Archaeological Survey in the Lower Liri Valley, Central Italy: under the direction of Edith Mary Wightman 9780860547693, 9781407348971

Table of contents :
Front Cover
Copyright
PREFACE AND ACKNOWLEDGEMENTS
TABLE OF CONTENTS
LIST OF FIGURES
LIST OF CONTRIBUTORS
Chapter 1: INTRODUCTION
Chapter 2: THE GENESIS OF AN ENVIRONMENT: ELEMENTS OF GEOLOGY, GEOMORPHOLOGY AND LAND USE
Chapter 3: PREHISTORIC SURVEY IN THE LOWER LIRI VALLEY
Chapter 4: BRONZE, IRON AGES AND ETRUSCANS
Chapter 5: THE LIRI VALLEY: LANDSCAPE AND PEOPLE IN ROMAN LITERATURE
Chapter 6: THE ROMANIZATION OF THE VALLEY
Chapter 7: COMMUNICATIONS
Chapter 8: SETTLEMENT PATTERNS AND SOCIETY
Chapter 9: SOCIETY, ECONOMY AND THE ENVIRONMENT
Chapter 10: THE END OF TRANQUILLITY: THE TRANSITION FROM ROMAN TO MEDIEVAL CONDITIONS
Chapter 11: THE LIRI VALLEY IN THE MIDDLE AGES
Chapter 12: SYNTHESIS
Figures
APPENDICES
BIBLIOGRAPHY
LOCATIONS

Citation preview

Archaeological Survey in the Lower Liri Valley, Central Italy under the direction of Edith Mary Wightman

J. W.

Edited by Hayes and I.

P. Martini

contributing authors R. G. V. Hancock, J. W. Hayes, P. Hemphill, M. E. Lewis, G. A. Lond, I. P. Martini, A. G. McKay, A. McPherron, E. M. Wightman, D. F. Williams

BAR International Series 595 1994

Published in 2019 by BAR Publishing, Oxford BAR International Series 595 Archaeological Survey in the Lower Liri Valley, Central Italy © The editors and contributors severally and the Publisher 1994 The authors’ moral rights under the 1988 UK Copyright, Designs and Patents Act are hereby expressly asserted. All rights reserved. No part of this work may be copied, reproduced, stored, sold, distributed, scanned, saved in any form of digital format or transmitted in any form digitally, without the written permission of the Publisher. ISBN 9780860547693 paperback ISBN 9781407348971 e-book DOI https://doi.org/10.30861/9780860547693 A catalogue record for this book is available from the British Library This book is available at www.barpublishing.com BAR Publishing is the trading name of British Archaeological Reports (Oxford) Ltd. British Archaeological Reports was first incorporated in 1974 to publish the BAR Series, International and British. In 1992 Hadrian Books Ltd became part of the BAR group. This volume was originally published by Tempvs Reparatvm in conjunction with British Archaeological Reports (Oxford) Ltd / Hadrian Books Ltd, the Series principal publisher, in 1994. This present volume is published by BAR Publishing, 2019.

BAR

PUBLISHING BAR titles are available from:

E MAIL P HONE F AX

BAR Publishing 122 Banbury Rd, Oxford, OX2 7BP, UK [email protected] +44 (0)1865 310431 +44 (0)1865 316916 www.barpublishing.com

PREFACE AND ACKNOWLEDGEMENTS This volume must stand, belatedly, as a tribute to the academic prowess of Edith Mary Wightman, who directed the Liri Project during all its field seasons from 1978 to 1983. Her premature and violent death in December 1983 robbed the project of its organizer, and resulted in the present delayed and somewhat truncated report. We hope that what follows will be seen as worthy of its initiator. The text had essentially been written by the contributing authors at the time of Professor Wightman's death. The texts of chapters 6 to 10 by E.M. Wightman remain essentially as written; Hayes is responsible for the present form of the notes. Serious delays were encountered in the transfer of computer records (held at McMaster University) and in obtaining the final survey maps. A consequence of the delay in publication of the volume is that bibliography cited is not fully up-to-date. In particular, a discussion of the parallel French work, published in G. Chouquer et al., Structures agraires en ltalie centro-meridionale: cadastres et paysage ruraux (Collection de l'Ecole Fran~aise de Rome 100; Rome 1987--specifically pp. 124-30, 263-67: G. Chouquer, F. Favory), is lacking; the reader will need to balance its conclusions against the primary evidence presented here. In view of the wide range of academic disciplines involved, each with its own publication conventions, we have eschewed a totally uniform style in this report for such things as orthography and footnote references: each author's preferred style, appropriate to her/his own discipline, is respected.

Acknowledgements Initial fieldwork (in 1977) was sponsored by McMaster University, Hamilton. During its main fieldwork phase (1978-83), the project was chiefly financed by grants from the (then) Social Sciences and Humanities Research Council of Canada, the funds being administered by the Research Office of McMaster University; both organizations are here thanked for their support. Additional funds for the geological and geomorphological research were provided by the National Engineering and Research Council to I.P. Martini. Both then and subsequently, the authors' universities and institutions have provided supplementary financing and/or technical and secretarial help. A substantial part of the cost of publication has been provided by the Edith Wightman Memorial Fund; we here, belatedly, thank all contributors to the Fund for their generosity almost ten years ago, and hope that the present volume matches their expectations. Individual thanks go to our colleagues in the project (most of whose names appear here as co-authors), and to the many students and others who contributed to the field-walking and the processing of information and finds, both in the field and subsequently in Canada and elsewhere. All the survey maps (apart from the plan of Aquinum) were expertly prepared for publication by Ms. P. Cutts (Washington, DC). On the Italian side, prof. Angelo Nicosia (Pontecorvo) was our chief collaborator; without his local knowledge and contacts, our project would have been much the poorer. We thank him for his invaluable participation in the field work, and for his readiness to share so much information. The Soprintendenza Archeologica per ii Lazio Meridionale, and its various officers, provided permits and storage space for the finds in the Cassino Museum, and generally facilitated our work. The local authorities and religious bodies in Pontecorvo, Roccasecca and S. Giorgio a Liri provided accommodation for the team during its field seasons; local maps and other services were supplied by these and others. Logistical support for fieldwork was provided by the Canadian Academic Centre in Italy, and the British School at Rome. Mr. J. Delmege kindly offered accommodation at Fiuggi Fonte during the preliminary fieldwork in 1977. Assistance with documentation was received from profs. L. Quilici, F. Castagnoli, and A. Giannetti, and (for the medieval documents) the archivists of the abbey of Montecassino Don A. Pantoni, Don T. Leccisotti, and Don . F. Avagliano. Finally, our thanks go to the numerous local informants throughout the study zone who offered the team members information and generous hospitality during the project; their interest in our work is remembered with pleasure. We thank L. Neuru for storing some of Liri Project files during the later editing stages and for searching for some elusive notes. Mention should be made here of two other initiatives at scientific sampling which proved unsuccessful. 1. A sample of wood from a possible mill installation in the bed of the Liri just downstream of the bridge at Pontecorvo was submitted for dendrochronological purposes. The wood proved to be of a softwood variety, providing insufficient rings for matching. 2. Core-sampling of sediments was attempted in the dried-up former lake to the E of the city walls of Aquinum, with the help of the Lerici Foundation (Dr. R. P. Linington), who provided equipment and personnel. Heavy washes of gravel derived from the west bank were encountered, preventing drilling to a sufficient depth to penetrate potentially datable ancient layers. We thank all those involved in these undertakings. We apologize to those contributors whose texts may have been adversely affected by the long delay in publication. J. W. Hayes, P. Martini, 1993

TABLE OF CONTENTS CHAPTER

1.

1

INTRODUCTION .................................................................................... Objectives and philosophy of the study ......................................................... E.M. Wightman Methods of study I..andscape .................................................................................. I.P. Martini

CHAPTER 2.

CHAPTER

3.

CHAPTER 4.

.

.

Archaeology ............................................................................... E.M. Wightman, P. Hemphill and J.W. Hayes

.

Statistical analyses of archaeological sites ............................................ E.M. Wightman, I.P. Martini and M.E. Lewis

.

THE GENESIS OF AN ENVIRONMENT: ELEMENTS OF GEOLOGY, GEOMORPHOLOGY AND LAND USE, ....................................................... I. P. Martini

. 5

PREHISTORIC SUR VEY IN THE LOWER LIRI VALLEY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 A. McPherron BRONZE, IRON AGES AND ETRUSCANS

.................................................

The Bronze Age ....................................................................................... E.M.Wightman

13 13

The Iron Age . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 E.M. Wightman The Etiuscans in the Liri Valley. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 P. Hemphill CHAPTER 5.

THE LIRI VALLEY: LANDSCAPE AND PEOPLE IN ROMAN LITERATURE ...... : 18 A.G. McKay

CHAPTER 6.

THE ROMANIZATION OF THE VALLEY .................................................... E. M. Wightman

CHAPTER

7.

COMM UNI CA TIO NS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 E.M. Wightman

CHAPTER

8.

SETTLEMENT PATTERNS AND SOCIETY .......................... E.M. Wightman and J.W. Hayes

CHAPTER 9.

. ..........

26

. .. .. . . .... . ... J.t

SOCIETY, ECONOMY AND THE ENVIRONMENT ...... .. ...... .. .... ........ ....... .. . .. . 41 E.M. Wightman and J.W. Hayes

V

CHAPTER 10. THE END OF TRANQUILLITY: THE TRANSITION FROM ROMAN TO MEDIEVAL CONDITIONS ............................................ ... ........ ................ E. M. Wightman CHAPTER 11. THE LIRI VALLEY IN THE MIDDLE AGES................... G.A. Loud

CHAPTER 12.

48

...........................

..... 53

SYNTHESIS OF PROJECT AND CONCLUSIONS .......................................... J.W. Hayes and LP. Martini

69

F1GURES,...............................

...................................

..... .......... ....... .... ..... ················

74

A}?PENDICES APPENDIX A.

POTTERY FIN'DS.................................... J. W. Hayes

· · · · · · ·· · · · · · · · · · · · · · · · · · ·· · · · · · · · ·· · ·· · ·· · · ·· · 127

APPENDIX B. PETROGRAPHIC STUDIES OF POTTERY AND TILE .................................... D.F. Williams

159

APPENDIX C. NEUTRON ACTIVATION ANALYSES OF POTTERY ... . ................ R.G.V. Hancock

164

... .............

APPENDIX D. AQUINUM SURVEY ............................................................................... J. W. Hayes and E.M. Wightman

171

APPENDIX E.

173

CATALOGUE OF SITES ..........................................................................

BIBLIOGRAPIIY . .. ..... .. .......... .... ..........................................

. ................

LOCATION INDEX ......................................................................................................

vi

. ..... .... .......... .-.237 240

LIST OF FIGURES 1.

