Age Coinage in South-​East England: The Archaeological Context. BAR British Series 174. Oxford: Archaeopress. Haselgrove, C. (2005). ‘A Trio of Temples: A Reassessment of Iron Age Coin Deposition at Harlow, Hayling Island and Wanborough’, in C. Hazelgrove and D. Wigg-​Wolf (eds), Iron Age Coinage and Ritual Practices. Studien zu Fundmünzen der Antike 20. Mainz: Verlag Philipp von Zabern, 381–​418. Haselgrove, C. (2009). ‘Small Change:  Two Recent Coin Finds from Humberstone (Leicestershire) and Long Distance Contacts in Later Iron Age Britain’, in G. Cooney, K.

848

848    Philippa Walton and Sam Moorhead Becker, J. Coles, M. Ryan, and S. Sievers (eds), Relics of Old Decency: Archaeological Studies in Later Prehistory. Festschrift for Barry Raftery. Dublin: Wordwell, 249–​258. Heesch, J. van (2007). ‘Some Aspects of Wage Payments and Coinage in Ancient Rome, First to Third Centuries ce’, in J. Lucassen (ed.), Wages and Currency: Global Comparisons from Antiquity to the Twentieth Century. International and Comparative Social History 10. Bern: Peter Lang, 77–​96. Hobley, A. S. (1998). An Examination of Roman Bronze Coin Distribution in the Western Empire ad 81–​192. BAR International Series 688. Oxford: Archaeopress. Hunter, F. (2007). ‘Silver for the Barbarians: Interpreting denarii Hoards in Northern Britain and Beyond’, in R. Hingley and S. Willis (eds), Roman Finds: Context and Theory. Proceedings of a Conference held at the University of Durham. Oxford: Oxbow Books, 214–​224. Hunter, F., and Painter, K. (2013). Late Roman Silver:  The Traprain Treasure in Context. Edinburgh: Society of Antiquaries Scotland. Johns, C. (1996). ‘The Classification and Interpretation of Romano-​ British Treasures’, Britannia, 27: 1–​16. Kent, J. P. C., and Painter, K. S. (1977). Wealth of the Roman World ad 300–​700. London: British Museum Press. Kenyon, R. (1987). ‘The Claudian Coinage’, in N. Crummy (ed.), Colchester Archaeological Report 4:  The Coins from Excavations in Colchester 1971–​ 9. Colchester:  Colchester Archaeological Trust, 24–​41. Kenyon, R. (1991). ‘The Copying of Bronze Coins of Claudius I in Roman Britain’. Unpublished Ph.D. thesis, University College London. Laycock, S. (2008). Britannia—​the Failed State: Ethnic Conflict and the End of Roman Britain. Stroud: The History Press. Leahy, K. (2007). ‘Soldiers and Settlers in Britain, Fourth to Fifth Century—​Revisited’, in M. Henig and T. J. Smith (eds), Collecteana Antiqua: Essays in Memory of Sonia Hawkes. BAR International Series 1673. Oxford: Archaeopress. Leins, I. (2011). ‘The Coins’, in V. Score (ed.), Hoards, Hounds and Helmets:  A  Conquest-​ Period Ritual Site at Hallaton, Leicestershire. Leicester Archaeology Monograph 21. Leicester: University of Leicester, 39–​60, 175–​277. Lockyear, K. (2000). ‘Site Finds in Roman Britain:  A  Comparison of Techniques’, Oxford Journal of Archaeology 19(4), 397–​423. Mattingly, D. J. (2006). An Imperial Possession: Britain in the Roman Empire. London: Penguin. Metcalf, W. (2012) (ed.). The Oxford Handbook of Greek and Roman Coinage. New York: Oxford University Press. Millett, M. (1990). The Romanization of Britain:  An Essay in Archaeological Interpretation. Cambridge: Cambridge University Press. Moorhead, T. S. N. (2001). ‘Roman Coin Finds from Wiltshire’, in P. Ellis (ed.), Roman Wiltshire and After. Devizes: Wiltshire Archaeological and Natural History Society, 85–​105. Moorhead, T. S. N. (2010). ‘Expanding the Frontiers: How the Portable Antiquities Scheme Database Increases Knowledge of Roman Coin Use in England’, in S. Worrell, G. Egan, J. Naylor, K. Leahy, and M. Lewis (eds), A Decade of Discovery: Proceedings of the Portable Antiquities Conference 2007. BAR British Series 520. Oxford: Archaeopress, 138–​155. Moorhead, T. S.  N. (2013). Roman Coinage in Britain:  The Contribution of the Portable Antiquities Scheme. Witham: Greenlight Publishing. Moorhead, T. S. N., and Stuttard, D. (2012). The Romans who Shaped Britain. London: Thames and Hudson. Moorhead, T. S. N., and Walton, P. J. (2014). ‘Coinage at the End of Roman Britain’, in F. Haarer with R. Collins, K. Matthews, S. Moorhead and P. Walton (eds), AD 410: The History and

Coinage and the Economy    849 Archaeology of Late Roman and Post-​Roman Britain. London:  Society for Promotion of Roman Studies. Moorhead, T. S.  N., Booth, A., and Bland, R. (2010). The Frome Hoard. London:  British Museum Press. Northover, J. P. (1992). ‘Materials Issues in the Celtic Coinage’, in M. Mays (ed.), Celtic Coinage: Britain and Beyond. The Eleventh Oxford Symposium on Coinage and Monetary History. BAR British Series 222. Oxford: Tempus Reparatum, 253–​299. Orna-​Ornstein, J. (1997). ‘Early Hoards of denarii from Britain’, in R. Bland and J. Orna-​ Ornstein (eds), Coin Hoards from Roman Britain X. London: British Museum Press, 23–​29. Portable Antiquities Scheme (2014). (accessed 1 March 2014). Reece, R. (1972). ‘A Short Survey of the Roman Coins Found on Fourteen Sites in Britain’, Britannia, 3: 269–​276. Reece, R. (1987). Coinage in Roman Britain. London: Seaby. Reece, R. (1988a). ‘Numerical Aspects of Roman Coin Hoards in Britain’, in J. Casey and R. Reece (eds), Coins and the Archaeologist. London: Seaby, 86–​101. Reece, R. (1988b). My Roman Britain. Cirencester: Cotswold Studies. Reece, R. (1988c). ‘Coins and Villas’, in K. Branigan and D. Miles (eds), The Economy of Romano-​ British Villas. Sheffield: Sheffield University Press, 34–​41; repr. in Reece (2003), 109–​117. Reece, R. (1991). Roman Coins from 140 Sites in Britain. Cotswold Studies 4. Cirencester: Cotswold Studies. Reece, R. (1995). ‘Site Finds in Roman Britain’, Britannia, 26: 179–​206. Reece, R. (2002). The Coinage of Roman Britain. Stroud: Tempus. Reece, R. (2003). Roman Coins and Archaeology: C ​ ollected Papers. Wetteren: Moneta. Robertson, A. S., with Hobbs, R., and Buttrey, T. V. (2000) (eds). An Inventory of Romano-​ British Coin Hoards. Royal Numismatic Society Special Publication 20. London:  Royal Numismatic Society. Scheers, S. (1992). ‘Celtic Coin Types in Britain and their Mediterranean Origins’, in M. Mays (ed.), Celtic Coinage: Britain and Beyond: The Eleventh Oxford Symposium on Coinage and Monetary History. BAR British Series 222. Oxford: Tempus Reparatum, 235–​299. Speidel, M. A. (1992). ‘Roman Army Pay Scales’, Journal of Roman Studies, 82: 87–​106. Sutherland, C. H.  V. (1935). Romano-​British Imitations of Coins of Claudius I. American Numismatic Notes and Monographs 65. New York: American Numismatic Society. Walker, D. R. (1988). ‘The Roman Coins’, in B. Cunliffe (ed.), The Temple of Sulis Minerva at Bath. Volume 2: The Finds from the Sacred Spring. OUCA Monograph 16. Oxford: Oxford University Committee for Archaeology, 281–​358. Walton, P. J. (2008). ‘The Finds from the River’, in H. E. M. Cool and D. Mason (eds), Roman Piercebridge:  Excavations by D.  W. Harding and Peter Scott 1969–​ 1981. Durham:  The Architectural and Archaeological Society of Durham and Northumberland, 286–​293. Walton, P. J. (2012). Rethinking Roman Britain: Coinage and Archaeology. Collection Moneta 136. Wetteren: Moneta. Walton, P. J. (2014). ‘Where, when and what for? Coin Use in the Romano-​British Countryside’, in M. Henig and B. Soffe (eds), Roman Villas: A Retrospective Review. Oxford: Oxbow Books. Worrell, S. (2012). ‘Finds Reported under the Portable Antiquities Scheme’, Britannia, 43: 355–​393.

850

Chapter 41

Ec onom y and P ow e r i n L ate Rom an Bri ta i n James Gerrard

Economics and power are two sides of the same coin. Their inter-​relationship is self-​ evident in the contemporary world, with its banks that are too big to fail and now jail (e.g. Sorkin 2009). The ability to mobilize and deploy economic resources also played an important part in creating and maintaining power relationships in the Roman Empire. At an imperial scale, this connection is perhaps most clear-​cut when considering the fall of the western empire. Many scholars would argue that the collapse of Roman power in the west was not caused by conflict in Gaul or court intrigue in Ravenna, but by the loss of tax revenue from North Africa, following its capture by the Vandals (Heather 2005; Ward-​Perkins 2005). What follows is a review of how the economics of Roman Britain have been characterized. It is argued that many of these characterizations have failed to problematize Romano-​British economics. As a consequence, the economics of Roman Britain have been generally understood as a set of proto-​capitalist systems. This view of economic activity has skewed our understanding of the foundations of late Roman power. By rethinking these issues and mobilizing new models, an alternative view can be presented that has the potential to recast our understanding of late Roman social, economic, and political dynamics. The study of the past can often be characterized in one of two ways. Ancient societies were either readily explicable and populated by rational beings, motivated by the same desires, fears, and emotions as us; or they were alien places inhabited by irrational people with alien desires and fears. Such visions of the past are probably deeply rooted philosophically and psychologically. Was the burial of a cow due to a farmer needing to dispose of the infectious carcass of some bovine plague, or part of a complex ritual involving ancestor worship and chthonic deities? Was the settlement built on top of a hill to place it beyond the reach of a flooding river, or so that the inhabitants could be perceived as being closer to the gods? Questions such as these, absurd though they sound, can be found answered one way or the other in a host of excavation reports.

Economy and Power in Late Roman Britain    851 The division of the past into periods that are perceived by modern viewers as being either rational or irrational has traditionally cleaved along a line determined by the presence or absence of textual sources. Irrationality is commonplace in the interpretation of temporally distant prehistoric societies, where exotic ethnographic analogies can be used to determine just how bizarre and alien the diverse societies of Homo sapiens sapiens are (and were) to modern western eyes (e.g. Parker Pearson and Ramilisonina 1998). However, the historic periods, with their textual sources and long-​standing positions in the origin myths of our modern nation states, always appeared more rational. The early Middle Ages (and particularly the fifth, sixth, and seventh centuries) were an exception to this rule, and this can arguably be linked to the near absence of contemporary textual sources (Moreland 2001, 2006). Roman Britain has often been seen as a kind of less sophisticated version of the present (e.g. Frere 1967; Salway 1981; de la Bédoyère 2006). The trappings of classical antiquity—​ towns, money, an army, taxation, government, and infrastructure—​are all familiar to modern eyes. Furthermore, our own world is founded on the intellectual principles of the Enlightenment and Graeco-​Roman revival. Like it or not, Rome remains one of the fundamental ideological touchstones of the western world (e.g. Dyson 2006). The study of Romano-​British economics has largely perpetuated this vision of a Roman past that can be intuitively understood. The historiography of this approach is complex and cannot be dealt with in detail here. However, in broad terms the economics of Roman Britain are usually presented as a development from a socially embedded Iron Age system. Such a system would be reliant on the social context of exchange, where personal and community obligations and reciprocity were paramount. The Roman conquest introduced towns, markets, and money, and, by the fourth century, this development had led to the emergence of socially disembedded economics and a fully monetized economy (Frere 1967; Fulford 1979) (see Walton and Moorhead, this volume, for a discussion on coinage). The problems with such an approach are legion, and scholars have tinkered with it in various ways. State intervention to explain peculiar or irrational distributions of material culture has been suggested (Allen and Fulford 1996: 269). The survival of Iron Age social systems has been argued to influence the distribution of late Roman pottery kilns (Millett 1990: 165–​169). Recently, Mattingly (2007) has invited us to accept a Roman Britain with many different kinds of economy. Nevertheless, when one studies the detail of material culture, the models for its distribution are often implicit. In the 1970s Hodder (1974) studied the distribution of various kinds of coarse pottery in south-​central England. He was explicitly looking for ‘marketing patterns’ in which pottery was manufactured, transported, and sold in local market centres. Applying the principles borrowed from geography, he mapped out ceramic distributions and explained the success or failure of one type of pottery over another by cost–​benefit analysis. Oxfordshire red-​slipped finewares were more successful than their New Forest rivals because they were functionally better and there was access to the cheap river-​ borne transport offered by the Thames (Fulford and Hodder 1975). Implicit within these interpretations is an economically rational world driven by concepts of profit, loss, and margins.

852

852   James Gerrard This uncritical acceptance of a vision of the past is not just confined to pottery studies. The large quantities of Roman period animal bones that have been excavated since the 1960s have been treated in a largely similar fashion (e.g. Maltby 2010). Animals were either raised for household consumption or driven to market and sold for consumption by townsfolk. There seem to be few alternatives, and what is worse is the apparent lack of any evaluation of why this type of model is thought to be appropriate (see also Maltby, this volume, for further discussion of the role of animals in the economy). Since 1990 the number and type of theoretical approaches applied to Roman Britain has blossomed. This rich diversity of approach, which is best encapsulated within the pages of the Theoretical Roman Archaeology Conference proceedings, show cases of just what a different, alien, and irrational place Roman Britain was. Yet this wider sea-​change in interpretation has rarely been equalled by a concomitant improvement in theoretical approaches applied to Romano-​British economics (Greene 1998). As an example of this, a recent study of pottery in the Midlands has followed the path blazed thirty years earlier by Hodder (Taylor 2004). There may be a simple, if somewhat depressing, explanation for this phenomenon. Hodder successfully navigated from working on Romano-​British coarse wares to being internationally renowned as an archaeological theoretician. There are few who can equal this career trajectory. Most people working on the nuts and bolts of the archaeological evidence for late Roman economic activity are specialists of finds and faunal remains. They are beavering away, dealing with things, not theories, and they are often working in environments, such as commercial archaeology, where engaging in theoretical musing is a luxury they can ill afford. One recent study on ancient economics has argued that the Roman world can be conceptualized economically as an agrarian society and tributary empire (Bang 2007). This model, which draws heavily on ethnohistoric parallels from the Moghul period, was developed explicitly for the early Roman period in the Mediterranean. However, it can also be utilized to great effect in the study of late Roman Britain. It emphasizes the coexistence of an agrarian society with urbanism and literary ‘high culture’, while addressing the importance of imperial surplus extraction from localized economic and social systems. Finally, the concept of the bazaar is introduced to describe an imperfect trading world ‘characterised by chronic imbalances and asymmetries in the supply of available information and goods’ (Bang 2007: 139). The application of this model to fourth-​century Britain would suggest that the foundation of the economy was not the archaeologically visible material culture like pottery, but agricultural commodities such as crops and animals. One way of understanding this economy is to consider the size of Britain’s population in the late Roman period. In the early 1990s a figure of approximately 3.67 million people (Millett 1990: 181–​186) was suggested. Such estimates are fraught with difficulties, but a number lying between the Domesday estimate of 1.76 million and the medieval maximum of the mid-​fourteenth century (4.81 million) (Broadberry et al. 2010) seems appropriate. Certainly, the number of Romano-​British rural sites discovered by commercially funded excavation since the early 1990s (Taylor 2007) would seem to indicate that Millett’s estimate is unlikely to be

Economy and Power in Late Roman Britain    853 too high. If we use the hypothetical figure of 3.67 million people, the economic impact of this population can be considered. Modern estimates for the bare minimum calorific requirements needed by a peasant cultivator suggest that each individual would require somewhere in the region of 250kg of wheat-​equivalent per annum (Clark and Haswell 1970: 57–​73). Roman-​period grain yields are uncertain and would, in any case, be variable according to climate, weather, and soil type. However, a ratio of one pound of seed to four pounds of harvested grain may be appropriate on slim Roman period evidence (Evans 1981; Mayerson 1984; Bang 2007: 89). This would add another 83kg of wheat, equivalent to the bare subsistence figure, as seed required for the next season’s planting. Thus, just to feed the population at subsistence level, over 1.2 million tonnes of grain had to be grown every year. This figure is admittedly uncertain, but it is also thought provoking. This is especially true when it is emphasized that this is a minimum food intake. The production of other commodities to provide further foodstuffs and clothing, as well as a surplus that could be used to meet the demands of taxation and exchange, ensures that this figure must be a mere percentage of total agricultural production. Archaeologically this element of the late Roman economy is both immediately obvious and frustratingly obscure. Field systems (Taylor 2007: 55–​7 1), rural farmsteads, palynological evidence (e.g. Dark 2000; Fyfe and Rippon 2004), and masses of animal bone (e.g. Maltby 2010) all exist in relative abundance. Yet the archaeobotanical evidence for the crops and cereals grown is still rare and depends on suitable preservational circumstances and recovery methods (van der Veen et al. 2007; see also van der Veen, this volume, for further discussion of crops). The use of the modern economic concept of gross domestic product (GDP) as a term is perhaps best avoided, owing to its modern and therefore anachronistic overtones. However, against the mountain of crops, cereals, meat, cheese, wool, and leather produced in fourth-​century Britain, the production, distribution, and exchange of material culture—​whether that be pots, building stone, mosaics, or items of jewellery—​must represent a minor element in late Romano-​British economic output. The recognition of this fact allows the model of an agrarianate society and a tributary empire to be deployed to its full effect. During the 1970s a nucleated rural settlement was excavated just north of Somerton in Somerset. This site, called Castgore (Leech 1982), has become something of a type site for what is sometimes referred to anachronistically as the Romano-​British ‘village’. It was formed of a series of rectangular buildings, often divided into three rooms, set within ditched and later walled plots alongside a minor road. The finds included all of the paraphernalia typical of late Roman farming community: coins, knives, agricultural implements, querns, assemblages of faunal and archaeobotanical remains. The pottery points to links with south-​east Dorset, the New Forest, Oxfordshire, and other regions. Over recent years similar sites have been dug elsewhere in the lowlands of Britain; at Kingscote in Gloucestershire (Timby 1998) and Higham Ferrers in Northamptonshire (Lawrence and Smith 2009). It is wrong to describe these settlements as typical, but they do provide a snapshot of fourth-​century life for rural communities (Figure 41.1).

854

854   James Gerrard

Gra

vel r

oad

surf ace

N

0

100 m

Figure 41.1  Plan of the roadside settlement at Higham Ferrers. Source: after Lawrence and Smith (2009: figure 2.18). © James Gerrard/​Andrew Agate.

Such settlements are usually characterized as being based on an agricultural economy, and this is surely true. However, this type of site and others like them are often unconsciously placed within an interpretative framework that argues that the inhabitants were producing a surplus to be marketed through a nearby urban centre. The publication of the finds often highlights the distance that goods, especially pottery, travelled before being consumed, for instance, in Somerset or Gloucester (Leech 1982: 159). This is certainly one way of interpreting the evidence. However, an alternative can be suggested. In a well-​known study of Hawaiian villages, Kolb and Snead (1997) used the concept of ‘small worlds’ to characterize their social and economic structure. This model can be applied to the fourth-​century rural situation. Did the inhabitants of Catsgore (Leech 1982), Kingscote (Timby 1998), or Cefn Cwmwd and Melin y Plas in Anglesey (Cuttler

Economy and Power in Late Roman Britain    855 et al. 2012) consider themselves to be part of a globalized network that ultimately connected them to Rome, Ravenna, Constantinople, and beyond? Or did they live in a ‘small world’ characterized by the rhythms of the agricultural season, occasional trips to the local ‘small town’, and episodic visits by peddlers and gatherings at fayres? Was the red-​slipped and black burnished pottery on their tables a constant feature on the market stalls in the local forum? Or did it appear only occasionally—​perhaps seasonally—​as a sudden glut, which was inevitably followed by a long period when no vessels were available (Bang 2007: 139)? Here a lone comment contained within an early Roman period letter found at Vindolanda is a powerful reminder that the road network was probably impassable for much of the autumn and winter: The hides which you write are at Cataractonium [Catterick]—​write that they be given to me and the wagon about which you write. And write to me what is with that wagon. I would have already been to collect them except that I did not care to injure the animals while the roads are bad. (Tab. Vindol. II. 343)

A chance observation like this indicates just how small and unpredictable a Roman rural community’s world could become. One element that has not yet been explored in this discussion is the relationship between the agrarianate society and the tributary empire. For these rural dwellers with their limited horizons these elements of social organization must have been articulated through the local elites. This social class and their control over the economy bring this discussion to the crux of the relationship between economics and power in the late Roman world. One of the defining aspects of the Roman Empire was the state’s ability to demand and raise revenues through taxation. Late Roman legal texts, such as the Theodosian Code (Pharr 1952), contain many laws relating to taxation, yet the intricacies of the tax system remain unclear (e.g. Jones 1964: 462–​469, 820–​823). Even such seemingly straightforward issues such as the scale of taxation are contentious topics. There is also plenty of room for temporal and geographical variation in how taxes were assessed and levied. These unquantifiable factors make any detailed discussion of the late Roman taxation system in Britain exceedingly difficult. However, this does not need to be concerning. The key issue regarding taxation is that it was the economic link between the small world of agricultural communities, like those at Catsgore and Kingscote, and the globalized empire. It is generally accepted that the collection of taxation was devolved upon regional administrative units known as the civitates. Within each individual civitas the town council or ordo was tasked with fulfilling the demands of the state and personally responsible for making good any shortfall in revenue (Jones 1964: 457; Wickham 1984). Much has been written about the negative impact on the curial class of this system. However, it also offered provincial elites the opportunity for self-​aggrandizement. The social structure of the late Roman world was highly stratified. Society was broadly divided into the honestiores and the humiliores (the honest men and the humble men)

856

856   James Gerrard (Garnsey 1970). The honestiores were a privileged class. If guilty of crimes, they were subject to fines, whereas the humiliores were likely to be subject to corporal punishment and torture (Harries 2007: 36). This general distinction marked the division between the elites and the lowly masses. Yet even within the elite groups further gradations of status and stratification were evident and formally marked by titles of status (Näf 1995: 20–​21; Mathisen 2001). This social stratification and the inequalities that it embodied produced a situation in which the humiliores came to depend on their powerful patrons (Brown 1971: 37). It was the local members of the elite who had the connections and money to grease the wheels of the legal system, or to negotiate with a powerful neighbouring landowner over property rights and long-​standing custom (Harries 1999: 79; Kelly 2004). This dependent relationship between the client and the patron will have been marked in other ways. Many clients will have been tenants, debtors, or otherwise obligated to their patron. Such power relationships and their inequalities offered significant scope for exploitation by elites, and the tax system was particularly vulnerable to such interventions. It is sometimes convenient to assume that late Roman taxation was a perfect system of annual assessments and payments. This view of a cumbersome and no doubt bureaucratic system is almost certainly erroneous (Brown 1971: 36). Taxation was likely to have been infrequent. Annual assessments may have stacked up and been demanded for payment in a single season, and bad harvests were likely to have been invisible to those charged with finding sufficient revenue to fill the state’s coffers. Periodic remissions and extra taxes, coupled with imperfect record-​keeping, probably further complicated the situation. For the agricultural producers coping with such a system, life would have been far from easy (Salvian De Gub. Dei IV.3). Here the local patron offered a means to intercede between those at the bottom of the social ladder and the inexorable demands of late Roman government. In so doing, the lines between what was due from the client or tenant to the patron or landlord and what was due to the state will have become blurred. Rent and tax may have been rolled together as one burden due to the local lord (Wickham 1984). That patron, in turn, could have used his position and influence to gain exemptions from paying tax, or could levy increased charges on his clients and tenants by claiming that state had demanded more. In short, the position of the local elite, sandwiched between those producing an agricultural surplus and the demands of the state, was an invitation for them to line their own pockets. Archaeologically these elite groups are visible through the architecturally elaborate, richly decorated and opulently furnished villas and townhouses that characterize much of lowland Britain in the fourth century (e.g. Smith 1997; Scott 2000). These residences, all too often seen as the dwelling places of supremely successful farmers or economically successful businessmen, were in fact the residences of the late antique aristocracy, and their wealth was staggering. The Hoxne hoard (Guest 2005; Johns 2010), with its plate and jewellery bearing the names of Aurelius Ursicinus and the Domina Juliane, is one manifestation of this wealth. It is equivalent to 5234.8 g of gold (Hobbs 2006: table 73), and this is a sum that would have sufficed to pay the taxes of Timgad in the fourth century (Kelly 2004: 145–​150). Yet it must represent only a fraction of the owner’s wealth.

Economy and Power in Late Roman Britain    857 The silver plate that might be expected is largely absent, and an antiquarian find of several hundred gold solidi nearby may indicate that the owners did not bury all their golden eggs in one basket (Guest 2005: 16–​33). On an imperial scale the owner(s) of the Hoxne hoard or a villa like Turkdean in Gloucestershire (Holbrook 2004) were small fry (Figure 41.2). A middling fourth-​century senator would have an annual income many times the value of the Hoxne hoard. Nevertheless, in lowland Britain this provincial elite exercised local power from a position far exalted above that of the average tenant cultivator. Yet it would be a mistake to characterize the fourth-​century Romano-​British elites in the singular. Social stratifications will have divided the honestiores—​both horizontally and vertically—​producing competing interest groups and nested patron–​client relationships (Wallace-​Hadrill 1989; Brown 1992). The analysis of late Roman settlement hierarchies has had a tendency to distinguish between villa sites and other forms of rural settlement. Implicit within this is the suggestion that such a division marks that between high-​and low-​status communities. This may well be true, although in some parts of Britain, like the Midlands, other form of architectural display, like aisled buildings, were prevalent (Taylor 2001: 52). However, the cartographic interpretation of settlement distributions flattens elite social stratification and provides a misleading view of power relationships in the fourth-​century landscape. The hinterland of the small town of Lindinis (Ilchester) in southern Somerset is one of the most densely occupied late Roman landscapes in lowland Britain. Sites such as Castgore (Leech 1982), Bradley Hill (Leech 1981), and Yeovilton (Lovell 2006) seem to be those of the rural population linked by roads and trackways that ran between complex

N

0

100 m

Figure 41.2  Plan of the Roman villa at Turkdean. Source: after Holbrook (2004: figure 4). © James Gerrard/​Andrew Agate.

858

858   James Gerrard field systems identified by aerial and geophysical survey. Within this densely populated landscape were a large number of villas (Figure 41.3). South of the town were villas at Seavington St Mary (Graham and Mills 1995), Dinnington (Adcock and Wood 2005; Prof. A. King, pers. comm.), and Lopen (Brunning 2001) in the south-​west; Lufton (Hayward 1952, 1972), Westlands (Radford 1928), and Bedmore Barn (Ham Hill) (Beattie and Pythian-​Adams 1913; Sharples et al. 2012: 39) in the centre, and East and West Coker (Moore 1862; Haverfield 1906: 330). All of these sites have yielded stone building, mosaics, wall plaster, and artefactual evidence of late Roman occupation. This density of settlement in a small area is striking, and, given that both Dinnington and Lopen are recent discoveries, it might be expected that further villa sites remain to be discovered within this region.

River

To

Lin

col

n

N

Yeo

Ilchester 4

ter

xe

E To

To

6

est

3

rch Do

1

er

2 5

9

0

4 km

7

8

Figure 41.3  The distribution of villas near Ilchester: 1) Dinnington, 2) Lopen, 3) Seavington St Mary, 4) Ilchester Mead, 5) Batemoor Barn, 6) Lufton, 7) West Coker, 8) East Coker, 9) Westlands. Source: © James Gerrard/​Andrew Agate.

Economy and Power in Late Roman Britain    859 It has generally been suggested that these sites were the pinnacle of the late Roman settlement hierarchy during the fourth century. This is surely true. Buildings like that excavated at Dinnington represent an architectural form that embodied high-​status civilian elite display and the aristocratic values of paideia that were shared across the Roman Empire (Ellis 1991; Brown 1992; Bowes 2010). A hypothetical fourth-​century Sicilian, familiar with the Piazza Armerina (Gentili 1999), might have sneered at the Romano-​ British interpretation of those values, but he or she would also have understood that the locals were attempting to share the same values and traditions that the visitor held dear (Bowes 2010). Unfortunately, the archaeological interpretation of these sites has reduced them to being mechanistic points in a cartographic landscape. Each site is often viewed as the dwelling of a single family, and interaction between sites reduced to hypothetical estates derived from nearest neighbour analysis. The reality of living in this landscape and these villas must have been very different from the flat abstractions of archaeological analysis. The close proximity of many of these sites is striking. The villas at East and West Coker are a few minutes’ walk apart, and many of the other sites are within easy reach of one another, either by foot or by horse. It is inconceivable that the inhabitants of these sites lived in isolation. A complex social network is likely to have linked these communities together. Many of these villas were occupied from at least the middle of the third century onwards. Thus, by the middle of the fourth century, perhaps five generations had been raised at some of these sites. Their physical proximity to one another and the duties of the curial class—​serving on the local town council or ordo, for instance—​will have thrown their inhabitants together. Friendship, marriage, business and political interests, feuds, and disagreements will all have characterized this social network (Brown 1992; Sivonen 2006). The Gallic world of Ausonius and Sidonius is brought to life by their surviving letters, which give glimpses of such a society (Sivan 1993). There are no surviving analogues from fourth-​century Britain, although Patrick’s early fifth-​century letters suggest that they may once have existed (Dumville 1993). The Gallic sources and settlement patterns, such as that identified south of Ilchester, suggest that these landscapes were characterized not by isolated estates, but by high-​status communities joined by ties of kith, kin, and patronage. Those at the pinnacle of this network are likely to have controlled directly and indirectly large swathes of agricultural land. There is little direct evidence of the size of late antique estates in the north-​western provinces. Ausonius’ ‘little inheritance’, one of four estates the orator and imperial tutor owned, is thought to have equated to some 238 hectares (Sivan 1993: 66–​69). How typical this is remains debatable. Estate sizes in the German provinces have been suggested as approximating to 200 hectares, but this is based on hypothesis rather than empirical data (King 1990). Nevertheless, a number of estates owned directly by an individual and controlled perhaps indirectly through their family friends and clients could easily total many thousands of hectares and be comparable in size to some of the smaller early medieval kingdoms known from documents like the Tribal Hidage (Dumville 1989; Yorke 1990: 9–​10). Of course, the Roman estates are unlikely to have been neatly distributed or geographically contiguous, but

860

860   James Gerrard the control of this land and its inhabitants and produce was the foundation of the late Roman elite’s power. The relationship between the state and the Romano-​British elites was probably characterized by symbiosis and paranoia. The state required the elite to serve as its agents and to enforce its demands, but to eschew the overt trappings of power that were an imperial monopoly. Thus the villa-​owner may have liked to be portrayed to his clients and tenants as an ‘emperor’ or ‘hero’ (Ellis 1991), but he had to stay firmly in the civilian world, wedded to the values of paideia (Sivonen 2006:  100–​104, 115–​117). Those who crossed this line, or were perceived to have done so, were branded by the state as latrones (bandits) (Grünewald 2004) and bagaudae (perhaps best thought of as insurgents) (Drinkwater 1992; Sánchez León 1996). In part, the elite were compensated for their lack of martial power by a legal system that was heavily biased towards those of elite social standing and wealth. There were few tenants or free peasants who could hope to triumph in the late Roman courts against the local landowner, his purse, and connections (Harries 1999; Kelly 2004). Ultimately the state provided a backstop to this system. Rebellion by the peasantry would also be characterized as banditry or bagaudic activity and met, if necessary, with the force of imperial arms.