A. Tractor driven plough; B. Deep ploughing of land ...................................................... 74 Phvsiographic map of Italy with location of Liri Valley .................................................... 75 Topographic map of Lower Liri Valley (scale £,.1:200 000) ................................................... 76 3. Map of Liri Valley region with locations of ancient towns ( after Salmon, 1967) ............................. 77 4. 5. A. Photo of Liri valley with M. Cairo in background; B. Bank of the Liri River ..................... 78 6. A. Archaeological team in dissected area of Lower Liri Valley; B. Freshly ploughed site near the Parito Villa .............................................................................................. 79 Quaternary geological map of the Lower Liri River valley ................................................ 80 7. A. Lacustrine sediments; B. Air-photograph of terrace and low-lying floodplain at the confluence 8. of the Gari and Liri Rivers (Giuntura) ......................................................................... 81 9. Map ofl,iri Valley with locations of cross-sections shown on Fig. 12 ....................................... 82 10. Longitudma1 sections of terraces ................................................................................ 83 A. Scarps separating first river terrace from the higher bottom valley main lacustrine plain 11 upstream from Pontecorvo. B. Gentler scarp separating last terrace from modem narrow floodplain ............................................................................................................ 84 12. Topographic sections across valley ............................................................................. 85 13. A. Southern flank of M. Cairo. B. Alluvial fan of Parito Villa ......................................... 86 14. Section across the valley showing major morphological features, Pleistocene sediments and surficial soils ........................................................................................................ 87 15. Modem bridge and partially destroyed retention walls in the Melfa River, just upstream from the river crossing at S. Vito .............................................................................. 88 16. Left pier of old bridge at the Roman crossing of S. Vito over the River Melfa ....................... 89 17. Drawings of lithics from Santa Scolastica area ............................................................... 91 Drawing of lithics (not Santa Scolastica) ...................................................................... 93 18. S. Scolastica site at foot of the steep slope of Montecassino1• ................................................. 94 19. Liri Valley Survey - Index Map of areas surveyed .......................................................... 95 20. 21. Liri Valley Survey - Zone 1, Index map of sites ............................................................ 96 22. Liri Valley Survey - Zone 1, Roman Republic sites ........................................................ 97 23. Liri Valley Survey - Zone 1, Early Roman Empire sites ................................................. 98 24. Liri Valley Survey-Zone 1: A. Mid/Late Roman Empire sites; B. Medieval sites ........................ 99 25. Liri Valley Survey - Zone 2, Index map of sites ........................................................... 100 26. Liri Valley Survey - Zone 2, Bronze/Iron Age sites ....................................................... 101 27. Liri Valley Survey - Zone 2, Roman Republic sites ....................................................... 102 28. Liri Valley Survey - Zone 2, Early Roman Empire sites ................................................. 103 29. Liri Valley Survey - Zone 2, Mid/Late Roman Empire sites ............................................... 104 30. Liri Valley Survey - Zone 2; Medieval sites ................................................................ 105 31. Liri Valley Survey - Roccasecca Roman village site ....................................................... 106 32. Liri Valley Survey - Zones. 3 and 4, Index map of sites .................................................. 107 33. Liri Valley Survey - Zones 3 and 4, Bronze/Iron Age sites .............................................. 108 34. Liri Valley Survey - Zones 3 and 4, Roman Republic sites .............................................. 109 35. Liri Valley Survey - Zones 3 and 4, Early Roman Empire sites ........................................ 110 36. Liri Valley Survey - Zones 3 and 4, Mid/Late Roman Empire sites .................................. ·.. 111 37. Liri Valley Survey - :Zones 3 and 4, Medieval sites .......................................................... 112 38. Liri Valley Survey - Parito Villa and surrounding sites ................................................... 113 39. Liri Valley Survey - A. Piumarola area sites; B. Aquinum ............................................... 114 40. Liri Valley Survey - Zones 5, 6, 7, Index map of sites ................................................... 115 41. Interamna: A,B. Site 354, with processing tank; C. Roman bridge over Spalla Bassa , ................. 116 42. A. Inscription (Interamna); B. Remains of water channel (Site 228). ...................................... 117 43. Liri Valley Survey - Zones 5, 6, 7, Roman Republic sites ............................................... 118 44. Liri Valley Survey - Zones 5, 6, 7, Roman Empire sites ................................................ 119 45. Liri Valley Survey - Zones 5, 6, 7, Mid/Late Roman Empire sites .................................... 120 46. Liri Valley Survey - Zones 5, 6, 7, Medieval sites ........................................................ 121 47. Liri Valley Survey - Medieval sites about 800 A.D ....................................................... 122 48. Liri Valley Survey - Medieval sites about 1040 A.D ...................................................... 123 49. Liri Valley Survey - Medieval sites about 1300 A.D ...................................................... i24 50. Liri Valley Survey - Medieval sites deserted after about 1350 A.D .................................... 125 51-53. Drawings of pottery ........ . ................................................................................ 126-134 54. A. Iron Age jug; B. Marble head from Parito; C. Fragment from Ara Murata ...................... 135 55-68. Drawings of pottery ....... . .......... . ..................................................................... 136-158

2.

vii

LIST OF CONTRIBUTORS

R.G.V . HANCOCK

Department of Chemical Engineering, University of Toronto, Toronto, Ont. Canada M5S lAl

J.W. HAYES

9 Linton Road, Flat 1, Oxford, OX2 6UH Great Britain

P. HEMPHILL 18 Aiken Ave., Princeton, New Jersey 08540, U.S.A. (now RHODES-JUDSON)

M.E. LEWIS

Department of Statistic, McMaster University, Hamilton, Ont. Canada, L8S 4M2

G.A. LOUD

Department of History, University of Leeds, Leeds, LS2 9JT Great Britain

A.G. McKAY

Department of Classics, McMaster University, Hamilton, Ont. Canada, L8S 4M2

A. McPHERRON

Department of Anthropology, University of Pittsburgh, PA. 15260, U.S .A.

LP. MARTINI

Department of Land Resource Science, University of Guelph, Guelph, Ont. Canada, NlG 2Wl

E.M. WIGHTMAN (Deceased)

Department of Classics, McMaster University, Hamilton, Ont. Canada, L8S 4M2

D.F. WILLIAMS

Department of Archaeology, University of Southampton, Southampton S09 SNH, Great Britain

viii

Chapter 1

INTRODUCTION

Objectives and Philosophy of the Study E.M. Wightman In the last thirty years, the archaeology of Italy has undergone a transformation. In large part this has been a response to the dual threat posed by the modem, tractordrawn deep plough (Figs. lA, B) and by the sharp upward spiral of building operations of all kinds in hitherto untouched rural areas. The second menace is more serious than the first, for the deep plough brings opportunities as well as destruction, provided these are seized in time. Churning the soil to a far greater depth than the old light ox-drawn plough, it throws up visible testimony of the sites buried below the surface. Though the superficial levels of the sites are disturbed or lost, potsherds and building materials indicate the nature and date of the settlement. Only the plough used for vineyards brings irremediable destruction, slicing the soil to a depth of over a metre, and even it cannot wholly obliterate the traces of previous occupation.

In choosing to follow the plough, as it were, archaeologists were reflecting, and at the same time reinforcing, a trend which diverted attention away from monumental sites to humbler settlements which more realistically mirror the total population of any given period, and which also offer greatest interest to the scholar no longer tied to classical texts, architecture and art history. Moreover, since the plough does not restrict its attention to the sites of any one period, chronological horizons had to be widened, and help sought when necessary from a range of experts. This new attention to rural sites over a wide chronological range entailed a still further widening of concepts, for it had long been recognized in other areas that rural settlement cannot meaningfully be studied apart from the landscape, and understanding of landscape means attention to geomorphology and to soils. As a result, team-work has become increasingly the rule. The lower Liri valley, between Ceprano and the confluence with the Gari (Figs. 2, 3, 4), was chosen as a conveniently definable new area in which to undertake a multi-disciplinary research project in which a geologist, archaeologists and medievalist could fruitfully collaborate in reconstructing the occupational history, with the now established techniques of archaeological surface survey playing a large part in the recovery of data. Because of modem developments attending on the building of the Autostrada del Sole and the encouragement given to industry to locate itself south of Rome, it was an area obviously under imminent threat. It was also an area significantly remote and different from the others in

1

Etruria, Molise and Apulia in which similar work had been, or was then being, undertaken, and it was a question of some importance to see whether similar patterns still emerged. Moreover, the peculiar geology and varied geomorphology of the lower Liri valley appeared, as will be seen, to offer particularly interesting scope for interdisciplinary collaboration. Any project develops its own life and persona as the participants familiarize themselves with the area of work. The landscape ceases to be an abstraction studied from maps and becomes a familiar backdrop against which they work in their endeavour to reconstruct the area's past life. As knowledge increases, it no longer forms a passive background but is recognized as one of the main protagonists in an unending drama, the other being man. The most recent acts are still being played out before our eyes even as we seek for clues to help us reconstruct earlier ones. From earliest times, man and the landscape have interacted; it placed numerous constraints on him, determining for example in which directions he could readily move, while also offering the potential of areas which could be tamed and increasingly used for his purposes as his numbers grew. Man's use of the land in tum altered it: as natural growth was cleared from hills and plain for grazing and agriculture, the inexorable processes of erosion set in, affecting both surface appearance and river behaviour . This use of the land did not form a steadily rising curve, for a marked expansion in the early Roman period was followed after some centuries by a contraction, it will be argued, in both population and land-use, forming a low point from which another growth curve then took off. The last phases of this are still unfolding. Intensive agriculture, including huge fields of maize in the valley floor, exist alongside of road building, new factories and the rapid growth of population in the towns. These last are now sprouting multi-storey apartment buildings whose massive foundations sometimes rip through earlier remains, while on the hill-sides the grass grows in huge tufts because no longer trimmed by grazing flocks. The spread of concrete puts constraints on survey, even if foundation trenches can be of value to both geologists and archaeologists, and when massive amounts of soil, sometimes containing ancient potsherds, are shifted from one location to another, this increases the feeling that the work is being done only just in time. For meanwhile, the deep plough has been at work for some two decades, and while for the first few years it throws up buried materials with great crispness and clarity, so that the dimensions of buildings may be simply measured off the ground, in time it breaks up and scatters these remains over a wide area, rendering them tenuous and hard to interpret.

The life of a project also has its own stages, not all of them predictable: personnel increases or decreases, and the best strategies may need modification in the face of intransigent facts. It also functions at different levels. At bottom are the multitude of practical decisions about dayto-day tactics, and the determined plodding and recording, be it literally at ground level or in the archives, that produces new data. These then require analysis and organiz.ation, so that they can be brought to bear on questions, be used to strengthen or discard previously formed hypotheses. Interdisciplinary team-work demands a particularly high degree of discussion and coordination from start to finish. Some of this is of limited interest except to those who were involved, some is of value to fellow-specialists while a certain proportion must be spelled out because it affects the results. Discussion of methodology and the raw data generated by the proj ect are here presented in the variou s appendices, which thus constitute the basement and ground floor , as it were, of the building , while the earlier chapter s with discussion and conclusions form the upper storeys. Various themes are consistently present throughout. One is the nature and hierarchy of settlements--from town to farmstead--, their location and spread. Another is communication s--the constraints placed on them by the landscape , the positive factors that caused them to develop as they did, be it the strategic needs of the Roman republic or the requirem ents of farmers going to market. The various recorded or likely types of agriculture and animal husbandry will be discussed against the background of soils and landform types, and the social pattern reconstructed on the basis of various types of evidence or, where evidence fails, analogies and hypotheses. All this will be viewed against the historical context afforded by literary or documentary evidence from the latest prehistoric period on, and by the chronological divisions that arise out of archaeological methods. The overriding theme will be the interaction between man and his environment, a changing landscape with moving figures. Finally, what of the valley itself ? Whether viewed from a satellite or by the modem traveller speeding along the autostrada, the lower Liri valley presents itself as a wide, flat, steep-sided corridor running, because of Italy 's overall tilt , as much east-west as north- south (Figs. 2, 4, 5A). Being a continuation of the Sacco valley which starts as far north as the Palestrina gap, it has all the appearance of a natural link between Lazio and the area of Capua, despite some topographical problems at the southern end , where a small but tricky pass must be negotiated. Travelling southwards, the observer will be aware of perched hilltowns at intervals, and may notice that after Ceprano these occur only on the north-eastern flank. The gleaming white surfaces of the rebuilt abbey of Montecassino , occupying the most south-easterly spur of the Monte Cairo massif, catches the eye from afar, and of course affords a magnificent bird ' s eye view to the visitor. The observant person may feel that the extremely flat land around Aquino is anomalous for the Italian landscape, and demands some explanation in terms of geological events, while curiosity

2

may impel the traveller to wonder whether the function of land corri dor was always as obvious as it seems today. These are among the many questions to which answers will be found in the following pages.