The End of Roman Britain The interconnection between economics and power was thrown into sharp relief during the early fifth century. The usurpation of Constantine III in ad 407 (Drinkwater 1998) and, more significantly, the failure of the western court to reimpose imperial control over Britain during the fifth century, mark a significant historical break that is also reflected archaeologically (Gerrard 2013). The early fifth century saw the disappearance of the defining characteristics of Romano-​British society and the appearance of new forms of material culture, architecture and—​particularly in eastern areas—​burial rites. This transformation and the speed with which these changes occurred are complex matters and ones that deserve more space than this chapter can devote to them (see also Esmonde Cleary, this volume). The traditional view has been to see this change as the product of a catastrophic economic collapse. The disappearance of the state and the money-​using economy led to the dislocation of market-​based exchange and the virtual disappearance of trade. Britain was plunged from a familiar economic system of markets into the alien, early medieval world of gift-​giving and socially embedded exchange (Wickham 2005). This chapter has argued that our fixation with archaeologically visible economic indicators, such as pottery or roof tiles, has skewed interpretations of the late Roman economy. The economic powerhouse of fourth-​century Britain was not the kilns manufacturing red-​slipped tablewares in Oxfordshire, but grain, crops, meat, and animal products such as cheese and leather. A significant (but unknown) proportion of the agricultural surplus had been rendered to the state as tax. In the early fifth century the

Economy and Power in Late Roman Britain    861 state disappeared. This presented the Romano-​British elites with both an opportunity and a dilemma. For the elite the late Roman tax burden offered an opportunity. After subsistence requirements for their tenants and clients were fulfilled, the slice of surplus that went to the landlord as rent and the slice that paid the tax—​which were perhaps already blurred—​became indivisible. Both the rent and the tax went to the landlord (Wickham 1984). However, this opportunity was not without its difficulties. A tax that must have been raised in kind was perishable. Without a means of transferring that surplus into specie, the grain would rot, the cheese would go off, and the meat would spoil. The clients and tenants, recognizing the weakening state and the traditional structures of elite power, may also have tried to renegotiate their position—​a development possibly spurred by the appearance of new and alternative ‘Germanic’ power structures in the east of Britain (Böhme 1986, 1997). Faced with these difficulties, the late Roman elites undoubtedly responded in different ways. For some the fifth century probably offered oblivion, but for those who could adapt to the changed economic situation and power structures, a new and post-​Roman path could be discerned. Some of the surplus from their large estates could be preserved and used to cement the loyalties of clients and retainers. These individuals could be converted into nascent war bands now that the prohibition on the elite adopting the emperor’s martial values had ended. The burdens on the agricultural producers could also be lessened without impacting on the elite’s position (Esmonde Cleary 1989: 145). A notional fourth-​century tax rate of 10 per cent could be lowered by, say, 8 per cent. This would reduce the burden on recalcitrant agricultural producers but still leave the elite taking a larger slice of the cake than they had been under the Roman state. The argument advanced here is an interesting hypothesis. However, can it be supported by an interpretation of the historical and archaeological evidence? The historical sources are limited and equivocal. Zosimus (New History 6.5.3), drawing on the contemporary and largely lost Olympiodorus (Blockley 1981: 28), tells us that the Britons took up arms and expelled the Roman state, and this may be a reflection of a society choosing to go its own way. The Life of St Germanus and Patrick’s writings (Dumville 1989) portray an elite in the late Roman mould. Finally, Gildas—​in the sixth century—​recounts how the superbus tyrannus supplied his Saxon mercenaries with foodstuffs, which suggests that some individuals and communities could accumulate and distribute surpluses. The archaeological evidence is just as problematic and capable of being interpreted in contradictory ways. The final phases of many fourth-​century villas display a shift to ‘messy’ deposits, often associated with grain-​driers and metal-​working hearths. This so-​ called squatter occupation is often taken as evidence of economic collapse. However, it could equally represent a realignment of priorities (Lewit 2003: 268). The old engines of elite display—​villas—​were now serving a less hierarchical form of social organization. Buildings that had previously been used to emphasize the separation of the elite from their clients were now being used for economically productive activities (Ellis 1988; Petts 1997; Ripoll and Arce 2000). Here it is interesting to note that grain-​driers may have been used in brewing (Van der Veen 1989). Beer would be one way of preserving a

862

862   James Gerrard grain surplus and turning it into a beverage that was suitable for use in the feasting that sought to bind the lord and his clients with ties of loyalty. The agricultural landscape also underwent significant changes during the fifth century. Palynological evidence presents a complex picture, but there is little clear support for reforestation or a significant national decline in the area of land being worked (Dark 1996, 2000; Dark and Dark 1997: 143–​144; Fyfe and Rippon 2004). Late Roman field systems were sometimes abandoned or partially maintained (Rippon 1991, 2006; Oosthuizen 1998, 2011; Baker 2002; Upex 2002). Again, this would seem at first sight to be evidence of a collapse in the rural economy. Alternatively, we might speculate that a shift occurred away from a late Roman rural arable economy driven by the demands of taxation. In the fifth century the reduced demands from the elite may have encouraged a less intensive agricultural regime that emphasized cattle and animals. We lack suitable animal bone assemblages to test this hypothesis (Crabtree 2010: 123), but the emergence of cattle as a measure of wealth would seem to be a strong possibility (Hamerow 2002: 129). Finally, recent analysis of osteological evidence has drawn attention to the fact that late Roman populations were shorter than early medieval individuals. This is a phenomenon that is visible across Europe (Köpke and Baten 2005) and also discernible in local and regional studies (Roberts and Cox 2003: 389–​390; Klingle 2011). Traditionally this height difference has been viewed as indicative of migration, with incoming Germanic groups being taller than the indigenes. Interestingly, recent work has suggested that height might be a proxy for health (Köpke and Baten 2005). The arguments are not straightforward, but perhaps fifth-​and sixth-​century populations ate better than their Roman forebears.

Conclusions Power and economics are inextricably intertwined. The study of the latter has been largely and unfortunately neglected, leaving much of the data to rest on implicit and poorly founded assumptions and models. A more critical interpretation of the fourth-​ century Romano-​British economy, using models developed for other regions of the Roman Empire, perhaps suggests that archaeologists, by focusing on what survives, have neglected the less visible agricultural economy that served as the foundation for late Roman society. A consideration of late Roman social stratification and the state suggests that the ability to control and exploit agricultural surpluses was critical to the maintenance and replication of power. The fifth-​century crisis, brought about by the failure of the western Empire to bring Britain back into the imperial fold after the defeat of Constantine III’s usurpation, left the insular power structures in a state of flux. Those capable of exploiting and maintaining their hold over the agricultural economy used their surplus to buy the loyalty of retainers and clients. By adopting the trappings of military power, these elites could reinvent themselves and their children as post-​Roman warlords and—​ultimately—​petty kings.

Economy and Power in Late Roman Britain    863

Acknowledgements I am grateful to Mr Andrew Agate for preparing the illustrations.

References Primary Sources Constantius, Life of St Germanus. ‘Constantius of Lyon: The Life of St Germanus of Auxerre’, ed. and trans. T. Noble and T. Head, in Soldiers of Christ: Saints Lives from Late Antiquity and the Early Middle Ages. Philadelphia: Pennsylvania State University Press, 1994, 75–​106. Patrick, Letter to Coroticus: S. Patrick: His Writings and Muirchu’s Life, ed. and trans. A. B. E. Hood. Arthurian Period Sources 9. Chichester: Phillimore & Co., 1978. Salvian, De Gubernatione Dei. The Writings of Salvian the Presbyter, trans. J. O’Sullivan. Washington: Catholic University of America Press, 1947. Tab. Vindol. = The Vindolanda Tablets Online (accessed 4 December ​2014). Zosimus. Zosimus: A New History, trans. and commentary by Ronald T. Ridley. Sydney: University of Sydney, 1982.

Secondary Sources Adcock, J., and Wood, E. (2005). ‘Geophysical Survey Report: Dinnington, Somerset’. Unpublished GSB Prospection Report 2005/​33. Allen, M., and Fulford, M. (1996). ‘The Distribution of South-​East Dorset Black Burnished Category 1 Pottery in South-​West Britain’, Britannia, 27: 223–​281. Baker, S. (2002). ‘Prehistoric and Romano-​British Landscapes at Little Wittenham and Long Wittenham, Oxfordshire’, Oxoniensia, 67: 1–​28. Bang, P. (2007). The Roman Bazaar: A Comparative Study of Trade and Markets in a Tributary Empire. Cambridge: Cambridge University Press. Beattie, I., and Pythian-​Adams, W. (1913). ‘A Romano-​British House near Bedmore Barn, Ham Hill, Somerset’, Journal of Roman Studies, 3/​1: 127–​133. Blockley, R. (1981). The Fragmentary Classicizing Historians of the Later Roman Empire: Eunapius, Olympiodorus, Priscus and Malchus. Liverpool: Francis Cairns. Böhme, H. (1986). ‘Das Ende der Römerherrschaft in Britannien und die Angelsächsich Besiedlung Englands im 5. Jahrhundert’, Jahrbuch des Römisches-​Germaniches Zentralmuseum, 33: 469–​574. Böhme, H. (1997). ‘Söldner und Siedler im spätantiken Nordgallien’, Die Franken, 1: 91–​101. Bowes, K. (2010). Houses and Society in the Later Roman Empire. London: Duckworth. Broadberry, S., Campbell, B., and van Leuwen, B. (2010). ‘English Medieval Population: Reconciling Time Series and Cross-​Sectional Evidence’. University of Warwick: Unpublished working paper. Brown, P. (1971). The World of Late Antiquity. London: Thames and Hudson. Brown, P. (1992). Power and Persuasion in Late Antiquity:  Towards a Christian Empire. Madison: University of Wisconsin Press.

864

864   James Gerrard Brunning, R. (2001). ‘Lopen Bridge Farm’, Somerset Archaeology and Natural History, 145: 142–​143. Clark, C., and Haswell, M. (1970). The Economics of Subsistence Agriculture. London: MacMillan. Crabtree, P. (2010). ‘Agricultural Innovation and Socio-​Economic Change in Early Medieval Europe: Evidence from Britain and France’, World Archaeology, 42/​1: 122–​136. Cuttler, R., Davidson, A., and Hughes, G. (2012). A Corridor through Time: The Archaeology of the A55 Anglesey Road Scheme. Oxford: Oxbow Books. Dark, K., and Dark, P. (1997). The Landscape of Roman Britain. Stroud: Tempus. Dark, P. (1996). ‘Palaeoecological Evidence for Landscape for Landscape Continuity and Change in Britain ad 400–​800’, in K. Dark (ed.), External Contacts and the Economy of Late and Post-​Roman Britain. Woodbridge: Boydell Press, 23–​51. Dark, P. (2000). The Environment of Britain in the First Millennium ad. London: Duckworth. de la Bédoyère, G. (2006). Roman Britain: A New History. London: Thames and Hudson. Drinkwater, J. (1992). ‘The bacaudae of Fifth-​Century Gaul’, in J. Drinkwater and H. Elton (eds), Fifth-​Century Gaul: A Crisis of Identity? Cambridge: Cambridge University Press, 208–​217. Drinkwater, J. (1998). ‘The Usurpers Constantine III (407–​411) and Jovinus 411–​413’, Britannia, 29: 269–​298. Dumville D. (1989). ‘The Tribal Hidage: An Introduction to its Texts and History’, in S. Bassett (ed.), The Origins of Anglo-​Saxon Kingdoms. Leicester: Leicester University Press, 225–​230. Dumville, D. (1993). St Patrick, ad 493–​1993. Woodbridge: Boydell Press. Dyson, S. (2006). In Pursuit of Ancient Pasts:  A  History of Classical Archaeology in the Nineteenth and Twentieth Centuries. New Haven: Yale University Press. Ellis, S. (1988). ‘The End of the Roman House’, American Journal of Archaeology, 92/​ 4: 565–​576. Ellis, S. (1991). ‘Power, Architecture and Decor: How the Late Roman Aristocrat Appeared to his Guests’, in E. Gazda (ed.), Roman Art in the Private Sphere. Ann Arbor, University of Michigan Press, 117–​134. Esmonde Cleary, S. (1989). The Ending of Roman Britain. London: Batsford. Evans, J. (1981). ‘Wheat Production and its Social Consequences in the Roman World’, Classical Quarterly, 31/​2: 428–​442. Frere, S. (1967). Britannia: A History of Roman Britain. London: Routledge and Kegan Paul. Fulford, M. (1979). ‘Pottery Production and Trade at the End of Roman Britain:  The Case against Continuity’, in P. Casey (ed.), The End of Roman Britain. BAR British Series 71. Oxford: Archaeopress, 120–​132. Fulford, M., and Hodder, I. (1975). ‘A Regression Analysis of Some Late Romano-​British Pottery: A Case Study’, Oxoniensia, 39: 26–​33. Fyfe, F., and Rippon, S. (2004). ‘A Landscape in Transition? Palaeoenvironmental Evidence for the “End” of the Romano-​British Period in South-​West England’, in R. Collins and J. Gerrard (eds) Debating Late Antiquity in Britain ad 300–​700. BAR British Series 365. Oxford: Archaeopress, 33–​42. Garnsey, P. (1970). Social Status and Legal Privilege in the Roman Empire. Oxford, Oxford University Press. Gentili, G. (1999). La Villa Romana di Piazza Armerina Palazzo Erculio. Osimo: Fondazione Don Carlo. Gerrard, J. (2013). The Ruin of Roman Britain. Cambridge: Cambridge University Press.

Economy and Power in Late Roman Britain    865 Graham, A., and Mills, J. (1995). ‘A Romano-​British Building at Crimbleford Knap, Seavington St Mary’, Somerset Archaeology and Natural History, 139: 119–​134. Greene, K. (1998). ‘Review Articles:  The Study of Roman Pottery in Britain in the Late 1990s: Progress and Prospects’, Britannia, 29: 447–​450. Grünewald, T. (2004). Bandits in the Roman Empire: Myth and Reality, trans. J. Drinkwater. London: Routledge. Guest, P. (2005). The Late Roman Gold and Silver Coins from the Hoxne Hoard. London: British Museum Press. Hamerow, H. (2002). Early Medieval Settlements: The Archaeology of Rural Communities in Northwest Europe 400–​900. Oxford: Oxford University Press. Harries, J. (1999). Law and Empire in Late Antiquity. Cambridge: Cambridge University Press. Harries, J. (2007). Law and Crime in the Roman World. Cambridge:  Cambridge University Press. Haverfield, F. (1906). ‘Romano-​British Somerset’, Victoria County History of Somerset, I. 207–​371. Hayward, L. (1952). ‘The Roman Villa at Lufton, near Yeovil’, Proceedings of the Somerset Archaeological and Natural History Society, 97: 91–​112. Hayward, L. (1972). ‘The Roman Villa at Lufton, near Yeovil’, Proceedings of the Somerset Archaeological and Natural History Society, 116: 59–​77. Heather, P. (2005). The Fall of the Roman Empire: A New History. London: Macmillan. Hobbs, R. (2006). Late Roman Precious Metal Deposits c. ad 200–​700: Changes over Time and Space. BAR International Series 1540. Oxford: Archaeopress. Hodder, I. (1974). ‘Some Marketing Models for Romano-​British Coarse Pottery’, Britannia, 5: 340–​359. Holbrook, N. (2004). ‘Turkdean Roman Villa, Gloucestershire: Archaeological Investigations 1997–​1998’, Britannia, 35: 39–​76. Johns, C. (2010). The Hoxne Late Roman Treasure:  Gold Jewellery and Silver Plate. London: British Museum Press. Jones, A. (1964). The Later Roman Empire 284–​602:  A  Social, Economic and Administrative Survey. Baltimore: Johns Hopkins University Press. Kelly, C. (2004). Ruling the Later Roman Empire. Harvard: Harvard University Press. King, A. (1990). Roman Gaul and Germany. Los Angeles: University of California Press. Klingle, D. (2011). ‘The Use of Skeletal and Mortuary Evidence to Understand the Transition from Roman to Anglo-​Saxon England in the Cambridgeshire Region’. Unpublished Ph.D. thesis, Department of Archaeology, Cambridge University. Kolb, M., and Snead, J. (1997). ‘It’s a Small World After All:  Comparative Analyses of Community Organization in Archaeology’, American Antiquity, 62/​4: 609–​628. Köpke, N., and Baten, J. (2005). ‘The Biological Standard of Living in Europe during the Last Two Millennia’, European Review of Economic History, 9/​1: 61–​95. Lawrence, S., and Smith, A. (2009). Between Villa and Town: Excavations of a Roman Roadside Settlement and Shrine at Higham Ferrers, Northamptonshire. Oxford: Oxbow Books. Leech, R. (1981). ‘The Excavation of a Romano-​British Farmstead and Cemetery on Bradley Hill, Somerset’, Britannia, 12: 177–​252. Leech, R. (1982). Excavations at Catsgore 1970–​73: A Romano-​British Village. Bristol: Western Archaeological Trust. Lewit, T. (2003). ‘Vanishing Villas: What Happened to Elite Rural Habitation in the West in the 5th–​6th Centuries?’, Journal of Roman Archaeology, 16: 260–​274.