Met hods of Stud y Landscape LP. Martini A review was made of existing information on geology, geomorphology, soils and hydrology (Accordi, 1964; Devoto, 1965; Sevink et al ., 1984). This information was augmented with detailed field observations of the areas of interest to archaeologists, interpretation of air-photographs, and analyses of soils samples. Furthermore the valley as a whole was considered to unravel the land use patterns through the ages. This led to the subdivision of the land into several areas: the high mountain, the mountain slopes (hills), the mountain rise , and the valley bottom within which further subdivision could be carried out into stable plain, dissected areas and river terraces (Martini and Wightman , 1987). Of particular interest was the analysis of the river crossings, particularly that of the torrent ial Melfa River, for the challenge they may have posed to Romans and more ancient travellers (Martin i et al. 1987). Detailed studies were made of the terraces and river deposits in the recent channel and preserved and exposed in quarries in the terraces. · Laboratory analysis of soils and bu ried organic matter consisted of chemical and clay mineralogical determinations according standard methods (McKeague, 1978). Some information was obtained also on the concentration s of phosphorous and nitrogen available to crop lands, but no detailed fertility experiments were made. The sediments were dated through C 14 and Urani um series methods, respectively of woody remains and stalactite in caves. Surficial deposits, soils and land surfaces were dated through potsherds and other artifacts studied by J.W . Hayes.

Archaeology (after unpublished annual progress reports for survey years 1978, 1979 , 1980 by E.M. Wightman, P. Hemphill and J. W . Hayes) .

An exhaustive review was made of bibliographic material relating to the Lower Liri Valley, including Roman, Medieval and recent literature and unpublished documents. The Archives of the Abbey of Montecassino proved to be a source of great information for the Medieval period. Such bibliographic studies preceded and paralleled the extensive field surveys which were carried out in the summers from 1978 to 1981. Three surv ey methods were used . (a) The first method was a purposive one whereby sites reported in the literature were visited to collect new artifacts. In most cases the reported finds derived from

modem excavations for sand and gravel and little or no new material was retrieved. Similarly, particularly for determining existence of prehistoric sites, selected mountain areas, caves and rock shelters were visited and analyzed for remains: little of significance was found, most caves having been inhabited as recently as the 1940-50's with clean floors. (b) The second, which may be described as extensive or longitudinal, was employed only where retrieving road paths and locations of river crossing points were considered the primary objectives, for instance during the first phase of study in the Melfa River area, and in areas proximal to the Liri River. Cultivated fields on either side of the possible lines of ancient roads were carefully examined by a pair of walking archaeologists looking for any manufactured remain. (c) The third method was employed whenever "settlement survey" was paramount, that is over all the territories chosen for survey (WardPerkins, 1964; Wightman, 1983). Blocks of land between 0.5 and 1 km in extent, normally bounded by roads and/or streams, were chosen each day and rigorously covered by teams of 2-4 walking archaeologists (Fig. 6A) spaced at intervals of 15-25 m, zig-zagging slightly, walking at the same speed and maintammg continuous vocal communication, calling out and warning each other of any manufactured material found of whatever nature: prehistoric, Roman, Medieval or recent. The presence even of ubiquitous modem tile fragments thus received comment, and whenever 2 or 3 pieces of tile or pottery were observed simultaneously, an intensive combing of the immediate surroundings, widening out from the point of observation, took place . All the artifacts retrieved were recorded and a day book was maintained for debriefing of the researchers. Accurate numbering of sites was made and photographs taken. Airphotos and topographic maps were constantly analyzed to establish the relationship of the newly found sites with the topography and the other sites. The manufactured material, primarily potsherds, was studied, drawn (Figs. 17-18, 5168), catalogued and stored at the Archaeological Museum of Cassino (Appendices A, E). A few selected potsherds and tiles were subjected to petrographic and neutron activation analyses to determine the source of the clay (Appendices B, C). Any analysis of sites discovered by surface survey is based on an assessment of what is actually seen on the ground. The assumption is made, however, that what seen at the surface bears some direct relationship to what is hidden buried undern eath. This assumption could be tested by excavation. In those parts of Italy where surface survey was carried out soon after the introduction of the deep plough, there appeared to be a fairly direct relationship between buried and surface remains. Concentrations of pottery, tiles and other building materials were thick and clearly defined , and criteria based largely on size of scatter and types of materials could readily be evolved. In the Liri Valley, some 20 years of deep plough use has had further effects, and sites are often marked by much larger and vaguer scatters of low density. Criteria based on a mixture

3

of overall size, density and the presence or absence of fmer types of pottery must therefore be used. The possibility of subjective and terrain bias are large. Erosion has also effect in some parts, particularly in the upper parts of hill-sides.

TABLE 1. Categories of sites.

Villa. Large area (2200 sq. m. or more) with building evidence such as plaster, marble, terracing and fine wares (tiles, mosaics, pottery, etc.). Major sites. The Liri Valley "major site" category is of about 1800-2200 sq. m. which would correspond to the small villa category of other areas such as Etruria, or a large farm. Dense scatter of potsherds present. The major sites show signs of careful selection, a gentle southward facing slope being almost invariably favoured. Minor sites. Smaller dimensions (1200 to 1800 sq. m.), less dense scatter ( ~ 100 sq. m.), usually coarse wares found. They probably represent work-a-day farms with no architectural pretensions.

Scatters. Small and undifferentiated, range of coarse tile and pottery sparsely scattered. They represent secondary buildings of some sort. It is difficult to establish the validity of the scatters. At the limit they can be artifacts associated with dumping of material such as manure and detritus in the field. However, after the careful field study done they must be considered as true sites for the sake of this analysis.

Selected areas of the Lower Liri Valley were selected for this type of settlement survey, and the site finds were dated from their pottery and the results are summarized in a series of maps (Figs. 20 to 50). Intensive surficial survey was conducted in a few special areas: such as on the site of Roccasecca village site (Fig. 31), Parito (Fig. 38), Piumarola (Fig. 39A), Interamna (Figs: 40, 41; Hayes and Wightman, 1984), andAquinum (Fig. 20, Appendix D). The area between Interamna and Casinum was resurveyed and sites identified in the 1978 season were re-visited during the 1979 season. No new material was found in the pre-identified sites. Some sites clear in 1978 were diffuse in 1979. No villa (Table 1) is present in the re-surveyed area.

Statistical analyses of archaeological sites E.M. Wightman, I.P. Martini and M.E. Lewis Statistical analysis were conducted of the distribution of Roman and modem sites located on the valley plain from Interamna to the base of M. Cairo. Three approaches were taken to analyze the spatial pattern: the first was based on area, such as quadrant count analysis, the second

considered the nearest neighbour distance between points, and the third involved estimation of the sample spectrum (Lewis, 1983). The statistical analyses were done primarily to test for randomness of spatial distribution of the Roman sites, in respect to a theoretical Poisson distribution, and to compare the settlement patterns of the Roman and modem periods in a semi-regular shaped surveyed region where topographic and soil properties do not vary much. The area was uniformly covered by the survey for Roman sites except for some gaps where modem habitations or factories were built. Such gaps had to considered as areas without Roman sites. For the sake of statistical analysis the surveyed area had to be contained in the smallest circumscribing rectangle. Random site points had to be added to the artificially created area of the rectangle not covered by the survey for the spectral analysis. A total of 77 real Roman sites were included in the analysis, and they were all considered of equal status for the statistical analysis, that is no difference were made on the size of the sites. The distribution of the recent sites was obtained form a 1:25,000 topographic map. In every case it was found that there was a significant statistical variation of the sites from that of a Poisson distribution, suggesting non-random distribution pattern, that is, the pattern of sites resemble a clustered distribution. Comparing the density of the Roman sites with that of modern sites utilizing a Chi Square test, it is found that the density of the two types is not the same over the region, "thus, the pattern of settlement density is different for the two periods, Roman and modem" (Lewis, 1983, p. 34). Furthermore, from the results of the nearest neighbours technique it is possible to establish that "since Roman sites have Roman sites as nearest neighbours unexpectedly often, and likewise for the modem sites, positive attraction amongst types is suggested, and consequently one can say that the sites are partly segregated" (Lewis, 1983, p. 39). Essentially the location of the Roman sites is not significantly correlated with the location of a modem site. A further statistical analysis was conducted on data obtained from blind re-survey (that is not divulging to the field archaeologists what was found the year before) of areas of the Lower Liri valley. The objective was to test whether the re-survey done one year later succeeded in retrieving the same spatial distribution of sites. The resurvey was conducted over 10 % of the area, on random squares of 0.5 km each on different terrains, that is, foothills, plain, and dissected areas. The major sites and most minor sites and large scatters were rediscovered. There was some variation in re-finding of less clear minor sites and smaller scatter sites. The comparison between the original findings and those of the re-survey was done both by measuring the proportion of sites found both times and by comparing distances among the sites that were found in both surveys and compute the deviations. It was found that there was not too much difference (statistically not significant) in the locations in both surveys, although a few extra sites were found in the re-survey indicating both

4

some terrain variation from year to year and the increased experience and familiarity of the surveyors with the general area.

References Accordi, B. (1964). Lineamenti strutturali del Lazio e dell' Abruzzo meridionali. Mem. Soc. Geol., 4:595633. Devoto, G. (1965). Lacustrine Pleistocene in the Lower Liri Valley (Southern Latium). Geologica Romana, IV: 291-368. Hayes, J.W. and Wightman, E.M. (1984). Interamna Lirenas: risultati di ricerche in superficie 1979-1981, Archeologia Laziale VI, Quaderni del Centro di studio per l'archeologia etrusco-italica, 8:13718, Lewis, M.E. (1983). Spatial Analysis applied to archaeological data. Unpublished M.Sc. thesis, McMaster University, Hamilton, Canada. 88 pp. McKeague, J.A. (1978). Manual on soil sampling and methods of analysis. Soil Research Institute, Ottawa, 212 pp. Martini, J.P. and Wightman, E.M. (1987). Geomorphology and ancient settlements of the southern flank of Mt. Cairo, Lower Liri Valley, Italy. Geoarchaeology, 2:131-147. Martini, LP., Wightman, E.M. and Hemphill, P. (1987). Bridging a seasonal torrent. The Melfa River, Southern Lazio, Italy. Geoarchaeology, 2:149-159. Sevink, J., Remmelzwaal, A. and Spaargaren, O.C. (1984). The soils of Southern Lazio and adjacent Campania. Fysisch Geografisch en Bodemkundig Laboratorium, Universiteit van Amsterdam, 143 pp. Ward-Perkins, J.B. (1964). Landscape and History in Central Italy. The second J.L. Myres Memorial Lecture, London. Wightman, E.M. (1983). McMaster University Liri Valley Project, Central Italy. In: Archaeological Survey in the Mediterranean Area, D.R. Keller and D.W. Rupp (Editors), pp. 183-185, British Archaeological Reports, International Series 155, Oxford.

Chapter 2

THE GENESIS OF AN ENVIRONMENT: ELEMENTS OF GEOLOGY, GEOMORPHOLOGY AND LAND USE I. P. Martini

Introduction Described in geological terms, the Lower Liri Valley is part of a large intramontane basin (the so-called Valle Latina) of the Apennine Mountains, a depression which owes its existence to faulting in rocks of the Mesozoic age and is partially filled by later Pleistocene volcanic and fluvio-lacustrine deposits (Fig. 7). The basin is still affected by tectonic movements along these faults and by ensuing differential uplift, that is continuing "neo-tectonic" activities with associated numerous earthquakes. The part of the valley we have studied does not contain volcanoes, but it has been affected by nearby ones. The large Roccamonfina volcano temporarily dammed the flow of the rivers during the mid-late Pleistocene period forming a large lake in the Lower Liri Valley, which previously drained to the sea by joining up with the Voltumo of Northern Campania (Lake Lirino: Ferrero (1879)). Pieces of volcanic material from Roccamonfina and other volcanoes of the Roman and Naples regions have been air transported and have ended up interstratified with other lacustrine deposits (Fig. 8A; Devoto, 1965). The geological history of this valley, and of the Latina Valley in general, is tied up with that of the still growing Apennine Mountains. This brief review will provide a basis for understanding the evolution of the local landscapes, and will indicate the conditions humans have had to cope with through the ages.

Mesozoic and Tertiacy Geology A geological flash-back story can best start in Mesozoic times, about 200 million years ago (Parotto and Praturlon, 1975). At that time the Atlantic Ocean did not exist. Instead, the large Thetis Sea cut a deep eastern indentation in a large supercontinent (the Pangea) which contained most of the landmasses of Earth clustered together (Dewey et al., 1973). Thick layers of carbonates were deposited in the wide shallow sea, and these are the rocks that now form the mountains bounding the Latina Valley of which the Lower Liri Valley is part. During Mesozoic and Tertiary times, just before and continuing during the Atlantic split of the Americas from Europe and Africa, the carbonate deposits were dissected by numerous faults into several blocks. Some of the blocks remained high in shallow sea conditions, others sank into relatively deep submerged troughs. These troughs were the embryos of the present valleys of the Apennines._

As the Atlantic widened, the European branch of the Thetis Sea started to close and several plate-like parts of the earth's crust collided and slid past each other (Vanderberg, 1979). The Alps emerged, and considerable amounts of sand and clay were funnelled through the deep submarine troughs by turbid, sediment-laden underwater currents. These formed the sandstone and shales of the bedrock of the valley floor of the Latina Valley. Further movements of the crustal plates led to the narrowing of the submarine troughs and sliding of material toward the deeper chasms. Large blocks of the ancient carbonate platforms (the present M. Lepini, Ausoni, and Aurunci (Fig. 9) were thrust eastwards toward the Abruzzi platform. Some blocks detached from the advancing thrusts and slid for several kilometres over the submarine shales of the valley bottom. These "olistholiths" (literally, "sliding stones") form several of the isolated mountain peaks of the area, such as M. Leucio (Fig. 9).

Pleistocene Geology The geomorphologic expression of the mountains and their emergence above sea level are in great measure ascribable to rejuvenation of the faults during Pliocene and Pleistocene times. Two major systems of "normal" faults developed, one parallel to the main Apenninic trend from the northwest to the southeast, and a secondary system almost perpendicular to the first one, from the southwest to the northeast (Accardi, 1964; Devoto, 1965). It is along those faults that the horsts (elevated parts) of the M. Cairo and the dissected hors ts and grabens (depressions or sunken parts) of the mountains to the south (M. Lepini, M. Ausoni and M. Aurunci) developed (Fig. · 9). The main northwest to southeast trend is responsible for the formation of the major longitudinal valleys. It is also responsible for the development of volcanic centres which in general evolved first to the northwest (M. Amiata in Tuscany, and the Colli Albani south of Rome) and become younger to the southeast with Vesuvius, Etna, Volcano and other minor ones still active in the south (Fig. 2). The largest volcanoes were localized at the intersections between the two major fault systems. One such large volcano was the Roccamonfina which during Pleistocene times exploded and evolved into a large caldera or sunken crater (Giannetti, 1970; Remmelzwall, 1978). Available radiometric dating indicates that the Roccamonfina formed approximately 1 to 1.4 million years ago, and was still active approximately 35 thousand years ago. The lava and tuff flows of this volcano dammed the Lake Lirino at Suio (Figs. 7, 9). During its existence, the Lirino Lake was partially filled

5

with a variety of sediments, most often containing fossils indicating cool to cold climates (Devoto, 1965; Settepassi and Verdel, 1965). In the southeastern, deeper parts of the basin, fine calcareous deposits (marls) were interstratified with volcanic tuffs (Fig. 8A). The tuffs and other volcanic debris were air transported from the adjacent Roccamonfina and, in the case of the finer cinder, from other volcanoes some as far away as the hills of Rome. The marls are fine grained toward the deeper southwestern part of the lake and become coarser toward the eastern shores, particularly near M. Cairo. They are capped by coarser calcareous muds and sands. These are in tum covered by gravelly sand to sandy gravels which formed primarily in the alluvial fans of rivers exiting from the narrow lateral gorges cut into the mountain. Locally, colluvial fans formed on the flank and toe of the mountain and are composed of angular pebbles, boulders and reddish mud moved down the mountain slopes by mudslides during periods of intense rainstorms (Fig. 7). While the Lirino Lake was progressively filled with sediments, its water level was also lowered by the incipient downcutting of the volcanic barrier near Roccamonfina at Suio. The last major lacustrine deposit (travertine) was formed near large hard groundwater springs. For instance, some thick travertine deposits formed in the Aquino area along the major normal fault paralleling the southern flank of M. Cairo (Fig. 7), about one quarter of a million years ago (as determined by radiometric dating, Schwartz, pers. comm.). Many travertine deposits, such as parts of those of Aquino and those of smaller pods at S. Angelo in Theodice and Pontecorvo, which are associated with minor transversal faults, contain fossils of encrusted swamp reeds and other vegetation which indicate a relatively cold climate, possibly associated with the penultimate glaciation of Europe (Devoto, 1965). Reactivation of movement along normal faults and further erosion of the volcanic dam eventually drained the lake. The baseline level dropped, the rejuvenated rivers downcut into the lacustrine sediments and locally deposited fluvial gravels. The resistant travertine deposits of Aquino, Pontecorvo and S. Angelo in Theodice became exposed above water level and for the first time parts of the valley floor, albeit somewhat swampy, became available to humans. Scattered Paleolithic tools found together with animal bones indicate that hunting expeditions followed the drier, gravelly, early fluvial terraces surrounded by swamps, and utilized the higher and drier travertine outliers (Biddittu and Segre, 1976). As progressive downcutting of the lacustrine and previously formed alluvial fan deposits (such as along the Melfa River) occurred, terraces developed, primarily along the Melfa and the Liri rivers. Remnants of these terraces are present along the Melfa as thin (2-3 m) gravels. Better preserved and wider terraces occur along the Liri River. Four major terrace surfaces are recognizable there, three along the left bank, and one along the right bank (Fig. 10). Each terrace is characterized by gentle slopes,

6

sometimes with minor steps, leading towards the present Liri River (Figs. 1lA, B). Some of the terraces, particularly the wider third one along the left bank, are backed by high (10 m) and steep back scarps which join the terrace surfaces by means of a rise formed by colluviated material coming down from above. Other terraces are separated from each other, or from the modem floodplain, by minor scarps. The youngest terrace forms also the modem floodplain, only partially abandoned · by the renewed entrenchment of the river, in a lower channel. Along the Melfa as well as along the Liri River, there are well developed, abandoned meander loops on several terraces. It is at the side of one such loop, high above a steep scarp, that the Roman colony of Interamna was built in the late 4th century B.C. (Salmon, 1967). The terraces are invariably covered by a relatively thin deposit (in the order of a few metres) of fluvial sands and sandy gravels, the one spread over the ancient floodplain, the other concentrated in abandoned braided as well as meandering channels. The various terraces of the Liri and in part those of the Melfa can be readily recognized and correlated both longitudinally and transversely along the alluvial valleys (Figs. 10, 12). However, the surfaces of the terraces differ. For instance, the surface of the third terrace is still regularly inundated near the junction of the Liri with the Gari (Giuntura area, Figs. 8B, 9), and if it were not for artificially improved drainage, it would be a wide wetland with local open stagnant ponds. By contrast, the surface of the same terrace is located 20 m above the river, and is totally divorced from it, in the upstream reaches in the Fontana della Fica area, near the Melfa confluence (Figs. 9, 12). If not used for agriculture such an area would be locally wet because of excessive seasonal drainage from the higher lands, but in general it would support a forest of oak (Quercus). The small tributaries of the Liri have entrenched the lacustrine sediments as well, some along structural alignments possibly associated with normal faults (Fig. 7). The area underlain by the more resistant travertine deposits of Aquino is not dissected, except alo~g the Forme d' Aquino (Fig. 3). The surface smoothness of the Aquino plain is also enhanced by spread-out deposit (order of a few metres thick) of argillaceous red soil, sheetwashed over the irregular leached carbonate surface from the nearby steep, high M. Cairo. The southeastern parts of the valley, underlain by softer sands and muds, are, by contrast, strongly dissected by the tributaries. Deep, narrow valleys alternate with sharp crested ridges. These sloping ridges link longitudinally the Aquino plain with the river floodplain of the Giuntura area.

Geomorphology and Land Use Geological processes and the different resistance of the materials to erosion have led to different landscapes. The non-volcanic portion of the valley we have studied can be subdivided into several units: the southern mountain slopes of M. Cairo, the valley floor, and the southern mountains.

Our primary concern here is with the slopes of M. Cairo and the valley floor, which can be subdivided further as follows. The slopes ofM. Cairo comprise the high slopes, the highlands, the lower slopes and the mountain rise, which in tum has both alluvial and colluvial fans (Fig. 13A). The valley floor can be subdivided into the Aquino plain, the dissected area, and the river terraces.

Mounta in slopes The various elements of the southern flank of M. Cairo have the following maj or characteristics (Martini and Wightman, 1987). High Slopes. They are fault bounded, steep, mostly showing bare rock (limestone), except where there are grassy meadows above the treeline. Forested lower areas retain some of the organic-rich , generally coarse-grained (sandy and gravelly) soils. The steepness of the slopes makes these areas unsuitable for intensive use. Only local grazing and wood cutting have been practised in the documented past, and this continues today. H ighlands . These are mountainous zones with gentler slopes. They form the flatter top parts of normal faulted block s. They have retained much of the soil washed downhill from higher slopes. Those areas have often been terraced, and have retained reddish gravelly and sandy loam soils, up to a few metres deep, suitable for agriculture, and capable of producing grains, fodder crops, some wine and fruits.