866

866   James Gerrard Lovell, J. (2006). ‘Excavation of a Romano-​British Farmstead at RNAS Yeovilton’, Proceedings of the Somerset Archaeological and Natural History Society, 149: 7–​70. Maltby, M. (2010). Feeding a Roman Town:  Environmental Evidence from Excavations in Winchester 1972-​1985. Winchester: Winchester Museums. Mathisen, R. (2001). ‘Imperial Honorifics and Senatorial Status in Late Roman Legal Documents’, in R. Mathisen (ed.), Law, Society and Authority in Late Antiquity. Oxford: Oxford University Press, 179–​207. Mattingly, D. (2007). An Imperial Possession: Britain in the Roman Empire. London: Penguin. Mayerson, P. (1984). ‘Roman Wheat Production:  An Addendum’, Classical Quarterly, 34/​ 1: 243–​245. Millett, M. (1990). The Romanization of Britain:  An Essay in Archaeological Interpretation. Cambridge: Cambridge University Press. Moore, J. (1862). ‘Roman Villa at West Coker, Somersetshire’, Journal of the British Archaeological Association, 18: 392–​395. Moreland, J. (2001). Archaeology and Text. London: Duckworth. Moreland, J. (2006). ‘Archaeology and Texts: Subservience or Enlightenment?’, Annual Review of Archaeology, 35: 135–​151. Näf, B. (1995). Senatorisches Standesbewusststein in spätrömischer Zeit. Beiträge zur Geschichte der altchristlichenen Literatur und Theologie 40. Freiburg, Paradosis. Oosthuizen, S. (1998). ‘Prehistoric Fields into Medieval Furlongs? Evidence from Caxton, South Cambridgeshire’, Proceedings of the Cambridge Antiquarian Society, 86: 145–​152. Oosthuizen, S. (2011). ‘Anglo-​Saxon Fields’, in D. Hinton, S. Hamerow, and S. Chapman (eds), Handbook of Anglo-​Saxon Archaeology. Oxford: Oxford University Press, 377–​401. Parker Pearson, M., and Ramilisonina, R. (1998). ‘Stonehenge for the Ancestors: The Stones Pass on the Message’, Antiquity, 72/​276: 308–​326. Petts, D. (1997). ‘Elite Settlements in the Roman and Sub-​Roman period’, in K. Meadows, C. Lemke, and J. Heron (eds), TRAC 96: Proceedings of the Sixth Theoretical Roman Archaeology Conference, Sheffield 1996. Oxford: Oxbow Books, 101–​112. Pharr, C. (1952) (trans.). The Theodosian Code and Novels and the Sirmondian Constitutions:  A  Translation with Commentary, Glossary and Bibliography. Princeton: Princeton University Press. Radford, C. (1928). ‘The Roman Site at Westland’, Proceedings of the Somerset Archaeological and Natural History Society, 74: 122–​143. Ripoll, G., and Arce, J. (2000). ‘The Transformation and End of Roman villae in the West (Fourth to Sixth Centuries): Problems and Perspectives’, in G. Brogilo, N. Gauther, and N. Christie (eds), Towns and their Territories between Late Antiquity and the Early Middle Ages. Transformation of the Roman World 9. Leiden: Brill, 63–​114. Rippon, S. (1991). ‘Early Planned Landscapes in South-​East Essex’, Essex Archaeology and History, 22: 46–​60. Rippon, S. (2006). Landscape, Community and Colonisation: The North Somerset Levels during the 1st to 2nd Millennia ad. London: Oxford University Press and the British Academy. Roberts, C., and Cox, M. (2003). Health and Disease in Britain: From Prehistory until the Present Day. Stroud: Sutton. Salway, P. (1981). Roman Britain. Oxford: Clarendon Press. Sánchez León, J. (1996). Les Sources de L’Histoire des Bagaudes: Traduction et Commentaire. Besancon: Annales Littéraires de l’Université de Franche-​Comté.

Economy and Power in Late Roman Britain    867 Scott, S. (2000). Art and Society in Fourth-​Century Britain: Villa Mosaics in Context. Oxford University School of Archaeology Monograph 53. Oxford:  Oxford University School of Archaeology. Sharples, N., Evans, C., Slater, A., Payne, A., Linford, P., and Linford, N. (2012). ‘Ham Hill’, British Archaeology, 123: 34–​39. Sivan, H. (1993). Ausonius of Bordeaux: Genesis of a Gallic Aristocracy. London: Routledge. Sivonen, P. (2006). Being a Roman Magistrate:  Office Holding and Roman Identity in Late Antique Gaul. Bibliotheca Historica 101. Helsinki: Bibliotheca Historica. Smith, J. (1997). Roman Villas: A Study in Social Structure. London: Routledge. Sorkin, A. (2009). Too Big to Fail. London: Viking. Taylor, J. (2001). ‘Rural Society in Roman Britain’, in S. James and M. Millett (eds), Britons and Romans: Advancing an Archaeological Agenda. CBA Research Report 125. London: Council for British Archaeology, 46–​59. Taylor, J. (2004). ‘The Distribution and Exchange of Pink, Grog-​Tempered Pottery in the East Midlands’, Journal of Roman Pottery Studies, 11: 60–​66. Taylor, J. (2007). An Atlas of Roman Rural Settlement in England. CBA Research Report 151. London: Council for British Archaeology. Timby, J. (1998). Excavations at Kingscote and Wycomb, Gloucestershire: A Roman Estate Centre and Small Town in the Cotswolds. Cirencester: Cotswold Archaeological Trust. Upex, S. (2002). ‘Landscape Continuity and Fossilization of Roman Fields’, Archaeological Journal, 159: 77–​108. Van der Veen, M. (1989). ‘Charred Grain Assemblages from Roman-​Period Corn Driers in Britain’, Archaeological Journal, 146: 302–​319. Van der Veen, M., Lavarda, A., and Hill, A. (2007). ‘The Archaeobotany of Roman Britain: Current State and Identification of Research Priorities’, Britannia, 38: 181–​210. Wallace-​Hadrill, A. (1989) (ed.). Patronage in Ancient Society. London: Routledge. Ward-​Perkins, B. (2005). The Fall of Rome and the End of Civilization. Oxford:  Oxford University Press. Wickham, C. (1984). ‘The Other Transition:  From Ancient World to Feudalism’, Past and Present, 103: 3–​36. Wickham, C. (2005). Framing the Early Middle Ages: Europe and the Mediterranean, 400–​800. Oxford: Oxford University Press. Yorke, B. (1990). Kings and Kingdoms of Early Anglo-​Saxon England. London: Routledge.

868

Index

Figures and tables are indicated by “f ” and “t” following page numbers. Acculturation, 276, 288, 386, 836 Actor–​Network Theory, 499 Adams, J. N., 578, 588–​589, 592n8 Adler, P. S., 225 Adolescence. See Childhood and adolescence ADS (Archaeology Data Service), 50 Aerial photography, 54–​55, 30, 700, 703, 715, 716f Age. See also Life course biological, 322, 323 burial patterns and, 313–​314, 322–​323, 325–331, 327f, 328t, 330t chronological, 322, 331 of decapitated burials, 391f, 392–​393 epitaphs and, 327–​328, 329 gender and, 322 identity and, 304, 314, 322 at marriage, 305, 329 of skeletal remains, 307, 323 social, 322, 324 Agricola (Tacitus), 4, 8, 110, 226 Agricola, military campaigns under, xxx–​xxxi, 184 Agriculture. See Farming Akerman, J., 65 Albarella, U., 798, 800 Allason-​Jones, Lindsay, 365, 366, 464 Allectus, xxxiii–​xxxiv Allen, Denise, 470 Allen, J., 517–​518 Allison, Pim E., 362, 366, 373, 450 Amalgam-​gilding technique, 534 Ammianus Marcellinus, 110 Amphitheatres, 759, 779, 780, 782 Anderson, A. W., 227 Androcentricism, 366, 367

Anglesey, Roman conquest of, xxx, xxxi Animals, 791–​806 birds, 654, 800–​801 cattle, 651, 792–​795, 814 dogs and cats, 653–​654, 799 donkeys and mules, 799 fish, 801 horses, 798–​799 hunted animals, 799–​800 overview, 791 pigs, 797–​798 ritual deposition of, 646, 650, 651, 653–​654 sheep and goats, 650, 795–​797 zooarchaeological evidence for, 791–​792, 802 Antiquarians, artefact studies by, 63–​64 Antonine Wall abandonment of, xxxii building process, 185 forts along, 14 as frontier, 224 inscriptions from, 8, 293 votive altars at, 622 Antoninus Pius, xxxi–​xxxii, 183f, 185 Appearance. See Personal appearance Arable farming, 808–​810, 812–​814 Archaeologia Aeliana (journal), 23, 28, 64 Archaeology. See also Bioarchaeology; Romano-​British archaeology behavioural archaeology, 486 cognitive archaeology, 487, 488–​489 funding for projects, 23, 24, 48–​50, 179–​180 in higher education, 32–​33, 35 memory and, 682–​684 multiculturalism in interpretations, 216–​217

870

870   Index Archaeology (Cont.) New Archaeology movement, 32, 382, 383, 485 professionalization of, 44, 48 rescue archaeology, 44, 47, 50, 26, 27, 32 symmetrical archaeology, 487, 488 in United Kingdom, 43–​44 zooarchaeology, 791–​802. See also Animals Archaeology Data Service (ADS), 50 The Archaeology of 'Race': Exploring the Northern Frontier in Roman Britain exhibition (2009), 216 Archaeology of Ritual and Magic (Merrifield), 82 The Archaeology of Roman Britain (Collingwood & Richmond), 77, 483 Arch of Constantine (Rome), 685 Aristotelian approach to identity, 246 Armour, 452, 465–​468 Army. See Military community Art, 599–​618 Celtic, 181–​182 coinage as, 600 dating challenges for, 602 distribution of, 600–​602, 601f interpretations of, 612–​614, 615 mosaics. See Mosaics overview, 599–​600 paintings. See Wall-​paintings patron–​client relationships in, 607–​608 pottery as, 600 producers of, 606–​609 quality of, 602–​604, 606, 614–​615 religious, 610, 613 sculpture. See Sculpture users and uses of, 609–​611, 615 Artefact studies, 63–​96. See also Material culture; Materiality; specific types of artefacts in 1930s–​1970s, 76–​79, 79 in 1970s and beyond, 79–​83 antiquarian interest in, 63–​64 archaeological approaches to, 64–​73, 66–​7 1t classification and interpretation of, 65, 69–​76 overview, 63

profile of artefact researchers, 82–​83 publication of, 73–​74 social use of space derived from, 188 technological innovations in, 83–​84 text-​based scholarship vs., 83, 84 by women, 72, 77, 78f, 81–​82, 82–​83 Arthropathy, 397 Arthur, P., 519 Ashanti tribe, 388 Aspöck, E., 389, 400 Assman, Aleida and Jan, 682 Atlas of Roman Rural Settlement (Taylor), 687, 708 Augustus, xxix, 152, 163, 685 Aurelian, xxxiii Aylesford–​Swarling tradition, 159–​160, 511, 689 Babenberg dynasty, burial practices of, 388 Baker, Patricia, 367, 555 Balty, J. C., 769 Baptism, 670–​672 Barates of Palmyra, 214, 238, 240, 298 Barbarino, J. L., 587 Barbarous radiate coinage, 842–​843 Barber, B., 387 Barber, E. W., 419 Barker, Phil, 27, 31 Barley, 808, 814 Barrett, A. A., 105 Basilicas, 122, 492, 495, 666, 769, 772 Bathing, 349, 351, 373, 495, 714, 780–​782 Bath tablet collections, 100, 108, 109, 580–​583, 581f, 583f Bayesian statistical analysis, 179 Bayford grave assemblages, 349, 351 Bayley, J., 420 Beaumont, J., 315 Beer production, 810, 812–​813, 861–​862 Behavioural archaeology, 486 Biddle, Martin, 27 Bignor villa, excavation of, 9, 10f, 26 Bilingualism, 577–​578, 579, 584, 588, 591 Bioarchaeology of childhood, 306–​308, 315, 316 in gender studies, 372, 374 medicine and, 567

Index   871 Biological age, 322, 323 Birdoswald site cemetery excavations at, 236, 457–​458 geophysical surveys of, 55 purse hoard from, 840 Birds, 654, 800–​801 Birley, Anthony R., 31, 37, 455 Birley, Eric, 43, 25–​26, 27–​28, 455 Birley, Robin, 26 Biró, M., 622 Bishop, Mike, 468 Black, E., 654 Blacksmithing, 548, 549 Blagg, Tom, 36, 774 Boar, 800 Bonsall, L., 374 Boon, G. C., 76, 78 Booth, Paul, 434, 522 Boudican Revolt (AD 60), xxx, 117, 119, 121, 459–​460, 514 Bowman, A. K., 96, 104 Boylston, A., 384 Bradley, Richard, 50, 683 Brailsford, J. W., 517 Brass, 532, 538, 541 Braund, D., 163 Bread wheat, 814 Breastfeeding, 308–​309 Breeze, David, 36, 227 Brigantes tribe, xxx Britain. See Roman Britain; United Kingdom (UK) Britannia (Camden), 3–​4, 5, 7 Britannia: A History of Roman Britain (Frere), 43, 30, 483 Britannia: The Creation of a Roman Province (Creighton), 125 British School at Rome (BSR), 28, 29 Brochs, 179, 180, 190, 194, 606 Bronze, 493, 532, 538, 540, 839, 844 Bronze Age burials, 314, 345 continental contacts in, 155 feasting in, 161 gold in, 157 hillfort construction in, 156 monuments of, 686–​687

Brooches artefact studies of, 71, 72, 73, 78 British adoption of, 159 at burial sites, 257–​257 Celtic nature of, 76, 82, 188 chronological and regional distribution of, 411–​413, 412t, 414f, 415–​416 coloured, 415–​416 cross-​bow, 275–​276, 370, 13, 417 evidentiary sources for, 420 gender and, 135, 370, 413, 415 metal content of, 540, 540f military, 469 in mobility studies, 207, 249, 250, 250f, 255–​258 ritual deposition of, 655 Brougham burial site age-​related patterns at, 343, 368 child burials at, 324 cremation at, 437 diversity of remains at, 214 gender patterns at, 343, 368, 372 inscriptions at, 439, 458 material culture at, 236–​237, 275, 346 Bruce, John Collingwood, 14 Bruhn, J., 275 BSR (British School at Rome), 28, 29 Building dedications, 287, 292 Burials, 341–​362. See also Cremation; Funerary process; Tombstones; specific burial sites age-​related patterns in, 313–​314, 322–​323, 325–​331, 327f, 328t, 330t Bronze Age, 314, 345 brooches at, 256–​257 of children, 310–​315 Christian, 672–​674 coffins for, 348–​348, 394f, 396, 428, 430–​431 decontextualization of, 145 demarcation of, 353–​351, 439 destruction and degradation of, 343 deviant. See Deviant burials family circle graves, 16, 17f in frontier zones, 236–​237 gender patterns in, 313, 326–​329, 328t, 330, 330t, 368–​369 of infants, 311–​312, 325, 369, 428, 715

872

872   Index Burials (Cont.) Iron Age, 151, 159–​161, 344, 345 material culture in, 145, 236–​237, 349–​352, 350f, 352f, 437–​438 in military community, 457 overview, 341–​342 plaster, 347, 430 pre-​interment rituals for, 345–​347 regional variations in, 331–​333, 332–​333f spatial organization of, 352–​353, 368–​369, 433–​434 status and, 309–​310, 314, 341–​345, 354–​356 urban vs. rural, 343, 353–​354, 431–​432, 735–​736, 736f Bushe-​Fox, J. P., 26, 28, 73, 74, 75 Butcher, S., 420 Butt Road cemetery age-​related patterns at, 322 baptismal font at, 671 child burials at, 309–​310, 313–​314, 325 church structures at, 667 clothing at, 407 gender patterns at, 313 spatial organization of, 433, 673 Caerwent site amphitheatre at, 779 bathing complex at, 781 cattle at, 794 coin assemblages from, 844 development of, 745, 755 forum at, 492, 768, 769, 770f, 772 pigs at, 797, 798 transformation of, 495 twentieth-​century excavations of, 23 Caesar. See Julius Caesar Caistor-​by-​Norwich bathing complex at, 781 development of, 755 forum at, 769, 772 public buildings at, 785 temples at, 776–​777, 778 theatre at, 779 Caligula, xxix Cambridgeshire, decapitated burials in, 389–​ 399, 390t, 391–​395f, 398f Camden, William, 3, 5–​6, 7, 18, 19

Camulodunum site construction of, 756 continuity at, 128–​130 Iron Age settlement at, 26 pottery from, 28 Canterbury bathing complex at, 495, 781 burial sites at, 167, 356, 467 coin assemblages from, 844 establishment of, 165, 755 grave goods at, 438 silver hoard from, 664 theatre at, 789, 779 urban structure of, 774, 775f Cantonal system, 748 Caracalla, xxxiii, 185, 255, 344 Carausius, xxxiii, 452 Carbon isotope analysis, 212–​213 Care of children, 307–​310 Carlisle site, 98, 99, 184, 746 Carroll, M., 204 Casey, John, 32, 80, 85 Cassius Dio, 110, 119, 152, 248, 746–​747 Castgore site, 853, 854–​855, 857–​858 Cats, 799 Catterick burials, 370, 371 Cattle, 651, 792–​795, 814 CBA (Council for British Archaeology), 29, 52 The Celt, the Roman and the Saxon (Wright), 11 Celtic–​Latin bilingualism, 577–​578, 579, 584, 588, 591 Celtic peoples art of, 181–​182 languages of, 575–​578, 580–​584, 581f as subalterns, 4–​5, 8–​10, 13, 18 Cementation technique, 541 Cemeteries. See Burials; Tombstones Centuries of Childhood (Ariès), 303 Ceramics. See Pottery Cereal crops, 808, 809, 813 Chamberlain, A. T., 311 Champion, Timothy, 150 Charles-​Edwards, T., 588 Charlesworth, D., 76 Cheesman, G. L., 24 Chesters, excavated forum at, 14, 16f