Lower mountain slopes. These are steep sloping areas bounded by normal faults (Fig. 13A). Some preserve the steep triangular shape which characterizes active faults. They have undergone considerable soil erosion and many rock outcrops exist. Vegetation grows quickly, but fires are frequent and the removal of the vegetation triggers renewed soil erosion . Small caves are present and in a few of them Paleolithic tools have been found (Biddittu and Segre, 1976). The lower parts of ~e slopes, transitional to the steeper colluvial fans, are suitable for terracing, and cultivation of olives has been practised. Mou ntain rise. The mountain rise is primarily characterized by both colluvial and alluvial fans. These fans can be viewed as traps which catch most of the mountain soils and materials which are washed downward. Only the finer materials escape the traps and are spread out on to the nearest parts of the valley floor. Towards the apex, the fans generally contain coarse pebbles and boulders in a matrix of sandy loam. Layers of waterlaid , well-sorted gravel alternate with the poorly-sorted gravelly and silty deposits characteristic of debris flows. Farther down the fan, the grain size fines out, debris flows decrease in importance, and gravelly layers alternate with progressively thicker sandy loam beds of reworked reddish mountain soil. The gravelly material is still present in minor measure at the toe of the slope, but disappears rapidly in the adjacent plain. The difference between the

7

colluvial and alluvial fans is more in terms of relative concentration than in drastically different types of sediments. The colluvial slopes are steeper, have more angular pebbles and have a predominance of debris flows. Sheetwash occurs, however, and reworks a lot of the finer reddish soils. The gentler portions of the fans are some of the best lands available, and have been utilized since antiquity. Bronze Age, Iron Age, Roman, Medieval and modem farmers have all used these lands. It was with good reason the Romans built some of the most luxurious villas of the region on them, such as Parito just east of Piedimonte (Fig. 13B). The well drained soil is a mixture of fertile loam mixed with coarser sandy and pebbly material which makes it easier to plough than the clay-rich soils of the adjacent plains. Furthermore, these areas are located above the mist of the valley floor; they have good views, southern exposures and easy access to the higher mountainous lands which are suitable for grazing and wood collecting. Water is available from nearby large springs at the foot of the mountain, or at relatively shallow depths below the sediments of the fans. Valley floor

The landscapes of the valley floor have the following major characteristics: Aquino Plain. This is a wide area to the northeast of the valley, flat to gently sloping to the southwest and underlain by travertine (Figs. 7, 14). The topmost soil, however , was not formed by weathering of the travertine, but it mostly derived from sheetwash of the adjacent slopes of M. Cairo. The soil is generally rich in clay and becomes easily waterlogged during the winter and spring, and experiences dryness and intense mud-cracking during the summer droughts. It can be worked only during certain times of the year when it is still moist but not soaked. If too dry it is at first difficult to break and is then pulveri zed as it is exposed. This destroys the required structure for optimum plant growth. When too wet, it physically hampers movement of workers on the fields, but more importantly, when ploughed, a clay film is smeared on the "peels" (soil units, small blocks) and clogs their surface. This inhibits the most effective transfer of nutrients from the centre of the peels to the roots of the crop plants, which tend to penetrate more readily between peels. When cultivated properly, with ploughing at the right time of year, this soil is fertile.

The swampy conditions which may have persisted in some places among the oak forest after the drying of the original lake, and the difficulty in cultivating the clay-rich soils, may have limited early colonization of the valley floor . It is not surprising that only a few Iron Age sites were established there, and that full colonization of the valley dates only to Roman times. Once colonized , the distribution of soils and landforms on the valley floor did not provide s~rong reasons for favouring any particular

location for farm houses, except for some high banks of streams, such as the Formed' Aquino. A more important anthropogenic factor for the location of farmhouses was the proximity to roads.

Dissected area. This is characterized by steep flank, narrow stream valleys alternating with narrow ridges elongated to the southeast (Figs. 7, 14). Much soil erosion has occurred along the slope of the valleys since the original forest vegetation was removed. Thus, a series (catena) of soils can be recognized ranging from wet colluvial to alluvial soils at the bottom of the valleys, through colluvial soils at the rises, to eroded flanks exposing lacustrine substratum, and finally to narrow, flatter tops of the ridges where deeper, more intensely weathered top-soils could develop because of reduced erosion. All these soils, even the continuously eroded (young) soils of the valley flanks, are relatively fertile, because they contain great amounts of nutrients, some associated with interspersed or interlayered volcanic materials deposited during the Lirino Lake stages. The steepness of the slopes and the humid narrow bottom valleys are the two major constraints for localization of settlements (houses) in the area. Therefore, since Roman times, the farm houses have been strung along the ridges, just off the crest, preferentially on the south facing upper slopes. The top of the ridges and part of the slopes are to-day cultivated for fodder-crops and grains, and some of the wider bottom valleys for com and cash crops, such as tobacco. Many suitable southward oriented slopes have good vineyards. The steepest slopes are not cultivated and are covered by underbrush (macchia).

River terraces. Along the major rivers and the Liri River in particular, five major planar surfaces are recognizable, the highest one merging laterally with the lacustrine plain of the valley floor and the other four being well defined river terraces (Figs. 9, 10, 14). There is a drastic longitudinal variation in the surface characteristics of these terraces. The soil, for instance, changes longitudinally along the terraces, consisting of gravelly sandy loams on the higher lands, and sandy loam in the lower areas, near the junction of the Liri and Gari Rivers. They are all fertile. However, the upstream higher plains are at present cultivated with cash crops, with the help of intense summer irrigation, while the downstream lower surfaces, which are still subject to yearly floods, bear grains and cash crops. The soil also changes transversally along the terraces, consisting of gravelly sandy loams in the higher lands, and sandy loam in the lower surfaces, near the junction of the Liri and Gari Rivers. These lower lands, as well as the whole of the valley of the Gari, were improved by drainage schemes undertaken before the Second World War (Mazzocchi, 1961). Relatively few pre-modem sites have been found on the terraces except where strung along roads, apart from special buildings such as the Roman bathhouse of Interamna located on the bank of the Liri River. Roman

8

farms, like the town of Interamna itself, were mostly placed at the edge of the topmost scarp of the terraces. Parts of the terraces have been used for cultivation or grazing. Although the landscape was much similar in ancient historical times to the present one, the vegetative cover and the drainage may have been rather different, and wetlands breeding noxious insects (perhaps but not certainly the malarial mosquito) may have existed, as is recorded in Medieval manuscripts. It is possible that the population pressure, even during Roman times, was never so strong as to require reclamation of the terraces for settlement.

Bridging the Rivers The communications network of the valley as a whole and the crossing of the various rivers, in particular, are intriguing aspects of the interaction of man and the environment. For instance, the localization of the Liri River along the southwestern side of the valley, against the rugged irregular flank of the southern mo:untains, forced most roads and activities on to its flatter northeastern, left side. As for river crossing, three major, very different rivers (Gari, Liri, Melfa) had to be negotiated, each requiring a different crossing strategy.

Gari River Perhaps the easiest river to deal with was the Gari (called the Rapido upstream from Cassino) (Fig. 9). The Gari River is a perennially flowing, meandering, relatively small river with comparatively little variation in flood conditions as it is fed primarily by large permanent springs near Cassino. It has developed a deep narrow channel (several metres deep, less than 50 m wide), cut into lacustrine volcanic and locally travertine deposits. Some reaches of the channel are now narrower than in the past due to reclamation of its floodplain. The main roads were routed during Roman times as they are now, upstream from the springs at Cassino, where the smaller Rapido River is bridged to the east. Temporary bridges were built by American soldiers over the Gari River itself, near the junction with the Liri, during the Second World War when the Cassino area was held by German troops. A medieval bridge is recorded at S. Angelo in Theodice where a recent bridge still exists. This Medieval bridge and the regular discharge of the Gari were set to good use by constructing mills on them: the number recorded, thirteen (Loud, personal communication), indicates a wider river-bed than at present.

Liri River The Liri River is also a perennial river, characterized by an irregular, entrenched meandering pattern. Straight reaches alternate with sharply turning tracts where faults may have influenced the river course (Figs. 7, 9). The river has migrated actively in historical times, and traces of the ancient courses are retained as scars on the floodplain in the Giuntura area (Fig. SB). Upstream, the

river is more deeply entrenched into the lacustrine deposits and except for parts of some of the larger point bars , flooding occurs rarely (such as in the proximity of S. Giorgio) or not at all. Neverth eless, the river undergoes high flood periods and the bridge build ers had to cope with them. Few bridges cross the Liri in the studied area. In the past there were several boat-crossing sites all along its course. A major bridge now links Cassino with S. Giorgio and the Tyrrhenian Sea. It is located in the vicinity of Interamna, not too far from the site probably used by the Romans. Although limestone blocks of possible Roman cut have been reported from the vicinity of the baths of Interamna on the river bed, we have never observed them as this perennial river has generally deep murky waters. No Roman remains of construction were observed on the river banks either, but this is not surprising because of the erosion of the steep outer bank cut into lacustrine marls, and the development of a large sandy bar along the depositional bank which may have obscured any existing remnants of piers. There is no record of permanent bridges over the Liri in this area during Medieval times, partly because the bridge at Pontecorvo had assumed the principal crossing function (Figs. 3, 9). Only after Worl d War II was a permanent bridge built over the Liri in the Giuntura area.

Melfa River Perhaps the most interesting and troublesome river to cross was the Melfa (Martini et al. , 1987). The river has torrential characteristics, alternating between a dry bed in summer and flash floods during winter and spring rainy seasons. The power of such floods is attested by the remnants of bridges, ancient and modem , and of fragmented retention walls strewn all along the stream (Figs. 15A, B). The river can be readily forded for most of the summer. This has been done since ancient times. Temporary wooden bridges may also have been built for winter travel, and reconstructed after each flood. As the importance of the roads increased and Roman conquest led to the strategic rapid route of the Via Latina, more permanent structures were required to assure passage throughout the seasons. The Medieval and modem traffic-intense routes span the river where it is deeply entrenched into a gorge and where consolidated gravels and sands exist on the banks. The Romans instead opted for fording localities and built their Via Latina bridge on a relatively wide, braided portion of the stream. The result was that the bridges were destroyed repeatedly by the uncontrollable floods. The remains of wooden piles in the stream floor, unearthed by recent gravel extraction, and the remains of probable massive stone piers indicate that the strategy of the Romans was to build solid piers, perhaps protected by leading walls which funnelled much of the waters, gravel and other debris in between them (Fig. 16). The roadway may have been of wood and thus readily detachable, so that when topped by the floods, it could give way without damaging the piers themselves,

9

thus making for easy and rapid repairs of the route (Martini et al ., 1987).

Evolution of the Alluvial Valley The origin of river terraces and their fluvial deposits and soils has been a matter of long standing controversy in the geomorphological literature of this century, It entered the archaeological literature after the synthesis of the archaeological and geomorphological evidence for the Mediterranean valleys was pub lished by Vita Finzi in 1969. Vita Fin zi argued that if time telescoping is allowed, then we can recognize maj or periods of river deposition alternating with periods of river entrenchment throughout the whole Medi terranean basin. This is presumed to be caused both by climatic variations and human actions. For instance, deforestation during the Iron Age, the various soil and water conservation practices during Roman times and their subsequent abandonment in the Middle Ages could well have led to some erosiondeposition cycles. Accordingly, Vita Finzi recognized an "old alluvium " possibly formed during the ultimate glaciation, followed by a period of entrenchment, followed by the deposition of a "new alluvium" which has buried some Roman artifacts, the whole then followed by renewed stream erosion which is still active today. Whereas the overall scheme has some general validity and serves as a useful synthetic and mnemonic model, its indiscriminate application undoubtedly obscures the understanding of particular valleys. The evolution of some valleys, such as the Lower Liri, has occurred through a complex series of geological, geomorphological and human events. This valley is typical of those formed in young, growing mountains such as the Apennines, where the influence of overall climatic changes associated with glaciations is modified by local orographic effects, and the geomorphology and sediments are also affected by rapidly rising volcanic and faulted ridges. The effect of human activities has indeed been strong in the Lower Liri Valley, but in geological terms they have been temporary, and of limited (superficial) effect. Deforestation, for instance, triggered a renewed soil erosion of the mountain slopes. However, all what was accomplished during historical times is perhaps equivalent to the results of a single instantaneous disruption of the whole vegetation by a relatively minor ancient eruption of the Roccamonfina volcano. In terms of the formation of river terraces, man had some influence by temporarily accelerating downcutting or triggering periods of more frequent flooding. Apart from protective measures that may have been constructed along the rivers (retention walls, canalizations of major and minor streams, damming, rerouting of waters, draining of inland swamps and lakes all undertaken in this century) the response of the river in the valley primarily depends on what is done on the mountain slopes. Deforestation and building (roads, habitation) practices trigger slope erosion, and therefore

more sediment is carried by the rivers. The energy of the streams is expended in transporting the sedimentary load, no erosion of the channel occurs, hence siltation of the lower reaches develops and that leads to overbank floodings, with deposition of sediment on the floodplain. Conversely, soil conservation practices on the mountain slopes and reforestation reduce the amount of sediment available to the streams. The amount of flowing water is reduced as well, but not as much. The excess energy of such streams is expended in eroding the bed and banks downvalley and thus entrenchment occurs. This energy balance system tends to equilibrium when the baseline of the rivers is stable (such as a stable or slowly changing sea level), and the drainage area is affected by stable or slowly changing climates and tectonics. However, the Liri River system has been affected by strong changes due first to the relatively rapidly built volcanic dam on the flank of the Roccamonfma Volcano, and by the subsequent faulting and rapid erosion of this dam near Suio. A lake (Lirino Lake) was formed first and sediment accumulation occurred in the Liri Valley when the baseline of the river was high. Subsequently, the rapid lowering of the dam, and thus of the river baseline, led to the entrenching of the channels, and to the abandonment of the lacustrine plain and of the various terraces. At the present time, the river has not reached equilibrium yet and continues to downcut.

References Accordi, B. (1964). Lineamenti strutturali del Lazio e dell' Abruzzo meridionali: Mem. Soc. Geol., 4:595-633. Biddittu, I. and Segre, A.G. (1976). Giocimenti preistorici e quatemario della provincia di F rosinone. Boll. Instituto Storia e Arte del Lazio Meridionale, IX: 21-44. Devoto, G. (1965). Lacustrine Pleistocene in the Lower Liri Valley (Southern Latium): Geologica Romana, IV:291-368. Dewey, J.F., Pitman III, W.C., Ryan, W.B.F., Bonnin, J. (1973). Plate tectonics and the evolution of the Alpine Systems. Geol. Soc. Am. Bull. 84:3137-3180. FAQ-UNESCO. (1974). Soil Map of the World. I, Legend. Paris: UNESCO, 498 pp. Ferrero, L.O. (1$79). Le terre della Provincia di Lavoro. - Profili sopra la costituzione del suolo della provincia, con attinenze della Geologia con lo studio del suolo agrario. "Per l'Esposizione Regionale di Caserta", 76 pp. Caserta. Giannetti, B. (1970). Contributo alla conoscenz.a delle lave leucitiche e delle piroclastiti della cinta calderica di Roccamonfina e petrochimica del complesso vulcanico. Mem. Soc. Geol. Ital. 9:497-556. Martini, I.P ., and Wightman, E.M. (1987). Geomorphology and ancient settlements of the southern flank of Mt. Cairo, Lower Liri Valley, Italy. Geoarchaeology. 2: 131-147. Martini , LP., Wightman, E.M ., and Hemphill, P. (1987) . Bridging a seasonal torrent: The Melfa River, Southern Lazio, Italy. Geoarchaeology. 2:149-159. Mazzocchi, A. (1961). La Bonifica della valle del Liri 10

Cassa per il Mezwgiorno - Studi Testi 3. 1st Edit. Mezzogiorno. Naples. 332 pp. Parrotto, M., and Praturlon, A. (1975). Geological Summary of the Central Apennines. Quaderni della ricerca scientifica. No. 90, 57 pp. CNR Rome. Remmelzwall, A. (1978). Soil growth and Quaternary landscape development in the 1yrrhenian coastal area of south-central Italy. Puhl. Fys. Geogr. Bodemk. Lab. Univ. Amsterdam. 311 pp. Salmon, E.T. (1967). Samnium and the Samnites. Cambridge University Press. 446 pp. Settepassi, F., and Verdel, U. ( 1965). Continental Quaternary Mollusca of Lower Liri Valley (Southern Latium): Geologica Romana. IV:369-452. Sevink, J., Remmelzwaal , A., and Spaargaren, O.C. (1984). The soils of Southern Lazio and adjacent Campania. Fysisch Geografisch en Bodemkundig Laboratorium, Universiteit van Amsterdam, 143 pp. Vanderberg, J. (1979). Reconstructions of the western · Mediterranean area for the Mesozoic and Tertiary timespan. Geologie in Mijnbouw. 58: 153-160. Vita-Finzi, C. (1969). The Mediterranean Valleys. Cambridge Univ. Press. 140 pp.

Chapter 3

PREHISTORIC

SURVEY IN THE LOWER LIRI VALLEY A. McPherron

Introduction

Prehistoric Occupation of the Valley

Objectives and methods of study

Palaeolithic

The study was designed to locate sites from the Palaeolithic through the Iron Age in the Lower Liri valley and immediately adjacent area. The methods employed were as follows. 1. Existing geological sections were examined for possible Pleistocene and earlier Holocene deposits and evidence of human activities; 2. caves and rock shelters on the slopes on both sides of the valley were located by purposive search and were inspected; 3. purposive survey was conducted in places which existing information on site location in different periods indicated to be promising. Since the aim was to locate sites and not study regional distribution, probabilistic or other systematized survey designs were not deemed appropriate. On the other hand, the surficial archaeological survey of over of the valley floor done to detect Roman and Medieval sites yielded only two occurrences of more than one flaked stone artifact per hectare. Work in the study area began over a hundred years ago, but has consisted mainly of salvage and study of chance discoveries. Although much useful information has been gained in this way, there has never been a systematic field survey, apart from some cave exploration.

Recent discoveries at Isernia La Pineta, about 50 km East of the Liri Valley, indicate the presence of hominids in Central Italy by the beginning of the Middle Pleistocene, about 730,000 years ago. The lithic industry there consists of choppers and flakes whose topological relationship to similar lithic industries elsewhere is as yet unassessed. The fauna is of Middle Pleistocene aspect (Coltorti et al., 1982).

Age of sediments The portion of the Liri Valley studied contains was inundated in middle Pleistocene by the now drained Lake Lirino (Devoto 1965). A palaeomagnetic study of lake-b ed sediments (" lower lacustrine mud" and "typical lacustrine facies" of Devoto, 1965) indicates normal polarity for all the samples collected, suggesting that the lake deposits are younger than 730,000 years ago, the time of the last Earth's palaeomagnetic reversal. That indicates that the damming by lava flows which led to the flooding of the valley took place in late Lower or middle Pleistocene times. On the other hand we were not able to obtain any Late Pliocene date from lacustrine sediments (Devoto , 1965).

In the Liri Valley north of the study area, lithic artifacts classified as "Pre-acheulean" comparable to those from Isernia La Pineta, have been reported near Ceprano where they are in association with a Middle Pleistocene fauna (Biddittu and Segre, 1977; Segre et al., 1982). Similar 1t1aterial occurs as surface scatters in several locations in the same area (Fig. 17). The location of these finds, near the northern shores of Lake Lirino, and their absence from the parts of the valley bottom examined by the project is probably to be explained by the presenc e at that time of the ancient lake. The earliest artifacts reported for study area are "evolved" Acheulean bifaces and flake tools . Geological cont ext, associated fauna, and bifa ce morphology sugg est a date toward the end of the Acheulean sequence, perhaps 300,000-150,000years ago. Their location within the area of ancient Lake Lirino indicates further that the lake had largely or entirely drained from the valley by this time. The finds of this time period were made near Pontecorvo , Aquino, Piedimonte S. Germano, Pignataro Interamna , and S. Angelo in Theodice. The artifacts were found in a matrix of sands and alluvium (presumably the "old terra ced gravels" of Devoto , 1965), suggesting that they came from campsites along the Liri or tributaries. Only a gravel pit near Pontecorvo yielded more than a handful of artifac ts; they occurred at a depth of about 3m (Biddittu and Cassoli , 1969). No site was excavated so as to expose ho rizontal surfaces, so one cannot judge now the artifacts and bon es came to be deposited where they were found, no r can we even be sure that all occurrences date to the same timeperiod. However the survey project encountered no artifact s clearly attributable to the Acheulean. Rather, the earli est phase represented in the survey (judging by tool typolo gy) is the Middle Pala eol ithic or Mou sterian (abou t 50 ,00035,000 years ago) . All finds came from an area at S.

11

Scolastica (Fig. 18). Although artifacts of other periods were not apparent in this collection, it is naturally unclear whether all the pieces found belong in fact to the Mousterian. The stratigraphic situation at S. Scolastica is worthy of note: the Palaeolithic artifacts were found on the surface, while an Iron Age site was found at a depth of 4m. The S. Scolastica site is located at the toe of the steep slopes of Mt. Cairo (Fig. 19A), and the stratigraphic position of the finds indicate that considerable slope erosion has occurred and the deposition at the mountain toe-area has created an inverted stratigraphy whereby younger material (iron age artifacts) have been buried by older Palaeolithic deposits. Several blades, some with heavy steep retouch, were found during survey in different parts of the valley. They are probably Upper Palaeolithic in age. The few them with diagnostic characteristics are typologically early Epigravettian (14,000-10,000 years ago). Again only one concentration was found in surface survey. The general conclusion is that geological factors make surface survey for the purpose of locating Palaeolithic sites of little value. The survey of the valley bottom did, however, confirm the evidence assembled by other workers (Biddittu and Segre, 1977). Early sites are either deeply buried or have been destroyed through erosion. Even those found at some depth below the surface appear to have been redeposited by natural agencies. The two concentrations (Mousterian and Upper Palaeolithic facies respectively) were both found on colluvial fans, and were probably redeposited in recent times onto their present location from primary sites higher on the hill slopes.

Early Holocene (Mesolithic, Neolithic, Eneolithic) No artifacts suggestive of the Mesolithic or late Epigravettian (sensu Barker 1981, ca. 10,000-5,000 years ago) were identified. Any encampment on the valley floor would no doubt be deeply buried in most places, and in any case would consist of little but thin lithic artifact scatters. The Neolithic/Eneolithic was also not represented, either in the study area or elsewhere in the valley. Biddittu (personal communication) reports a near total absence of such evidence in the lower and middle Liri Valley. It may be that the heavy soils of the valley were not tillable by hand cultivation methods. Such a hypothesis would fail, of course, if sites of this period elsewhere in Italy were found on similar soils. It also needs to be established that soils in the Neolithic were of the same composition as those of today.