Index   873 Chickens, 800 Childhood and adolescence, 303–​321. See also Infancy; Life course age and sex determination, 306–​307 bioarchaeology of, 306–​310, 315, 316 classical perspectives on, 304–​306 culturally specific constructions of, 303, 304, 306 diet in, 308–​309 diseases of, 308, 309 funerary evidence of, 310–​315 future directions for research, 315–​316 health and care considerations in, 307–​310 overview, 303–​304 rites of passage in, 305 Childhood in the Past (journal), 304 Chi-​rho symbol, 663, 664, 668–​670, 669f, 671 Chorography, 5–​6 Christianity, 660–​680. See also Religion baptism in, 670–​671 in Britain, 11, 143, 387 churches and, 143–​144, 409, 454, 665–​670, 667f, 669f death and burial in, 672–​674 documentary evidence for, 661–​663, 674 early Christianity, 660–​661 in Gaul, 143 hairstyles influenced by, 419 imagery and symbolism in, 663–​664, 668–​670 martyrdom and, 662 material culture and, 663–​665, 665f in military community, 454 in mosaics, 669–​670, 671 overview, 660 paganism, relationship with, 675 regional variations in, 675–​676 in wall-​paintings, 143, 661, 668–​669, 669f Chronological age, 322, 331 Churches, 143–​144, 409, 454, 665–​670, 667f, 669f Cicero, 681, 682 Cirencester site amphitheatre at, 759, 780 cattle at, 793, 794 fort at, 754 forum at, 769, 771, 772

graves goods at, 329, 330t organization of excavations at, 27, 45, 45f pottery at, 523 public buildings at, 785 sheep at, 795 sustained occupation of, 14 temple at, 776 theatre at, 782 Cities. See Urban areas Citizenship, Roman, 232–​233, 452 Civil districts, 14, 15f, 18 Clarke, G., 208 Clarke, S., 188, 642, 652 Classicianus, 103, 103f, 110, 293, 353, 434 Claudia Rufina, 7, 254, 410 Claudius, xxix, 6, 119, 127, 341 Clay, C., 204 Cleere, H. F., 546, 549 Client-​kingdoms, 510, 511–​513, 514 Clodius Albinus, xxxii Clothing, 406–​424. See also Brooches; Personal appearance accessories, 135–​136, 159, 249 evidentiary sources for, 406–​410, 419–​420 footwear, 410, 420, 438, 450 gender and, 369–​371, 409 manufacturing process for, 407–​408 military, 416–​417, 419–​420, 452 overview, 406 religion and, 418 styles of, 409–​409 togas, 409, 418 tube dresses, 409, 413 tunics, 409, 416 Coffins, 348–​349, 394f, 396, 428, 429–​430 Cogidubnus inscription, 6–​7, 7f, 8 Cognitive archaeology, 487, 488–​489 Cohen, R., 214 Coinage, 834–​849 artefact studies of, 64, 65, 75, 76, 79, 85 as art object, 600 bronze, 493, 839, 844 cessation of coin supply, 844–​845 distribution of, 600 as economic evidence, 77 in first century AD, 838–​840 in fourth century AD, 843–​844

874

874   Index Coinage (Cont.) in Gallo-​Roman horizon, 162, 266, 512, 512f, 513 gold, 137, 157–​158, 159, 163–​164, 532 inscriptions on, 9, 101, 164, 575–​576 Iron Age, 107, 152, 157–​159, 161–​162, 836–​838, 837f Latin on, 575–​576 overview, 834, 836 pioneering studies on, 32 radiate coinage, 841–​843 regional distribution of, 512, 512f, 513, 516f ritual deposition of, 161–​162, 644–​645, 647, 648, 841 Romanization of, 16, 163–​164, 510 in second and third century AD, 840–​843 silver, 137, 164, 195, 493, 537, 541, 541f striking and distribution of, 136–​137 textual evidence from, 107 transformations in, 491–​492, 493–​494 Coins and the Archaeologist (Casey & Reece), 80 Coins of the Ancient Britons (Evans), 151 Colchester site age-​related burial patterns at, 326, 327f, 328 cattle at, 793, 795 church structures at, 666–​667 construction of, 752–​754, 754f continuity at, 128–​130 dating evidence at, 128 in early Roman horizon, 121, 122 forum at, 772 industrial activity at, 783 industrial production at, 164–​165 infant burials at, 325 location of, 750 military settlements at, 120 pigs at, 797 plaster burials at, 347 pottery from, 513, 519, 730 sheep at, 796 site reports from, 80 temple complex at, 774, 776, 785 theatre at, 779, 781 veterans at, 459–​460 Coleraine hoard, 845, 845f Collective memory, 681, 682, 691

Collingwood, R. G., xxvii, 24–​25, 76, 100, 105, 482–​483, 602, 603, 720–​721, 722, 731, 737 Collins, R., 417 Collyrium stamps, 560–​562 Colonial diasporas, 214, 214 Colostrum, 309 Coloured brooches, 415–​416 Commemoration stage of funerary process, 438–​439 Commius, 163, 264 Commodus, xxxii Community archaeology groups, 52 Condron, F., 549 Conkey, M. W., 363 Conolly, R., 579 Constans, xxxiv Constantine, xxxiv, 454, 662, 685 Constantine III, xxxv, 860, 862 Constantius, xxxiii–​xxxiv Construction process for urban areas, 752–​ 760, 753–​754f, 757f Cook, James, 150 Cool, Hilary, 214, 236–​237, 240, 327, 328, 406, 420, 451, 458, 520, 522, 526 Cooper, N., 731 Coote, Henry, 11 Copper and copper alloys, 532, 538–​542, 540–​ 541f, 548–​549 Corbridge hoard, 467, 472 Core-​periphery models, 689, 721 Corn-​driers, 808, 813 Cosh, Stephen, 37 Cosmetics, 411, 416 Cotswold Archaeology, 28, 50, 699 Council for British Archaeology (CBA), 29, 52, Council of Arles, 662, 663 Cox, John, 72 Craniomorphometric analysis, 215–​216, 215f Creighton, J., 125, 163, 164, 267, 277, 510, 511–​ 513, 516, 749, 755, 758 Cremation. See also Funerary process ageing and sexing remains from, 323, 368 Aylesford–​Swarling tradition, 159–​160, 511, 689 British adoption of, 168, 434 containment methods, 159–​160, 347–​348, 437

Index   875 of horses, 236 Late Iron Age, 151, 263–​264, 263f, 265f, 428, 436–​437 locations for, 431–​433 pre-​interment rituals for, 345–​346 rituals in, 435 as technology of remembrance, 686 Creolization, 214, 437 Crerar, Belinda, 381 Cribra orbitalia, 309 Croom, A., 419 Crops. See Farming Cross-​bow brooches, 275–​276, 370, 413, 417 Croxford, Ben, 599 Crummy, N., 75, 80, 82, 154, 560 Cult practices, 453–​458, 778, 779 Cunetio hoard, 834, 841, 843 Cunliffe, Barry, 27, 28, 32, 127, 153–​154, 156, 517 Curle, Alexander, 73 Curle, James, 73, 74, 80, 186 Currency bars, 155 Curse tablets (defixiones) content and subject matter of, 108–​110 deities and, 100, 108, 109, 622, 627 in Gaulish, 582 in Latin, 110, 585–​587, 586f sociolinguistic evidence from, 574 textual evidence from, 99–​101, 99f Uley curse tablets, 99f, 100, 108–​109, 574, 582, 628 Cybele, cult of, 778 Dalton, O. M., 73, 76 Danebury Environs project, 706, 708, 809 Daniel, Glyn, 52 Davidson, C., 368 Decapitated burials, 389–​399 age of, 391f, 392–​393 coffins for, 394f, 396, 428 explanations for, 384–​385, 387–​388 funerary process for, 428 gender of, 392f, 393–​394 health, disease, and disability in, 397 material culture in, 393f, 394–​396 methodology for analysis, 389–​392, 390t as pagan rites, 387 posture of, 395f, 396–​397

spatial organization of, 397–​399, 398f Deer, 799–​800 Defixiones. See Curse tablets Deities, 619–​640 curse tablets and, 100, 108, 109, 622, 627 distribution and dating of, 622–​626, 623f, 625f double-​named, 633–​635 evidentiary sources on, 620–​622, 635–​636 geographic nature of, 630–​632, 631f healing powers of, 564 imperial vs. local, 629–​630 inscriptions dedicated to, 108, 292, 621–​622 of military community, 456, 619 offerings to. See Ritual deposition; Votive altars overview, 619–​620 sculptures involving, 610 vows pledged to, 626, 627 DeLaine, J., 781 Demographic determinism, 311 Demography, statistical analysis on, 292–​293 Dental diseases, 733–​734 Deposition stages of funerary process, 436–​438 Derks, T., 237, 272, 633, 634, 635 Deviancy, defined, 381, 382 Deviant burials, 381–​405 characteristics of, 385 decapitated. See Decapitated burials explanations for, 384–​385 funerary process for, 428 historical approaches to, 381–​386 individuals associated with, 385–​386 overview, 381–​382 in Romano-​British archaeology, 388–​389, 400 vicarious human sacrifice and, 384 Diasporas in Roman Britain project (2007–​9), 207, 211–​213, 215–​216, 216 Diaspora theory, 213–​214, 216–​217 Díaz-​Andreu, M., 363 Dickinson, Brenda, 470 Diet and nutrition. See also Farming in childhood, 308–​309 cultural influences on, 272–​274

876

876   Index Diet and nutrition (Cont.) dental diseases and, 733–​734 gender and, 374 lead in, 308, 427 meat in, 792–​793, 794, 802 North African, 821 for ritual and ceremonial activities, 822 in urban vs. rural areas, 734 variety and diversity in, 81, 819–​823 Dietary deficiency diseases, 308, 309, 315 Dietler, M., 161 Diocletian, xxxiii, 454 Diodorus Siculus, 152, 542 Disability among decapitated burials, 397 Diseases. See also Medicine agricultural, 810, 811–​812f of childhood, 308, 309 decapitation burials and, 397 dental, 733–​734 dietary deficiency, 308, 309, 315 gender and, 567 Diversity. See Ethnicity; Multiculturalism; Race Division of labour, 371–​372, 374 Dobney, K., 524, 794 Doctors, 563, 564 Dogs, 653–​654, 799 Domestic structures, 373–​374 Domitian, 184 Donkeys, 799 Dorchester settlement amphitheatre at, 780 bathing complex at, 781 cattle at, 794, 795 chronology of occupation, 157 construction of, 752, 753f, 755, 783 lead coffin liners at, 347, 348 pigs at, 797, 798 ritual deposition in, 653–​654 sheep at, 796 Double-​named deities, 633–​635 Dress. See Clothing Druids abolishment from Gaul, 560 military campaigns against, xxx pictorial representations of, 153 training of, 270

Duality of structure, 246 Ducks, 801 Dungworth, David, 532 Dunning, G., 76, 151 Durham University, 25, 27, 45, 51, 33, 216 Early Roman horizon, 117–​133 Colchester in, 121, 122 Colonia Claudia in, 121, 122 continuity in, 126–​130 evidentiary challenges and biases in study of, 117–​118 Londinium in, 118–​119, 121, 122–​123 material culture of, 124–​125, 130–​131 overview, 117 public buildings in, 123–​124 regional diversity in, 118 towns in, 123–​126 urbanization in, 125–​126 Verulamium in, 121–​122 East Anglia building types in, 142 Christianity in, 675–​676 folk customs in, 65 hinterlands of, 725–​726, 728–​730 pottery in, 511 revolts in, xxx Eckardt, Hella, 203, 683 Economics gender and, 371–​372 gross domestic product and, 853 laissez-​faire, 722 power and, 850, 856, 860, 862 systems of, 138–​140, 608 taxation and, 136, 227, 526, 721–​722, 855, 856 theoretical approaches to, 851, 852 urban and rural relationships, 720–​722, 723–​724, 729–​731 Edict of Milan (AD 313), 660, 661, 663 Edict of Thessalonica (AD 380), 661 Educational networks, 33, 34–​35t Educational programs for archaeology, 32–​33, 36 Elite auto-​representation, 142, 143, 144 England historic preservation responsibilities in, 44

Index   877 research framework in, 47 English Heritage historic preservation responsibilities of, 44 National Mapping Programme, 54–​55 research framework developed by, 47 research funding from, 50 Entertainment in urban areas, 779–​780, 781 Epigrams (Martial), 7 Epigraphic habit dynamic nature of, 293 inscriptions and, 288–​290, 456, 622 in military community, 95, 237 mortuary, 439 sculpture and, 609 Epigraphy Christian, 673–​674 inscriptions and, 204–​205, 247, 248–​249, 253–​254, 253f military community and, 471 in mobility studies, 204–​206, 205f, 248–​249, 251–​255, 252–​253f Epitaphs age-​related patterns in, 327–​328, 329 on coffins, 348 content of, 291–​293 decline in, 300 function of, 107, 287 gender and, 326, 328, 329 prevalence of, 287, 289 Erdrich, M., 193 Esmonde-​Cleary, Simon, 129, 427, 767–​768 Ethnicity. See also Multiculturalism; Race definitions of, 223–​224, 246 identity and, 246, 257, 278, 458–​459 material culture and, 136, 207 in military community, 458–​459 Evans, Arthur John, 16–​17, 73, 151, 153 Evans, C., 75 Evans, Jeremy, 108, 208–​209, 211–​212, 510, 521, 523–​524, 592n6 Evans, John, 151 Everill, P., 45 Excarnation, 434 Excavations in Cranborne Chase (Pitt-​Rivers), 71 Eye remedies, 561, 562 Eyers, E., 311–​312

Family circle graves, 16, 17f Family structures, 372–​374 Farley, J. M.-​A., 168 Farming, 807–​833 arable farming, 808–​810, 812–​814 decline in, 862 diet impacted by, 817, 819–​823 future research needed, 827 germination of grains in, 809 horticulture and, 814–​817, 815f, 818f modes of preservation for, 815–​816, 815f overview, 807–​808 pest and diseases in, 810, 811–​812f regional variations in, 823–​827, 828 scale of production in, 809, 813–​814 vineyards, 817 Farmsteads, 127, 128–​129, 131, 706, 708 Faulkner, N., 783 Faussett, B., 64 Feasting deposits, 161, 690, 715 Females. See Gender; Women Field, D., 687 Figs, 819, 820 Fish, 801 Fishbourne site dating evidence at, 127 pigs at, 797 research framework for, 47, 27 villa at, 270, 704–​706, 705f Fitzpatrick, A. P., 161 Flavian invasion of Scotland, 183f, 184–​185 Flavius Cerialis, 104 Folkestone villa, 26, 74 Folly Lane burial site age-​related patterns at, 323 creation of, 344 cremation at, 264 laying-​out stage at, 428 military items at, 167 pottery from, 730 wealth at, 160, 165 Food. See Diet and nutrition Footwear, 410, 420, 438, 450 Forensic ancestry assessment, 215–​216 Forum, as central feature of urban areas, 768–773, 770–​771f Fowl, 800

878

878   Index Fox, Aileen, 27, 65 Franks, A. W., 71–​73, 306 Freestone, I. C., 511 Frere, Sheppard, 43, 27, 30, 31, 32, 79, 121, 483, 748, 755, 769, 783 Frisian peoples, 235–​236, 459, 470 Frocester villa, 515, 706, 707f Frome hoard, 54, 841, 842f, 843 Frontier zones burial sites in, 236–​237 identities in, 225–​227 indigenous populations of, 226 lifestyle in, 182, 185–​189, 187f, 194 material culture in, 188–​194, 191f, 226 Fulford, Mike, 26, 36, 36, 234, 517–​518, 644, 652 Fuller, B. T., 308–​309 Funding for archaeological projects, 44, 45, 48–​50, 179–​180 Funerary process, 425–​447. See also Burials commemoration stage, 438–​439 deposition stage, 436–​438 evidentiary sources for, 425, 426t interpretation and translation of, 440–​441 laying-​out ceremonies in, 429–​431 locations for, 431–​434 modification stage, 434–​435 preparation stage, 427–​431 selection stage, 426–​427 Gaffney, V., 725 Gager, J. G., 109 Gale, Robert, 6 Galen, 307, 735 Galinsky, Karl, 684 Gallic empire, xxxiii Gallicization, 262, 273, 274, 276–​278 Gardening, 814, 816, 817, 823, 826 Gardner, Andrew, 228, 406, 481 Garments. See Clothing Gaul abolishment of druids from, 560 building types in, 141, 142, 143–​145 cemeteries in, 145 Germanic peoples in, 144, 145 immigrants from, 274 inscriptions in, 288

interactions with Britain, 262, 263–​272, 265f, 269f, 273–​275, 276–​278 militarization of, 136, 141, 143 patron–​client relationships in, 158–​159 pottery from, 73, 139, 140, 273, 470, 511 religion in, 143 revolts in, xxxiv Roman conquest of, xxviii, xxxiii, 491 withdrawal of imperial presence from, 135, 137 GDP (gross domestic product), 853 Geese, 801 Gell, Alfred, 485, 488, 499 Gender, 363–​381. See also Women age and, 322 artefact studies and, 71, 76, 78f, 80–​81, 82–​83 being vs. doing, 364 bioarchaeology of, 372, 374 brooches and, 135, 370, 413, 415 burial patterns and, 313, 326–​329, 328t, 330, 330t, 368–​369 cultivation techniques and, 826–​827 of decapitated burials, 392f, 393–​394 defining, 363–​365 diet and, 374 diseases and, 567 division of labour and, 371–​372, 374 in dress, adornment, and appearance, 369–​371, 409 economic activity and, 371–​372 epitaphs and, 326, 328, 329 in family and domestic structures, 372–​374 future directions for research, 374 identification of, 367–​374 in life course, 305 material culture and, 367–​371 mobility and, 205, 252, 254, 256 overview, 363 Romano-​British gender studies, 365–​367 as social construct, 363–​364 Geographia (Ptolemy), 110, 181, 226, 747 Geophysical surveys, 55, 55f, 38, 700, 701–​703, 702f Germanic Mode of Production, 154 Germanic peoples in Gaul, 144, 145 on Hadrian's Wall, 204

Index   879 languages of, 560, 561 in Roman Britain, 144 Germanization, 262, 273, 275, 276 Germany diplomacy and interference in, 194–​195 immigrants from, 274 inscriptions in, 288–​289 institutional support for archaeology in, 73 interactions with Britain, 262, 268, 272, 273–​278 Germination of grains, 809, 810 Gerrard, James, 850 Gildas (monk), 8, 454, 662, 861 Gillam, John, 28, 79 Gini coefficient, 735 Globalization, 484, 487, 499, 500, 737 Goats, 795–​796 Gods. See Deities Gold coinage produced from, 137, 157–​158, 159, 1663–​164, 532 sites for extraction of, 533, 533f techniques for working with, 533–​534, 535, 549 uses for, 532, 534–​535 Golden, M., 303 Gosden, C., 167, 168 Gowland, Rebecca, 303, 307, 308, 311, 313, 322, 324, 325, 328, 330, 331 Graffiti scratches, 100, 101f, 105, 108, 592n6 Grain weevils, 810, 811f, 819–​821 Gratian, xxxv Gratwick, A. S., 577, 585 Graves. See Burials; Tombstones Greenfield, Ernest, 44 Griffin, R., 310, 735 Grimes, W. F., 27 Grog-​tempered pottery, 513, 514, 515, 516 Grooming tools, 411 Gros, P., 767 Gross domestic product (GDP), 853 Grupe, G., 209 Guildhall Museum exhibition (1962), 29 Gutteridge, A., 685 Habitus, 246 Hadrian, xxxi, 185

Hadrian's Wall armour and weaponry from, 468 coinage from, 844 conflicts regarding, 185 as defensive system, 137 forts along, 14, 138 frontier life beyond, 182, 185–​189, 187f, 194 geophysical surveys of, 55 Germanic peoples on, 204 inscriptions from, 8, 291 invasions across, xxxii keyhole excavation on, 24 materiality of, 234–​235 military presence along, 227–​231, 229f, 230–​231t multiculturalism at, 217, 224–​225, 228–​2331, 229f, 230–​231t, 240–​241 Newcastle excavations on, 45, 51 purpose of, xxxi research framework for, 47 as tourist attraction, 52 votive altars at, 622 writing tablets from, 26 Hairstyles, 410, 413, 419, 431 Halbwachs, Maurice, 682 Halkon, Peter, 46 Hall, J., 387, 782 Hamp, E. P., 592n8 Handley, M., 205, 206 Hanson, W. S., 579 Hapsburg–​Lorraine dynasty, burial practices of, 388 Hares, 800 Harris, W. V., 102, 622 Haselgrove, C. C., 721 Hassall, Mark, 33, 582 Haverfield, Francis, 3, 4–​5, 13–​14, 16–​17, 18–​19, 22, 23–​24, 73, 74, 75, 482, 576, 747–​748 Hawaii, social transformation of, 150 Hawkes, Christopher, 26, 28 Hawkes, S., 76, 151 Hayling Island temple, 270, 272, 689, 690, 837 Haynes, Ian, 228, 448 Health. See also Diseases; Medicine in childhood, 307–​310 decapitated burials and, 397 inequality and, 731, 735 in urban vs. rural areas, 373–​734, 733f

880

880   Index Health stress, 307 Heidegger, Martin, 485 Henig, M., 105, 603, 621, 633, 635 Herbs, 815–​817, 820–​821, 822, 824–​825 Heritage Lottery Fund, 44, 52 Herodian, 344, 434 Herodotus, 542 HERs. See Historic Environment Records Heterarchy, 154 Heterosexual bias, 364, 365 Higham, N. J., 590 Higham Ferrers settlement, 644, 646–​648, 708, 709–​7 10f, 853, 854f Hill, J. D., 82, 154, 365, 373, 511, 512, 525, 689 Hillforts, 154, 156, 491, 690, 741–​742 Hingley, Richard, 3, 216, 224–​225, 240, 373, 484, 644, 652, 721–​722, 729, 730, 745 Historic Environment Records (HERs), 44, 50, 54, 579 Hoare, Richard Colt, 11, 17 Hodder, Ian, 32, 82, 487, 489, 851, 852 Hodgson, N., 226, 227 Hod Hill site, 28 Holland, Sally, 310 Holocaust memorials, 684 Holtorf, C., 683 Homogenization, superficial, 248 Homosexuality, 364–​365 Honorius, xxxvi Hope, Valerie M., 237, 285 Horses, 798–​799 Horticulture, 814–​817, 815f, 818f Hospitals, 563, 565–​566 House churches, 409, 668–​669 Housesteads fort architecture of, 470–​471, 495 church structures at, 669, 670 medical evidence at, 564, 566 textual evidence at, 472 Hoxne hoard, 856–​857 Hull, M. R., 28, 77 Hunter, Fraser, 179, 226 Hunting, animals for, 799–​800 Hurst, Henry, 95, 784, 786 Hutton, R., 687 Hybridization, 214, 276 Hydrocephalus, 397

Identity age, 304, 314, 322 appearance and, 327, 406, 417–​421 categories of, 248 cult practices and, 453–​458 defined, 225 diasporic, 214 ethnic, 248, 257, 278, 458–​459 fluidity of, 233, 248, 277, 304 in frontier zones, 225–​227 individual expressions of, 237–​240, 239f inscriptions and, 295–​298 material culture as indicator of, 206–​207, 226 of military community, 417, 448–​449, 451–​459 mobility and, 245–​246 Identity and Violence (Sen), 448 Immigrants. See also Mobility foods brought by, 821 from Gaul and Germany, 274 identification of, 214–​217, 248–​250, 250f Imperial diasporas, 214–​215, 217 Imperial economy, 138, 139 Imperialism materiality and, 481, 482, 489, 492–​493 rural areas impacted by, 721, 732, 736 An Imperial Possession (Mattingley), xxvii, 449 Inchtuthil site, 28, 184–​185, 566 Indigenous populations. See also specific groups of frontier zones, 226 land incorporated into Roman Empire, 117 Romanization of, 4, 13–​14, 17 Inequality in rural areas, 732–​736, 736f, 737 social stratification and, 856 Infancy. See also Childhood and adolescence; Life course burial practices in, 311–​312, 325, 369, 428, 715 dependency in, 306 dietary deficiency diseases in, 308, 309 feeding practices in, 308–​309 personhood in, 312, 325–​326 Infanticide, 311–​312, 325 Ingold, Tim, 488

Index   881 Inscriptions, 285–​302. See also Epitaphs; Textual evidence abbreviations in, 291 building dedications, 287, 292 categories of, 287 Cogidubnus inscription, 6–​7, 7f, 8 on coinage, 9, 101, 164, 575–​576 communication to audiences through, 290–​295 dating, 289 for deities, 108, 292, 621–​622 epigraphic analysis of, 204–​205, 247, 248–​ 249, 253–​254, 253f epigraphic habit and, 288–​290, 456, 622 from Hadrian's Wall, 8, 291 identity and, 295–​298 in Latin, 5, 8, 18, 106, 240, 290–​291 literacy and, 291 medical-​related, 562–​564 military, 14, 287, 288–​291, 295–​296, 628 Ogam inscriptions, 588–​590, 589f on pottery, 101 prevalence and distribution of, 286–​287 statistical analysis on demography using, 292–​293 on stone, 286 on tombstones, 237–​240, 239f, 289, 296–​298, 439 on votive altars, 292, 294, 297, 298, 622 Institute of Archaeology (London), 25, 33, 483 Instrumentation, medical, 558–​560 Inventory of Romano-​British Coin Hoards (Robertson), 834 Inveresk military complex, 186, 187f Iron, 532, 542–​548, 544–​545f, 549 Iron Age archaeological evidence for, 152–​153 burials, 151, 159–​161, 344, 345 Celtic languages in, 575 changing perceptions of, 153–​154 coinage from, 107, 152, 157–​159, 161–​162, 836–​838, 837f conceptualizing societies in, 180–​182 hillforts in, 742 late. See Late Iron Age (LIA) material culture in, 10, 26, 155, 159, 167, 689 metalworking in, 490

political authority in, 153, 154, 162 pottery from, 128, 155, 159–​160 ritual practices in, 159–​161, 272 Roman, 180 rural landscape during, 49 in Scotland, 179–​182 settlement and economy of, 10, 26, 155–​157 social organization in, 153–​154 wealth and power in, 157–​162 Iron Age and Roman Coin Finds from Wales Project, 834 Isotope analysis archaeological impact of, 38 carbon, 212–​213 lead, 212–​213 in mobility studies, 204, 207–​214, 212f, 217–​218 nitrogen, 212–​213, 308 oxygen, 208, 209, 210–​212, 212f strontium, 208, 209, 210–​211, 212f of tooth enamel, 308, 309, 315 Itinerarium Curiosum (Stukeley), 6 Ivleva, Tatiana, 207, 245 Ivory Bangle Lady, 214, 216, 216f Jackson, K. H., 576–​577, 578–​579, 584–​585, 586, 587–​588, 592n7 Jackson, R., 560 James, A., 306 James, Simon, 228, 231, 277, 451 Jenny, L. L., 309–​310, 313 Jobey, George, 30, 235 Johns, C., 621 Jones, Barri, 33 Jones, M., 813 Jones, R. F. J., 80, 342–​343 Jongman, William, 732, 735 Journal of Roman Military Equipment Studies, 466 Journal of Roman Pottery Studies, 80 Joy, J., 161 Joyce, James Gerald, 14 Julius Caesar on appearance of British, 410 expeditions to Britain, xxviii–​xxix, 162–​163, 510–​511, 575

882

882   Index Julius Caesar (Cont.) Gallic War account by, 264, 268 writings of, 4, 9, 128, 152, 154, 158 Jupiter columns, 272 Kakoschke, A., 204 Kamash, Zena, 681 Kampen, N. B., 371 Kennedy, D., 251 Kenyon, Kathleen M., 27, 28, 75–​76 Kiernan, P., 646 Kincheloe, J. L., 225 King, Anthony, 273, 274, 646, 647, 690, 723, 792 King, Charles, 72 King Harry Lane burial site, 165, 343, 411, 413, 429, 436 Kingscote settlement, 602, 709–​7 10 Kipling, Rudyard, 13 Kirk, J., 77 Knight, J. K., 673 Knook Castle, 11, 12f Kolb, M., 854 Küchler, Susanne, 683–​684 Kyphosis, 381 Labour, division of, 371–​372, 374 Laissez-​faire economics, 722 Landownership, 701, 745 Landscapes. See also Rural areas; Urban areas aerial photography of, 30, 54–​55, 700, 703, 715, 716f chorography and, 5–​6 geophysical surveys of, 55, 55f, 37–​38, 700, 701–​703, 702f memory and, 683 reorganization of, 687, 688 research philosophy on, 32 Language. See Sociolinguistics Language and History in Early Britain (Jackson), 576–​576 Lankhills cemetery age-​related patterns at, 313, 322, 324, 326–327, 327f, 328 developer-​funded excavations of, 49 deviant burials in, 384, 428 diversity of remains at, 209

gender patterns at, 313, 326–​327, 328, 330 infant burials at, 325 isotope analysis of, 211, 212f material culture at, 275, 276, 329, 342, 350, 417 spatial organization of, 433 Late Antique period, mobility in, 205–​206, 205f Late Iron Age (LIA) coinage in, 836–​838, 837f community identities in, 226–​227 continuity in, 126, 128–​129 cremation in, 151, 263–​264, 263f, 265f, 428, 436–​437 feasting in, 161 funerary process in, 428, 430, 432–​437 models of interaction in, 166–​169 modes of interpretation for, 151–​152 political authority in, 153, 154 pottery from, 128 relationship with early Roman horizon, 117, 118, 126, 128 shrine structures in, 270–​271 transitions in, 125, 126, 127 Latinitas Britannica (Jackson), 584–​585, 588 Latinization, 574, 579, 591 Latin language adoption of nomenclature from, 583–​584 Celtic–​Latin bilingualism, 577–​578, 579, 584, 588, 591 on coinage, 574–​576 on curse tablets, 110, 585–​587, 586f in elite contexts, 584, 588, 592n11 emergence in Britain, 575 filiation markers in, 582–​583, 583f inscriptions in, 5, 8, 18, 106, 240, 290–​291 levels of, 577 literacy in, 168, 576, 579 loanwords from, 577–​578, 584–​585, 587, 593n12 nature of British Latin, 584–​588, 592n8 on Ogam stones, 588–​590 in post-​Roman Britain, 591 on tombstones, 439 Latour, Bruno, 485, 488, 499 Laurence, R., 307, 767–​768

Index   883 Laying-​out ceremonies, 429–​431 Lazzari, M., 233 Lead in coffins, 309, 348, 430, 431 in diet, 308, 427 isotope analysis using, 213–​214 sites for extraction of, 535, 533f techniques for working with, 535, 537 uses of, 532, 538, 540, 540f, 548 Leaf-​gilding technique, 534 Leather products, 410 Lewis, M. E., 307–​308, 309 Lewis, M. J. T., 29 Lexden burial site, 160, 165, 352, 429 LIA. See Late Iron Age Life course, 321–​340. See also Age age identity and, 304 childhood in. See Childhood and adolescence future research needs, 334 gender in, 305 infancy in. See Infancy local and regional identities in, 331–​333 older adulthood in, 329–​331, 330t overview, 321–​322 as social construct, 305, 306 in theory and practice, 322–​324 young adulthood in, 328–​329, 328t Lilley, I., 214 Limes Congress, 56, 27 Ling, R., 105 Linguistics. See Sociolinguistics Literacy. See also Inscriptions; Textual evidence defined, 102 inscriptions and, 291 in Latin, 168, 576, 579 in military community, 104–​105, 471–​472 in Roman Britain, 102–​102, 104–​106, 166 Liversidge, J., 76, 77, 78, 384 Livia, 408 Loanwords, 577–​578, 584–​585, 587, 593n12 Locker, A., 801 Lockyear, K., 84 Londinium in early Roman horizon, 119–​120, 121, 122–​123

material culture in, 124–​125, 130 public buildings at, 123, 124 Lost-​wax casting technique, 539 Love, J. R., 534 Lucas, G., 65, 71, 73–​74, 75 Lucy, S., 225 Lullingstone site burials at, 348 church structures at, 409, 668–​669 sculptures from, 608 textual evidence from, 105, 106f wall paintings at, 151, 409, 668–​669, 669f Lysons, Samuel, 9 Macaulay, Thomas, 150, 151 MacDonald, George, 24 MacDonald, J. L., 384 MacKinnon, M., 798 Mackreth, D. F., 412, 415, 420 MacMullen, R., 287, 622 Magnentius, xxxiv, 663 Magnesian Limestone ridge, 49, 703 Magnetometry, 701, 702f Magnus Maximus, xxxv, 135 Make-​up, 409, 416 Males. See Gender Maltby, Mark, 791 Mann, J., 592n5, 622 Mann, M., 233 Manning, W., 79 Manufacturing process for textiles, 407–​408 Marchitello, Howard, 5–​6 Market economy, 138–​139, 608 Market gardening, 817, 823, 826 Marriage, age at, 305, 329 Martial, 7, 254, 410 Martin, M., 109 Martin of Tours, 143 Martyrdom, 662 Maryport site, 55, 294, 298, 454 Material culture. See also Artefact studies; Materiality at burial sites, 145, 236–​237, 349–​352, 350f, 352f, 437–​438 Christianity and, 663–​665, 665f in decapitated burials, 393f, 394–​396