References Barker, G. (1981). Landscape and Society: Prehistoric Central Italy. Academic Press, London. 12

Biddittu, I. and Cassoli, P.F. (1969). Una stazione del Paleolitico inferiore a Pontecorvo in provincia di Frosinone, Quaternaria 10:167-197. Biddittu, I. and Segre, A.G. (1977). Giacimenti preistorici e quateranrio della provincia di Frosinone, Bolletino dell 'Istituto di Storia e Arte per il Lazio Meridionale IX.1-2:21-44. Coltorti, M., Cremaschi, M., Guereschi, A., Peretto, C., and Sala, B. (1982). L'accampamento preistorico di Isernia La Pineta. In: Atti della XXIII Riunione Scientifica. Il Paleolitico inferiore in Italia (Firenze, 7-9 maggio 1980). Istituto Italiano di Preistoria e Protostoria, Firenze, pp. 577-587. Devoto, G. (1965). Lacustrine Pleistocene in the Lower Liri Valley (Southern Latium): Geologica Romana, IV:291-368. Segre, A.G., Biddittu, I., and Pipemo, M. (1982). 11 Paleolitico inferiore nel Lazio, nella Basilicata e in Sicilia. In: Atti della XXXIII Riunione Scientifica, 11 · Paleolitico inferiore in Italia (Firenze, 7-9 maggio 1980). Istituto Italiano di Preistoria e Protostoria, Firenze, pp. 177-206.

Chapter 4

BRONZE, IRON AGES AND ETRUSCANS The Bronze Age

The Iron Age

E.M. Wightman

E.M. Wightman

Only a few Bronze Age sites were discovered by the survey, and all were on the edges of the valley. Sherds clearly in the Apennine Bronze Age tradition came from site 949963 (Appendix E: 159). This site, lying at a height of 250 m., well illustrates the problems of surveying in rough, hilly territory. The pottery was discovered only because of quarrying operations, and it may well have slipped down the hill, since further hand-made prehistoric pottery was observed at 948967 (157), some 500 m north and 250 m higher, on the summit of a spur descending from M. Aglietta, itself a south-westerly spur of the M. Cairo range. The exact location and size of the site is therefore not determinable, the more so since further Bronze Age pottery, mixed with Iron Age, was observed at 949962 (160) some 100 m below the quarry. A similar problem affects the site of S. Scolastica, where Bronze Age material was found earlier during culvert construction but only Iron Age material was observed by the survey; this is an area where there is evidence of extensive hillwash (Fig. 19A). M. S. Silvestro above Capo d' Acqua may have been occupied during the Bronze Age, though the certainly identifiable pottery, which here too had slipped down the hill, was of Iron Age date (Figs. 19B, 54A). On the south side of the valley, south of S. Apollinare, another easily defensible site may have Bronze Age roots: this is Morroni, 042815 (514), where the westerly of two curious limestone protuberances, and the tiny valley between them, contained much early Iron Age and some possibly BA pottery. The hill commands extremely extensive views of Liri, Gari and part of the Garigliano valleys. High sites thus seem to have been favoured, whether or not for reasons of defence, and these situations are consistent with an economy in which pastoralism rather than agriculture predominated. The one exception is a small site on the right bank of the Rio Moscosa west of Roccasecca at 854033 (85), though it too lies over 200 m high, and the valley slopes are not especially attractive for agriculture.

When the Iron Age is reached, the most salient feature of the archaeological record is undoubtedly the hill-top fortifications - Rocca d' Arce, Cassino and S. Vittore along the north side of the valley, M. Nero (near Castro dei Volsci) on the south. These have obvious analogies to Volscian and Latin sites further north and west - Sora and Arpinum in the middle Liri, the Latin colony of Segni, the cities on the seaward side of the M. Lepini - as well as to fortifications in the heart of Samnite territory. They indicate the growth of political entities with organizing power, centred on sites that had attained salience within a certain territory. There are however problems in the dating of these fortifications, for which there is no direct archaeological testimony within the area of the valley, so that recourse to other places (themselves rarely having a secure date) must be made. Moreover, the fortified sites logically represent the end of a process, and search must be made for prior stages.

Since the survey did not concentrate on the higher ground, other sites have doubtlessly been missed. But a high population does not seem to be indicated, and the evidence is consistent with a view of the valley which also emerges in other periods - that it was relatively cut off and untouched by developments which have occurred further north and south in the peninsula. The hints of continuity of sites through into the Iron Age are to be noted, but cannot be pressed too far on the examples available. Chronological refinements are also not warranted.

13

Recent researches in northern Lazio close to Rome have shown that the larger fortified sites, with bank or wall of tufa blocks, were preceded by smaller sites of 1-10 ha with earthwork or rubble defenses, while in some cases defence may have come from the topography alone. In the Liri valley, no examples of earthworks have come to light, but given the lack of excavations this does not preclude their existence. Four sites at least - M. S. Silvestro, the spur of M. Aglietta, Morroni and M. Trocchio - are obviously defensible from topography, and artificial defenses could well have been added as needed (on M. Trocchio they hardly were). Size can be determined only for Morroni, where the occupied area does not exceed 2 ha; here, it is perfectly possible that the rough terracing still visible on the northern flank of the hill (the other sides are precipitous) represents the remains of an earthwork, and stone revetting, though not visible, cannot be ruled out. As seen, the pottery from Morroni suggests an early phase of the Iron Age, probably not later than the 8th-7 th centuries. That Casinum (Cassino) was already inhabited at that time is shown by the cemetery near the later amphitheatre, which has graves of 8th-6 th century date, and an earlier phase of fortifications, whether on the sloping mountain foot or (more likely) some kind of refugeon the hill-top above, is perfectly feasible. A defence around the inhabited southern terraces of M. Leucio is also thinkable, whether or not doubled by a small one on the summit. Graves of 8th -6 th century date at S. Giorgio show that some settlements had already moved out into more open land. The obvious attraction here was the copious sorgente,

comparable with the one that feeds the Gari at Casinum. Pottery from M. S. Silvestro and the settlement below it, close to the other sorgente at Capo d' Acqua, also suggest occupation before the 6th century, and at S. Scolastica and the spur of M. Aglietta continuity from the Bronze Age is likely. At M. Leucio, a fragment of heavy bucchero (the only one found from a settlement site) is datable to the 7th century. The bulk of the pottery from Iron Age sites is rough hand-made poorly-fired impasto ware, which readily breaks into small, undiagnostic fragments. It does not correlate precisely with the better quality material (some in buccheroid style, though not including actual bucchero) from the Casinum cemetery. A date in the 8th-6th centuries is indicated for most of it.

If this is indeed the case, then there were a considerable number of open, or relatively open, settlements by the 6th century, although with a markedly uneven distribution. Many still hug the foot, or the lowest slopes, of the northern mountains, and one small, ill-defined site in the plain south of Piedimonte stands out as an exception. Only west of Roccasecca, in the area traversed by the Rio Moscosa, are there a number of sites on flatter land. Some of the foothill sites are well located for the use of the mountain slopes, or the higher mountain terraces, for pasture, and the lower colluvial or alluvial slopes for agriculture; two striking instances are sites 158 and 169, both situated on natural pathways to the areas of level land found at higher altitudes. In the Rio Moscosa valley, lighter soils were probably the attraction. The plain of Aquino, with its heavier clayey soils, has no certain Iron Age sites, nor has the dissected area northeast of Interamna. It is harder to make pronouncements on the south side of the valley, where work was less systematic, and the only sites known are Morroni, S. Giorgio and M. Leucio. Sites vary in size, but if the site at Capo d' Acqua be taken as a yardstick, only a few are of comparable importance 85 in the Rio Moscosa valley, 129 and 130 below Capri le, perhaps 169 above Piedimonte. These look like small nucleated settlements, whereas most of the others along the northern edge could be single homesteads. M. Leucio and S. Giorgio were also large sites. A reasonable, if hypothetical, reconstruction is to see several interrelated changes taking place in the early Iron Age up until the 6th century. One is simply a rise in the population, bringing with it the need for new settlements and an increase in the agricultural base - hence the higher number of sites, and the gradual creeping out into open territory. We have no means of knowing how common iron was in the Liri valley in this period, but the discovery of slag on the terraces of M. Leucio shows that the metal was worked locally, whatever the source of the ore (very possibly the M. Meta where the Melfa rises, but some small, more local source is not to be ruled out). The technology of primitive iron-smelting can demand the cutting of large quantities of timber or brushwood; thus the manufacture of iron demands clearances of woodland

14

which iron implements can in tum accelerate. The bringing of new areas under agriculture is thereby encouraged, and the new areas can support a further population growth. It was then probably this period which saw the first indications of radical change in the relationship between · man and environment. For the first time, changes in the landscape were not the slow, longterm products of climate changes and modified fauna! life, but were engineered by the human occupants. The ground is rather less secure in the ensuing period down to the Roman conquest. The problem stems from the almost complete absence of pottery types datable to the 6 th -4th centuries. Some signs of influence either from Campania or from northern Lazio or Etruria might logically have been expected: these would take the form of wheel-made pottery, south-Italian Greek pottery (or imports from Greece itself), early forms of black glaze ware or late forms of bucchero. The rough hand-made pottery previously mentioned is however not found associated with these finer types, and it is not even clear whether the Casinum cemetery continues without interruption, for there is also an absence of the forms of fibulae and other implements that might be expected. When wheel-made coarse wares in the Iron Age tradition are finally found, they are associated with other wares that take us into the late 4th or early 3rd centuries, and are usually found on sites that saw occupation of the Roman period so that they appear to represent a transition period belonging to the conquest or just after, when coarse pottery was still being made in the old indigenous tradition. Two exceptions to the general rule are of interest. The Iron Age site at S. Giorgio produced some wheel-turned wares in Iron Age tradition, but no imported wares, and it would be reasonable to see it coming to an end with the founding of nearby Interamna in 312. One rim fragment of the 5th century Attic (or Atticisi.ng) cup was found in a northern part of the city of Aquinum, in an area which also produced Iron Age sherds. This area, though within the later Roman walls, lies just to the south of the extramural area known as Mefete, where earlier discoveries suggested an important sanctuary area, and finds included hand-made Iron Age pottery and coins of Cumae and Neapolis attributable to the 5th and 5 th /4 th centuries (a quantity of 4 th and 3rd century pottery was also found in this area). It is then possible that occupation of the site of Aquinum started, or markedly increased, in the otherwise dim period of the 5th century. Two solutions (not necessarily exclusive) to the problems of this period can be offered, but neither suggests that the valley was in the forefront of developments, or even keeping up with adjacent regions. Rough, hand-made pottery may simply have continued to be the main form available, in which case there is little possibility at present of distinguishing sites dating to· the ~-6 th century from those that are a little later. Many of them may simply have continued, or shifted location slightly. The lack of contact with the outside world is noteworthy, suggesting

either a general poverty or that political conditions discouraged exchange (or both). The alternative is to suppose that quite extensive changes in the settlement pattern took place, and that many of the minor sites died out during this period. This cannot be proven, but there are good arguments for supposing that it was this period which saw a process of selection at work which led to the formation of the larger fortified sites mentioned earlier, and dispersed settlement may have given way to a more nucleated form. The problems of dating the stone fortifications must therefore be addressed. They are constructed in the technique known as "polygonal", since the blocks are not squared, even when quite carefully dressed. Essentially the materials are the rough limestone boulders which lie naturally to hand because of the geological formation of the hills. In the roughest style, sometimes called "cyclopean", they are put together in dry-stone manner, without (or virtually without) any dressing. More refined styles have the stone being first roughly, or even quite carefully, trimmed to ensure a good fit, all without mortar: the finest examples are works of no little virtuosity, with stones of varying shapes and sizes fitting very precisely along all faces. The stylistic labels I, II and III are usually given to . the cyclopean, rough and smooth types respectively, and a 4 th style is also distinguished in which the stones have a trapezoidal shape, but are laid in courses. Fortifications of this kind are common throughout southern Lazio and the Abruzzi - the territories of Latins, Hemici, Aurunci, Volsci and Samnites. In northern Lazio and southern Etruria fortifications built of squared tufa blocks are found, while some native sites in Lucania have fine ashlar walls that show obvious Greek influence. It is now widely accepted that the groups of walls are broadly contemporary, and that the differences of technique are due to the differences in the materials to hand as well as to different building traditions. Previous attempts to push the polygonal walls back into the Bronze Age and to see them as testifying to Mycenaean influence are now largely abandoned. The hardness of the local limestone, and its natural tendency to fracture into roughly polygonal blocks, are seen as sufficient explanation for the styles of middle Italy. The dating is now seen to lie within a bracket of 7th to late 4th or early 3 rd centuries, though some later repairs took place. Within this still rather wide bracket, there are two ways of trying for greater precision without having direct archaeological dating evidence from excavations. The first assumes that the various styles, from rough to smooth, correspond to chronological divisions, and for long this was the accepted method of procedure. Faith in it was shaken when a number of rough walls, which should on this principle have belonged to the 7th or at latest the 6th century, failed to produce evidence of so early a date. The hills of Montefortino above modem Artena (which may or may not be the ancient site of that name) are a case in point. The site on the northern tip of M. Lepini guards