884

884   Index Material culture (Cont.) of early Roman horizon, 124–​125, 130–​131 ethnicity and, 136, 206 food as, 822, 826 in frontier zones, 188–​194, 191f, 226 gender identification through, 367–​371 of Hadrian's Wall, 234–​236 homosexuality represented in, 364–​364 hybrid forms of, 188 identity and, 206–​207, 226 in Iron Age, 10, 26, 155, 159, 167, 689 memory and, 682–​683 of military community, 451–​452 mobility and, 206–​207, 255–​258 multiculturalism and, 233–​234 in rural areas, 712, 714 selection processes for, 190 transformations in, 491–​492, 493–​494, 496 Materiality, 481–​519. See also Artefact studies; Material culture agency and, 484, 486–​487, 489, 497 in contemporary archaeological thought, 487–​489 continuity and, 495 globalization and, 484, 489 historical approaches to, 482–​484 imperialism and, 481, 482, 489, 492–​493 interdisciplinary studies of, 485, 487–​488 material horizons in Roman Britain, 489–​496 theoretical approaches to, 485–​487, 488–​ 489, 499–​500 Matthews, C. L., 384 Matthews, K., 364–​365 Mattingley, David, xxvii, 56, 102, 109, 228, 238, 449, 579, 624, 628, 689, 722, 745, 749, 851 Mauss, Marcel, 485 Maximian, xxxiii, xxxiv Mays, S., 311–​312 McKinley, Jacqueline, 435 McManus, D., 585 McWhirr, Alan, 27 Media coverage of Romano-​British archaeology, 51–​52 Medicinal herbs, 822 Medicine, 555–​572 anthropological studies of, 555, 556

bioarchaeology and, 567 collyrium stamps and, 560–​562 doctors and, 563, 564 herbs in, 822 historical overview, 556–​557 hospitals and, 563, 565–​566 instrumentation used in, 558–​560 in military community, 562–​567 Romanization of, 555–​556, 557 Vindolanda tablets and inscriptions regarding, 562–​564 vision remedies, 561, 562 Memoria Romana project (Galinsky), 684 Memory, 681–​696 archaeology and, 682–​684 collective, 681, 682, 691 history of, 681–​682 landscapes and, 683 limitations of, 688, 691 material culture and, 682–​683 overview, 681 prospective, 683, 688, 691 retrospective, 683, 687, 688, 691 in Roman Britain, 685–​691 in Roman society, 684–​685 visibility of, 691 Men. See Gender Menarche, 305, 307 Merrifield, R., 82, 440 Meskell, L., 233 Metabolic disease, 308 Metal detectors, 53–​54 Metals and metalworking, 532–​554 blacksmithing, 548, 549 brass, 532, 538, 541 bronze, 493, 532, 538, 540, 839, 844 copper and copper alloys, 532, 538–​542, 540–​541f, 548–​549 iron, 532, 543–​548, 544–​545f, 549 in Iron Age, 490 lead. See Lead as motivation for Roman conquest of Britain, 548 overview, 532 pewter, 532, 542 precious metals. See Gold; Silver in Scotland, 181

Index   885 tin, 532, 542, 548 in urban areas, 782, 783 zinc, 532, 538, 540, 541 Middle ground theory, 168 Migration. See Mobility Miles, David, 32 Military community, 448–​477 armour and weaponry of, 452, 464–​468 burial practices within, 457 Christianity in, 454 clothing of, 416–​417, 419–​420, 452 cult practices within, 453–​458 daily life in, 470–​471 epigraphic evidence for, 471 epigraphic habit in, 95, 237 extramural settlements associated with, 450–​451, 470, 471, 472–​473 on Hadrian's Wall, 227–​231, 229f, 230–​231t identity of, 417, 448–​449, 451–​459 inscriptions by, 14, 287, 288–​291, 295–​296, 628 literacy in, 104–​105, 471–​472 material culture of, 451–​452 medicine and hospitals within, 562–​567 multiculturalism in, 227–​231, 229f, 230–​231t, 458–​459 overview, 448–​449, 464–​465 personal belongings of, 468–​470 population estimates for, 464 in rural areas, 712 servants of, 450 sociolinguistics within, 578–​579 veterans in, 459–​460 women in, 473 Military districts, 14, 15f, 16, 17–​18 Miller, D., 233, 488 Millett, Martin, xxvii, 46, 22, 231, 343, 483–​484, 519, 699, 721, 722, 723, 724, 729, 730, 748, 755, 774 Milne, J., 557 Mionnet, Théodore Edme, 64 Mithras, cult of, 456 Mnemonics, 682 Mobility, 203–​223, 245–​258 archaeological evidence for, 204 cultural change and, 274–​275 diaspora theory and, 214–​215, 217

epigraphy in study of, 204–​206, 205f, 248–​ 249, 251–​255, 252–​253f gender and, 205, 252, 254, 256 identity and, 245–​247 immigrants, identification of, 215–​217 isotope analysis and, 204, 207–​214, 212f, 215–​217 in Late Antique period, 205–​206, 205f material culture and, 206–​207, 255–​258 models for interaction, 213–​214 overview, 203 scientific techniques in study of, 207–​213 Modification stage of funerary process, 434–​435 Modood, T., 225 Molleson, T. I., 308 Mommsen, Theodor, 13, 482 Money. See Coinage Mons Graupius, battle of (AD 83), xxxi, 184 Montgomery, J., 212 Moore, Alison, xxvii, 312, 321, 322, 363, 372 Moore, Tom, 28, 50, 154, 226–​227, 262 Moorhead, Sam, 834, 836 Moreland, J., 104, 112 Morris, J., 654 Mortaria, 107, 470, 517, 520 Mosaics Christian, 669–​670, 671 distribution of, 600 early interpretations of, 9 figurative imagery on, 105 function of, 610 subject matter for, 141, 142, 351 textual evidence from, 105, 106f on villa flooring, 29, 141 Moulding techniques, 539 Müldner, Gundula, 203 Mules, 799 Mullen, Alex, 251, 573 Multiculturalism. See also Ethnicity; Race in archaeological interpretations, 215–​216 definitions of, 225 Hadrian's Wall and, 216, 224–​225, 228–​231, 229f, 230–​231t, 240–​241 materiality of, 233–​234 in military community, 227–​231, 229f, 230–​ 231t, 458–​459

886

886   Index Nash-​Williams, V. E., 30, 77 National Planning Policy Framework (2012), 44, 51, 58 National Roman Pottery Fabric Reference Collection, 82 Neal, David, 37 Neolithic monuments, 686–​687 Nero, xxx, 610, 758 Nesbitt, Claire, 224 Networks, educational, 33, 34–​35t New Archaeology movement, 32, 382, 383, 485 Newborns. See Infancy Newstead site artefact studies from, 69, 73, 80, 457 geophysical surveys of, 186 ritual deposition at, 654, 655 twentieth-​century excavations of, 23 Niblett, R., 127, 429 Nitrogen isotope analysis, 212–​213, 308 Nora, Pierre, 682 Normanby hoard, 834, 841 North African populations diet, 821 at Hadrian's Wall, 234–​235, 459 pottery from, 234, 459 Nucleated settlements, 156, 266, 526, 708–​7 10, 709–​7 10f Numismatics. See Coinage Nutrition. See Diet and nutrition Occupational communities, 451 Offerings. See Ritual deposition; Votive altars Ogam inscriptions, 588–​590, 589f Older adulthood, 329–​331, 328t. See also Life course O'Neil, J., 76 Oppida (settlements) in Britain vs. Gaul, 266 construction of, 759 early Roman horizon in, 118, 121, 122, 128, 129 emergence of, 742 in Iron Age, 154, 156, 490 ownership of, 746 pottery assemblages in, 514, 515 Romanization of, 512 typology of, 156 villas as, 704

Oriental cults, 779 Orthopraxy, 635 O'Shea, J. M., 314 Osteoarthritis, 397 Osteological analysis, 306–​307, 323, 324, 345–​346, 427 Osteological paradox, 732 Oswald, F., 74 Owles, E., 549 Oxford Archaeology, 45, 45f, 49, 33 Oxford History of England, xxvii Oxygen isotope analysis, 207, 208, 210–​212, 212f Pader, E. J., 385 Paganism, relationship with Christianity, 675 Paintings. See Wall-​paintings Parker, A., 800 Parks, K., 809 Parsons, D. N., 592n4 Parting technique for goldworking, 535 PAS. See Portable Antiquities Scheme Patron–​client relationships, 158–​159, 161, 607–​608 Paul (apostle), 232 Peachin, M., 355 Peacock, David, 32, 80, 511, 525 Pearce, John, 325, 341, 368, 426–​427, 440 Pearson, Joseph, 684 Perring, D., 514 Personal appearance. See also Clothing burial items related to, 328, 328t, 351 cosmetics and, 411, 416 gender and, 369–​371 hairstyles and, 410, 413, 419, 431 identity and, 327, 406, 417–​19 toilet instruments and, 166, 264, 265, 369–370, 411, 512 Personhood, 312, 325–​326, 488 Pertinax, xxxii Petts, David, 660 Pewter, 532, 542 Pheasants, 801 Philpott, Robert, 36, 80, 425, 434, 438 Phosphoric iron, 543, 545 Photography, aerial, 54–​55, 30, 700, 703, 715, 716f

Index   887 Picts tribe, xxxiv, 186, 195, 196 Piddington excavations, 46, 46f Piece-​moulding techniques, 539 Pigs, 797–​798 Pine cones, 824, 824f Pitt Rivers, A. H. L. F., 17, 65, 71, 73 Pitts, Martin, 161, 514, 720, 735 Planning Policy Guidance 16: Archaeology and Planning (PPG 16), 43–​44, 45, 49, 50, 484 Planning Policy Statement 5: Planning and the Historic Environment (PPS 5), 44, 50–​51, 52 Plaster burials, 347, 430 Plautius, 119 Pliny, 65, 74, 561 Plouviez, J., 415 Politics and religion, 627 Polluter pays principle, 44 Polybius, 158 Portable Antiquities Scheme (PAS) coinage recorded by, 834, 835f, 836, 837 databases of, 54, 56, 82, 188, 451 metalwork findings recorded by, 54, 38, 82, 152 styli distributions recorded by, 579, 580f Postcolonial theory, 214, 216, 484, 485, 487, 500 Post-​processualism, 84, 383, 384, 385, 483 Posture of decapitated burials, 395f, 396–​397 Pottery, 510–​531 artefact studies of, 64, 71, 73–​74, 75, 77, 78–​79 as art object, 600 in burials, 349, 350f characterization and sourcing of, 32 chronological studies of, 27–​28 civilian traditions, 518–​521 in Claudio-​Neronian periods, 510–​513 for cremation containment, 159–​160 development of Romano-​British traditions, 513–​521, 525 as economic evidence, 77, 139–​140 Frisian, 235–​236, 459, 470 Gaulish, 73, 139, 140, 272, 470, 511 graffiti scratches on, 108, 592n6 grog-​tempered, 513, 514, 515, 516 inscriptions on, 101 Iron Age, 128, 155, 159–​160

late Roman period, 522–​525 marketing patterns and, 851 North African, 234, 459 overview, 510 regional types, 521–​522, 525–​527 Romanization of, 16, 17 stamps on, 101, 105–​106, 107 transformations in, 492, 494 in urban vs. rural areas, 514–​515, 518, 723–​ 724, 725–​731, 727f, 729f wheel-​thrown, 159–​160, 511, 515 The Pottery Kilns of Roman Britain (Swan), 47 Poundbury cemetery age-​related patterns at, 326, 327f child burials at, 308, 309, 323, 325 coffins in, 348 demarcation of burials at, 354 gender patterns at, 326, 330, 372 material culture at, 329, 370 spatial organization of, 433, 672 wall paintings at, 409 Powell, L. A., 308–​309 Power economics and, 850, 856, 860, 862 wealth and, 157–​162, 856–​857 Powleslend, Dominic, 701 PPG 16. See Planning Policy Guidance 16: Archaeology and Planning PPS 5. See Planning Policy Statement 5: Planning and the Historic Environment Precious metals. See Gold; Silver Pre-​interment rituals, 345–​347 Preparation stage of funerary process, 428–​431 Preservation by record, 44, 31 Price, J., 80 Priorities for the Preservation and Excavation of Romano-​British Sites (Roman Society), 47 Processualism, 84, 383, 385, 483, 486, 488–​489 Professionalization of archaeology, 44, 48 Prone burials, 384, 387, 388, 395f, 396–​397, 428 Propertius, 410 Prospective memories, 683, 688, 691 Prowse, T. L., 209 Ptolemy, 110, 180–​181, 226, 747 Puberty, 305

888

888   Index Public engagement with archaeology, 51–​33, 53f Puck of Pook's Hill (Kipling), 13 Purbeck marble plaque, 626 Puttock, S., 645 Querns, 155, 470, 712, 715 Rabbits, 800 RAC (Roman Archaeology Conference), 47, 57 Race. See also Ethnicity; Multiculturalism controversy in identification of, 215–​216 mixed-​race individuals, 215, 240 Radiate coinage, 841–​843 Rahtz, Philip, 44 Ralph, S., 161 Reciprocity, in economic systems, 139 Redfern, R. C., 307, 308, 567 Reece, Richard, xxvii, 32, 33, 80, 85, 111, 276, 483, 836, 840–​841 Regina (Catuvellaunian freedwoman), 238–​ 240, 239f, 298, 369 Religion. See also Christianity; Deities; Ritual deposition; Temples in art, 610, 613 in Britain, 11, 143, 387 clothing and, 418 cult practices, 453–​458, 778, 779 in forum, 772 in Gaul, 143 politics and, 627 social renegotiation through, 688 theories of transmission and adaptation, 619 worship spaces in urban areas, 772–​779, 775f, 777f Remote-​sensing techniques, 55 Renfrew, Colin, 487 Rescue archaeology, 44, 47, 50, 26, 27, 31 Res Gestae (Augustus), 152 Retrospective memories, 683, 687, 688, 691 Revell, Louise, xxvii, 321–​322, 326, 328, 330, 332, 367, 371, 372, 373, 635, 767 Rice, 822 Richborough site artefact studies from, 65, 71 baptismal font at, 670 church structures at, 668

coin assemblages from, 844 dating evidence at, 127 textual evidence from, 106 Richmond, Ian, 28, 79, 483 Rickets, 308, 309, 397 Rigby, V., 511 Rigor mortis, 429 Rites of passage, 305, 322, 326 Ritual deposition, 641–​659 of animals, 646, 650, 651, 653–​654 of coinage, 161–​162, 644–​645, 647, 648, 839 offering zones for, 646–​649 overview, 641–​643 in settlement contexts, 651–​654, 653 special deposits, 642–​643, 650, 653–​654, 653 structured, 649–​651 in temple settings, 642, 643–​644, 649–​651 votive offerings within cult sites, 644–​646, 645t in water, 160, 642, 671–​672 Rivet, A. L. F., 49, 31, 723 Roach-​Smith, Charles, 64–​65, 73 Robinson, H. Russell, 467 Robinson, M., 813 Rogers, Adam, 741 Roman Archaeology Conference (RAC), 47, 57 Roman Britain animals in. See Animals archaeological study of. See Romano-​ British archaeology artefact studies of. See Artefact studies art in. See Art building types in, 142, 144 burials in. See Burials Celtic subaltern image of, 4–​5, 8–​10, 13, 18 childhood in. See Childhood and adolescence civil and military districts of, 14, 15f, 16, 17–​18 civilizing influence of Romans in, 4, 5–​7, 8, 10–​11 clothing in. See Clothing coinage in. See Coinage continental influences in, 268–​274, 269f, 271f contrasting views of, 10–​11, 13 defensive systems in, 135–​136

Index   889 diet and foodways in, 272–​274 division into provinces, xxxii early studies of, 3–​10 economic systems of. See Economics elite auto-​representation in, 142, 143, 144 emergence of, 489–​492. See also Early Roman horizon end of, 858–​860 farming in. See Farming fragmentation of, 494–​496, 498 framework for study of, xxvii, xxviii, 25 gender in. See Gender Germanic peoples in, 144 independence of, xxxv interactions with Gaul and Germany, 262–​ 272, 265f, 269f, 273–​278 language in. See Sociolinguistics life course in. See Life course material culture of. See Material culture; Materiality medicine in. See Medicine memory in, 685–​691 metalworking in. See Metals and metalworking military in. See Military community mobility in. See Mobility political evolution of, xxix population estimates, 700, 852–​853 religion in. See Religion rural areas of. See Rural areas textual evidence from. See Textual evidence timeline of, xxviii–​xxxvi urban areas of. See Urban areas withdrawal of imperial presence from, 135–​138, 144 Roman Britain and the English Settlements (Collingwood), 24, 25 Roman Crafts (Strong & Brown), 80 Roman Empire expansion of, 151, 152, 153, 232 indigenous lands incorporated into, 117 literacy in, 102 multiculturalism of, 225, 247–​248 unification of populations across, 5, 213 The Roman Era in Britain (Ward), 73 Roman Finds Group, 81 Roman Grey Literature Project, 50

The Roman Inscriptions of Britain (Collingwood & Wright), 37, 95, 100–​101, 105–​107, 561–​562, 609 Roman Iron Age, 180 Romanization age construction and, 324 citizenship and, 232–​233, 452 of coinage, 16, 163–​164, 510 conceptualizations of, 231–​232 emergence of, 482 inadequacy in explanation of cultural transformations, 262 of indigenous populations, 4, 13–​14, 17 of medicine, 555–​556, 557 of pottery, 16, 17 of rural areas, 723 in urban areas, 14, 16, 492–​493, 747–​749 The Romanization of Britain (Millett), 483, 748 The Romanization of Roman Britain (Haverfield), 4, 23, 482, 747–​748 Romano-​British archaeology, 43–​64 in 1940s–​1960s, 26–​30 in 1970s and 1980s, 30–​33, 36 in 1990s and beyond, 36–​38 approaches and techniques in, 54–​55, 55f, 38 developer-​funded, 48–​50 deviant burials in, 388–​389, 400 future challenges for, 57–​58 history and evolution of, 44–​45, 45f, 56–​57 in interwar years, 24–​26 metal detectors in, 53–​54 non-​publication of, 50–​51, 37 overview, 43 public engagement with, 51–​53, 53f research frameworks, 46–​48 specialist groups in, 48 twentieth-​century foundations of, 22–​24 Romano-​Celtic temples baptismal fonts at, 670 design of, 270–​272 distribution of, 271, 271f, 277 reuse of ancient monuments by, 689 ritual deposition at, 690 spatial configuration and patterning of, 691 in urban areas, 773–​774, 776–​778 The Roman Wall in Scotland (MacDonald), 23 Rooke, Hayman, 9

890

890   Index Ross, Ann, 29 Roundhouses, 18, 180, 190, 267, 496–​497, 706, 759 Rudling, D., 652 Rummymede Trust, 216 Rural areas, 699–​740 burials in, 343, 353–​354, 431–​432, 735–​736, 736f characteristics of, 700–​703, 701f diet in, 734 dynamic understanding of, 710–​7 17 economic relationship with urban areas, 720–​722, 723–​725, 729–​731 farmstead diversity in, 706, 708 health conditions in, 733–​734, 733f imperialism impacting, 719, 732, 736 inequality in, 732–​736, 736f, 737 in Iron Age, 49 material culture in, 712, 714 military community in, 712 nucleated settlements in, 708–​7 10, 709–​7 10f overview, 699–​700 population estimates for, 700 pottery in, 514–​515, 518, 723–​724, 726–​731, 727f, 729f regional variations in, 699–​700 Romanization of, 723 small worlds concept applied to, 854–​855 transformation of, 721, 722–​724, 737 of urban hinterlands, 725–​731, 727f, 728t, 729f villas in, 703–​706, 705f, 707f, 723 Rural Settlement in Roman Britain (Hingley), 30, 484 Russell, P., 251, 592n3 St Albans site cemetery at, 343 coinage from, 844 inscriptions at, 287 organization of excavations at, 27 pottery from, 730 wealth at, 160 St Germanus, xxxvi St Joseph, J. K., 54, 30 Saller, R., 292 Salway, Peter, xxvii, 30, 387–​388, 465

Sandwich Islands, 150 Saxe, A., 382, 383–​384, 388 Saxon Shore site, 137–​138 Scarth, Henry Mengden, 11 Scheidel, W., 209 Schiffer, Michael, 486 Schrijver, P., 591 Schwartz, B., 683 Schweissing, M. M., 209 Scotland coinage in, 839 diplomacy and interference in, 194–​195 frontier life in, 182, 185–​189, 187f, 194 historic preservation responsibilities in, 44 Iron Age in, 179–​182 long-​term impact of Roman influence on, 195–​196 material culture in, 188–​194, 191f, 226 metalworking in, 181 military community in, 469 military history of, 182–​186, 183f research framework in, 47 Roman conquest of, xxxi, xxxiii, xxxiv, 181–​185, 183f settlement patterns in, 180–​181 twentieth-​century excavations in, 23 Scott, Eleanor, 365, 366, 369, 483 Sculpture destruction of, 683 distribution of, 600–​602, 601f, 614 funerary, 353 military-​related, 416, 606 producers of, 607–​609 quality classifications for, 604–​606, 605f religious use of, 610 as status display medium, 141 subject matter for, 141, 606 textual evidence from, 292 uses and users of, 610–​611 in villas, 139, 142, 610–​611 Scurvy, 308 Sears, G., 767–​768 Selection stage of funerary process, 426–​428 Sen, Amartya, 448 Septic arthropathy, 397 Septimius Severus, xxxii–​xxxiii, 185, 216, 234, 344, 434, 743

Index   891 Sewing kits, 468 Sex determination of skeletal remains, 306, 327, 368 SFBs (sunken-​featured buildings), 144, 145 SGRP (Study Group for Roman Pottery), 48, 79 Sharples, Niall, 154, 689, 690, 759 Shaw, B. D., 292, 305 Shay, T., 383, 385 Sheep, 650, 795–​796 Sheepen settlement, 128 Sherratt, Melanie, 363 Shiel, N., 592n5 Shiptonthorpe settlement, 369, 712, 713, 713f, 715 Shoes. See Footwear Silbury Hill site, 687 Silchester site baptismal font at, 671 bathing complex at, 781 cattle at, 794 Christian symbols at, 663 church structures at, 665, 667f construction of, 755–​756, 757f cupellation at, 537 foods at, 819, 824 forum at, 768, 769, 770–​771, 771f, 772 housing at, 784 inscriptions at, 287 pigs at, 797, 798 public buildings at, 785 report from, 65 ritual deposition at, 652–​653, 655, 687 sustained occupation of, 14 temples at, 776, 777–​778 theatre at, 780 twentieth-​century excavations of, 23 university excavations at, 46 urban features of, 7, 266, 492 Silver coinage produced from, 137, 164, 193, 493, 537, 541, 541f hoards of, 664 sites for extraction of, 535, 536f techniques for working with, 535, 537, 538 uses of, 535, 537 Simonides of Ceos, 682

Sims-​Williams, P., 592n10 Site reports, 65–​69, 66–​68t, 73–​74 Skeleton Green site, 126 Small finds, origins of term, 74–​75 Small worlds concept, 854–​855 Smiles, S., 152 Smith, Alex, 641 Smith, C., 585, 592n7 Smith, David, 29, 31 Smith, K., 654 Snead, J., 854 Social age, 320, 324 Social mapping, 453 Social stratification, 128, 855–​856, 857, 862 Society of Antiquaries founding of, 63 meetings of, 8 research reports of, 74, 75 Silchester excavations by, 14, 23 subject catalogue, 64, 72 women in, 77 Sociolinguistics, 573–​598. See also Latin language of Celtic languages, 575–​578, 580–​584, 581f evidentiary sources for, 574 Latinization and extent of bilingualism, 576–​584, 588, 591 loanwords and, 577–​578, 584–​585, 587, 593n12 in military community, 578–​579 nature of British Latin, 584–​588, 592n8 of Ogam stones, 588–​590, 589f overview, 573–​574 post-​Roman inscriptions and Germanic languages, 588–​591, 591f in pre-​Roman Britain, 575–​576 Soden Smith, Robert, 72 Soldiers. See Military community Spector, J., 363 Spelt grains, 808, 809–​810, 811–​812f, 812–​813 Springhead temple complex, 162, 325, 348, 647, 651 Stair, John, 7 Stamps collyrium stamps, 560–​562 on pottery, 101, 105–​106, 107 Stanford University, 51

892

892   Index Stanway burial site continuity in, 128, 129 creation of, 344 evidence of literacy at, 166 individualized nature of, 267, 268 laying-​out stage at, 429 material culture at, 350 medical instrumentation discovered at, 559–​560 pottery from, 730 warrior burial, 264 wealth at, 160, 165 Statistical analysis on demography, 292–​293 Status and burial, 309–​310, 314, 341–​345, 354–​355 Steinberg, S. R., 225 Stewart, P., 621, 622 Stilicho, xxxv Stinking mayweed, 813 Stolpersteine plaques, 684, 686 Stonehenge, 6, 8, 47 Strabo, 152, 163, 267, 542, 837–​838 Strathern, Marilyn, 488 Strontium isotope analysis, 207, 208, 209–​210, 212f Structural Marxism, 154 Study Group for Roman Pottery (SGRP), 48, 79 Stukeley, William, 6–​7, 9, 11, 18 Stylus tablets, 96, 98–​99, 98f Sumner, G., 419–​420 Sunken-​featured buildings (SFBs), 144, 145 Superficial homogenization, 248 Supine burials, 395f, 396, 437 Surgical instrumentation, 558–​559 Surgical Instruments in Greek and Roman Times (Milne), 557 Swan, Vivien, 47, 234, 235, 274, 459, 470, 724 Swift, Ellen, 63, 207, 417, 420 Symbols in Action (Hodder), 82 Symmetrical archaeology, 487, 488 Tacitus, 4, 8, 9, 110, 111, 185, 224, 268, 459–​460, 548, 575, 747, 758, 839 Taxation, 138, 227, 524, 721–​722, 855, 856 Taylor, Jeremy, 35, 37, 687, 700, 708, 717 Technological determinism, 525 Temples. See also Romano-​Celtic temples

ritual deposition in, 642, 643–​644, 649–​651 in urban areas, 773–​779, 775f, 777f Temples, Churches and Religion in Roman Britain (Rodwell), 644 Tertullian, 454, 661 Textile production, 407–​408. See also Clothing Textual evidence, 95–​116. See also Inscriptions; Literacy ancient uses and responses to writing, 102–108, 103f, 106f of Christianity in Britain, 661–​663, 674 classification as artefacts, 104 curse tablets. See Curse tablets evolution of study of, 112 overview, 95–​96 relationship with archaeological evidence, 110–​112, 118 stylus tablets, 95, 98–​99, 98f types of, 96–​101, 97–​99f, 101f of urban development, 746–​747 Thalassemia, 309 Theatres, 773–​774, 776, 778, 779–​780, 782 Theodosian Code, 855 Theodosius, xxxv, 660–​661 Theoretical Roman Archaeology Conference (TRAC) artefact studies and, 81 diversity of approaches represented by, 852 establishment of, xxvii, xxviii, 483 on gender, 363, 366 on identity, 406 influence of, 35, 36–​37 internationalization of, 57 on materiality, 484, 488 Thomas, J. D., 96 Tilley, C., 488 Timby, J. R., 515 Tin, 532, 542, 548 Titus, 184 Togas, 409, 418 Togidubnus, 758 Toilet instruments, 166, 264, 267, 369–​370, 411, 512 Tollgate grave assemblages, 347, 349, 350f, 351 Tombstones. See also Burials expressions of identity through, 237–​240, 239f inscriptions on, 237–​240, 239f, 289, 296–​298, 439

Index   893 Latin on, 439 stock images for, 608 on Via Appia route, 685 Tomlin, R. S. O., 99, 108, 580, 582, 635 Tonnochy, A. B., 77 Tooth enamel health stress and, 307 isotope analysis from, 308, 309, 315 Toponyms, 592n4 Town and Country in Roman Britain (Rivet), 29 Towns. See Urban areas The Towns of Roman Britain (Wacher), 748 Toynbee, Jocelyn, 29, 76, 348, 602–​603 TRAC. See Theoretical Roman Archaeology Conference Trade diasporas, 213, 216 Trajan, 292, 293, 467 Traprain Law site, 179, 192, 193, 664, 845 Treggiari, S., 371 Tube dresses, 409, 419 Tunics, 409, 416 Turner, E. G., 96 Tyers, Paul, 37, 64, 85, 519 Tylecote, R. F., 549 Uley temple curse tablets at, 99f, 100, 108–​109, 572, 582, 628 goats at, 795 images at, 607, 613 infant burials at, 326 ritual deposition at, 644, 649 United Kingdom (UK). See also England; Scotland; Wales archaeology in, 43–​44 higher education in, 32–​33 University of Birmingham, 27, 46 University of Cambridge, 51 University of Newcastle, 45, 51 University of Reading, 46, 50, 203, 699 University study of archaeology, 32–​33, 36 Urban areas, 741–​790 bathing in, 780–​782 burials in, 343, 353–​354, 431–​432, 735–​736, 736f categorization of, 742–​746 construction of, 752–​760, 753–​754f, 757f, 767

diet in, 734 in early Roman horizon, 125–​126 economic relationship with rural areas, 720–​722, 723–​724, 729–​731 entertainment and spectacles in, 779–​780, 782 forum as central feature of, 768–​773, 770–​771f health conditions in, 733–​734, 733f hinterlands of, 725–​731, 727f, 728t, 729f housing in, 782, 783–​784 industrial activity in, 782–​783 legal hierarchy of, 742–​744 location of, 749–​752 major settlements in, 744, 744f overview, 741–​742 political and administrative spaces in, 768–​ 773, 770–​771f pottery in, 514–​515, 518, 723–​724, 726–​731, 727f, 729f regional variations in, 784–​786 Romanization in, 14, 16, 492–​493, 747–​749 size of, 745, 784 textual evidence of urban development, 746–​747 worship spaces in, 773–​779, 775f, 777f Vale of Pickering, geophysical survey of, 55, 55f, 701–​703, 702f Vale of York, 515, 711 Vance, Norman, 4 Van der Veen, Marijke, 807 Van Driel-​Murray, Carol, 235, 365–​366, 450 Verulamium site bathing complex at, 781 continuity at, 127 cult practice at, 778 in early Roman horizon, 121–​122 forum at, 769, 772 industrial activity at, 165, 783 material culture in, 130–​131 mortuary complex at, 341, 758 organization of excavations at, 25, 27 pottery from, 513–​514, 519 public buildings at, 123–​124, 755, 784–​785 sustained occupation of, 14 temples at, 773–​774, 776, 778 theatre at, 780

894

894   Index Vespasian, xxix, 184 Veterans, 459–​460 Via Appia, tombstones lining route of, 685 Vicarious human sacrifice, 384 Victim diasporas, 213 Victor (North African freedman), 235, 237–​238 Victoria County Histories, 23 Villas architecture of, 10, 10f, 141–​142, 268–​269, 269f, 493, 857, 857f baptismal fonts at, 671 churches in, 668–​669 mosaic flooring in, 29, 141 regional distribution of, 704, 858–​859, 859f in rural areas, 703–​705, 705f, 707f, 723 sculpture in, 141, 142, 610–​611 Vindolanda site churches at, 454 church structures at, 668 clothing at, 406–​407, 408 cult practices at, 455 extramural settlements at, 450 medical-​related tablets and inscriptions at, 562–​564 sociolinguistic evidence from, 574, 578 textual evidence from, 37, 37, 96–​98, 97f, 104–​105, 471–​472, 839 volunteers at, 51, 53, 53f Vineyards, 817 Vision remedies, 561, 562 Vitamin C deficiency, 308 Vitamin D deficiency, 308, 309 Viticulture, 817 Volunteer archaeology groups, 51, 51, 27 Votive altars distribution of, 622–​624, 623–​624f, 630, 631f female construction of, 632 inscriptions on, 292, 294, 297, 298, 622, 628–​629 offerings within cult sites, 644–​645, 645t Vows pledged to deities, 626, 627 Wacher, John, 44, 27, 31, 721, 741, 746, 748, 768 Wales historic preservation responsibilities in, 44 pottery in, 513, 514–​515, 522–​523, 525–​526 pre-​Norman inscriptions in, 592n10

research framework in, 47 Roman conquest of, xxx, xxxi Wallace, Lacey, 117, 784 Wallerstein, Immanuel, 721 Wall-​paintings Christian, 143, 661, 668–​669, 669f clothing styles in, 409 distribution of, 600 function of, 610 textual evidence from, 105 Walton, Philippa, 834, 842 Ward, John, 73 Warner, Richard, 64 Warton, Thomas, 3, 10–​11 Water Newton hoard, 143, 663, 664, 675 Way, Albert, 72 Wealth and power, 157–​162, 856–​857 Weaponry, 465–​468 Webster, Graham, 27, 30, 77, 79 Webster, J., 558, 621, 633, 634 Weekes, Jake, 423, 440 Weiss-​Krejci, E., 388 Welch, K., 780 Wells, P., 272 Welwyn burial site, 151, 164, 165, 264, 265f, 349 Wheeler, Mortimer, xxvii, 52, 57, 25, 75 Wheeler, Tessa Verney, 25, 75 Wheel-​thrown pottery, 159–​160, 511, 515 White, Richard, 166, 725 Whittaker, C. R., 627, 629 Wild, J. P., 409, 419 Wild boar, 800 Wilkes, John, 33 William, John, 11 Williams, Howard, 351, 683, 686, 689 Williams, J. H. C., 107, 166, 168, 574 Wilmott, T., 779 Wilson, David, 54 Wilson, Peter, 43 Winbolt, S. E., 74 Winchester site cattle at, 794, 795 cupellation at, 537 gold artefacts at, 161, 413 isotope analysis of, 211, 212f occupation of, 157 pigs at, 797, 798

Index   895 research framework for, 27 sheep at, 796, 797 Windle, Bertram, 11, 13 Witcher, R., 225 Women. See also Gender artefact studies by, 71, 77, 78f, 80–​81, 82–​83 breastfeeding by, 308–​309 gardening by, 826 hairstyles of, 410, 413, 419, 431 in military community, 473 mobility of, 205, 252, 254, 256 social roles of, 329, 331 in Society of Antiquaries, 77 in textile production, 408 visibility of, 413, 415 votive altars constructed by, 632 Woodfield, C., 524 Woolf, G. D., 107, 152, 155, 166, 231–​232, 635, 785 World-​systems theory, 689, 721 Wright, R. P., 100 Wright, Thomas, 11 Writing on the Wall project (2006), 216 Written evidence. See Inscriptions; Textual evidence

Wroxeter site bathing complex at, 781, 782 cattle at, 795 coinage from, 844 construction of, 754 cupellation at, 537 forum at, 771 hinterlands project, 725 inscriptions at, 287 pigs at, 797 public buildings at, 785 remote-​sensing techniques for, 55 temples at, 775, 777f twentieth-​century excavations of, 23 university excavations at, 46, 27 urban features of, 492 Yalden, D., 800 Young adulthood, 328–​329, 328t. See also Life course Zinc, 532, 538, 540, 540–​541f, 541 Zoll, Amy, 619 Zooarchaeology, 791–​802. See also Animals Zosimus, 861

896

898

900

902

904

906

908