15

the northern entrance to the Sacco valley corridor, looking across to Praeneste (Palestrina) on the other side. It thus enjoyed a strategic position, and had either habitation or a sanctuary of 6th century date. The walls, of Cyclopean or 1st style, might be thought to go with this, but excavation showed that one part of them at least could not be earlier than the late 4 th or early 3rd century, and there are no signs of re-building or a double circuit. Other fortifications in the heart of Samnite territory are thought to belong to the period of the wars with Rome in the 4th century. Nonetheless, while the rough manner of building may have persevered (especially for inaccessible sites), it may still be generally true that the more polished styles belong to a slightly later date, and are unlikely to be found before the 5th or 4 th centuries. The use of size as a criterion, albeit a rough one, has only just become possible with recent discoveries in northern Lazio. Absolute distinctions cannot be made, but there seems to be a tendency for smaller sites to be replaced or overtaken by larger ones (of 20 ha or more) in 5th and 4th centuries (the 300 ha of the Servian wall at Rome remains extraordinary even for the 4 th century, much less the possible 6 th century date). Most fortifications in the SaccoLiri valley (Artena, Segni, Ferentino, Alatri, Sora, Arpino, M. Nero, perhaps Rocca d'Arce, Cassino, .S. Vittore) fall into this larger class, as do those on the seaward side of the M. Lepini (such as Cora, Norba, and Sezze though it is smaller). M. Nero (some 30 ha) and S. Vittore (some 70 ha) have rough walls of style I. The walls of Cassino include both an acropolis at the top of the hill (site of the later monastery) and an area of sloping ground on the lower flank, coming near to but apparently not including the Gari with its sorgente. The area is some 75 ha, the style of the walls II, like the greater part of the walls at Arpino (and also Atina). Neither the extent nor the style of the walls (almost entirely destroyed) at Rocca d' Arce is clear, but it is likely that, as at Cassino, they enclosed a citadel (Rocca) and some lower ground (Arce). A further fortification may exist to the south-west of Boville Emica, on Monte del Fico, west of the Liri and almost due north of Monte Nero. A stretch of walls here is in the 3rd style tending towards the 4th , and enclosed an unknown area. The site, to judge from a local place name, might be the Luca mentioned by Livy, in which case Monte Nero may be the Volscian Satricum. A date subsequent to the Roman consequent can be ruled out with some confidence, since S. Vittore, Monte del Fico and Monte Nero were not communities that continued under the Romans: indeed, the first is so inconveniently situated that it may always have been more of a refuge than a defended commuhity, and this might also be true of the last. The degree of organization required for the construction of such massive walls nevertheless shows them to be the work of groups with a certain political cohesion. Effectively, they must belong either to the period of Volscian domination or to that of the Samnites. The latter, it may be felt, was almost too brief and certainly it demands a late (mid-4 th century) date for the

walls, whereas the arguments for seeing at least some of the Sacco valley and Monte Lepini and Monti Aurunci series as earlier have some weight. On balance, then, a dating within the period of Volscian domination seems justified. The organized communiti es responsible were then the outcome of the Volscians establishing themselves in the area, perhaps as a military elite. Some changes in settlement pattern might well be involved, and the valley could well have become less open to passage, more aloof from the outside world, than it had previo usly been. It is to be noted that the infilt ration of the Volsci in the 6th century ante-dated by only two or three generations the period of decline in Etruscan influence in Campania, so that the valley may have become more self-contained because of more than one factor.

It was probably the wars between Latins and Volscians that led to the more northerly series of fortifications, and the ones in the lower Liri valley could well be a little later, belonging to the 5 th and early 4 th centuries. One or two might be later still; the style of the wall at Monte del Fico suggests this, and the site of S. Vittore, with its remarkable view commanding most closely one of the passages into Samnite territory, would make best strategic sense in the context of Samnite expansion, or even the wars between Samnites and Romans.

The question remains of whether the fortifica tions are to be regarded as the efforts of individual communities, each protecting itself against its neighbour as well as any more distant foe, or whether they form a system. The degree of intervisibility between the sites has often received comment, though this could be seen as the accidenta l result of the choice of commanding positions. The most striking instance is perhaps Arpino and Rocca d' Arce, the summits of which are intervisible across the interve ning hills; Arpino and Sora are also intervis ible, as are some of the sites further up the Sacco valley. There is no doubt that a system of signalling, by fire or smoke, would have transmitted simple news, indicating the need to muster armies down from the Alban Hills to Rocca d' Arce in a very short time. South of Rocca d' Arce the situation is less clear, for the sites along the M . Cairo massif are not intervisible, and relaying from the plain (Aquinum?) would be necessary. M. Cassino and S. Vittore are intervisible, and each is also with M. Trocchio and Morroni, which however by this time may not have been occupied. While the intervisibility of some sites may indeed have been exploi ted for signalling, this does not necessarily mean that the sites were systematically chosen for this purpose. For that, some form of supra-local political organization would have been required, and there is no evidence that this existed among the Volscian cities. Admittedly, we are here dependent on Roman sources, but these show quite clearly the existence of a league among the Samnites, while the Volscian cities are shown acting in a more piecemeal fashion. The question of a communications system must therefore remain open.

16

Be that as it may, Volscian commuruttes did develop regular forms of civic life and administration, even if we have no direct evidence for this from the lower Liri valley . Inscripti ons from Velitra e and Antinum show that there were Volscian magistrates called meddices (the same word as was used by the S~unnites), and that in rare instances the title was kept even after Roman conquest. While the 5th and 4 th centuries may have been in some respect a dark and troubled period, they saw eventually the development of civic organization in the Liri valley. Along with this may have gone an increased nucleation of settlement, the abandonment of certain sites in favour of others. Even some of the potentially defensible sites may have been abandoned , perhaps as a result of warfare or political developme nts. On the other hand , it is likely that agriculture increased in the vicinity of the favoured, larger settlements. These now gained their circles of mighty walls, which must have been a prominent feature in the landscape. One major new site development which, so far as is known, had no defenses, is Aquinum, where the evidence suggests an important sanctuary which went on in use throughout the Roman republ ican period . Just why this site on the plain, probably still wooded, above one of the lakes formed by a small stream, was sacred, can only be guessed. The name Mef ete suggests a connection with Mefitis, an Italic goddess connected with life-giving supplies of water. Later, Aquinum was to be a city in its own right , but perhaps it started life as a sacred spot where the various communities from the valley could meet for the common celebration of sacred rites , and enjoy the cessation of hostilitie s which such circumstances can bring.

The Etruscans in the Liri Valley P. Hemphill The Liri Valley lies between the cities of Etruria proper and the Etruscan cities of Campania. At first glance the valley would appear to have provided the same land route as it has in later periods. However, archaeological and histor ical evidence suggests that this was not so. No evidence presently exists for Etru scan settlements as opposed to some vaguer Etru scan influences, north of the Voltumo and into the Liri Valley . Archaeological evidence for Etruscan sites in South Etruria, even such small sites as tombs and huts, consistent ly shows a wide variety of contemporary materials. Such concentrations have not been found in the survey of the Liri Valley. This rules out colonization, and with it any immediate politi cal domination.

In considering the Etrusca ns one must never lose sight of the fact that they were centered on the large cities, each of which maintained a sturdy indepen dence. In addition:, it should be remembered that_, from their original development through to their falling under the Roman hegemony, the power and wealth of these cities varied considerably. At no point from the 8th century into the 2rd century B.C. can we speak of the Etruscan s as a cohesive

group either in military or trade terms. Territorial expansion or the creation of an Empire does not appear to have been a part of their political organization. Wealth and power were based on overseas or inter-city trade and already by the 8th century B.C. such places as Vetulonia, Tarquinia and Vulci[a] had far-ranging shipping fleets. By the 7th and 6 th centuries B. C., the Etruscans had fleets throughout the eastern and central Mediterranean, successfully challenging the Greeks and the Carthaginians. Etruscan influences in Campania and Latium appear to have reached a high point in the 7 th and 6th centuries B.C. which coincided with the high point of greatest wealth and sea-power in Etruria. It is reasonable to suppose that the Etruscan support in the development of the Campanian towns was one of trade expansion. An example of this is the evidence of a colony of Vulcian metal workers in Capua. Once the Etruscans had lost their foothold in Latium after the beginning of the 5th century B.C. and lost naval supremacy, especially after the defeat by the Syracusan naval forces in 474 B.C., the Etruscan influence in Campania decreased dramatically. With the descent of the Samnites on Capua in 430 B.C. the Etruscans were no longer able to protect their interests. Loss of sea power meant loss of the Campanian connection. All of which suggests that a sea-borne trade, maintained and even enforced by sea power, was the basis of the relationship between the major Etruscan cities and Campania. Therefore, since this connection was commercial rather than political it must have been dependent on large scale movement of goods. Ships and ports existed and to think in terms of a trade carried on overland is ludicrous. To be sure it was no doubt possible for a determined person to travel overland from, for example, Palestrina to Capua. But there is no reason to suppose anything other than woodland paths and thus, even on horseback, this journey of almost 200 km. would probably have taken a week. It is unreasonable even to imagine an overland route for anything larger than small scale forays--and then only from inland towns. It appears, therefore, from lack of any archaeological evidence, from the turbulent history of the valley before the Roman settlement and from the essentially trade based connection between Campania and the Etruscan cities that the Etruscans did not extensively use the Liri Valley as an overland route. Yet the discovery of Etruscan bucchero pottery at a number of sites shows that there was some overland contact. If these are plotted, it appears that the middle Liri valley and upper Melfa valley have more such sites than the lower Liri, where only two or three fragments of bucchero have come to light, and pottery from graves at Cassino, while it shows the influence of bucchero, does not include imports. This pattern may indicate that.

17

Chapter 5

THE LIRI VALLEY: LANDSCAPE AND PEOPLE IN ROMAN LITERATURE A.G. McKay

Salve, magna parens frugum, Saturnia tellus magna virum (Hail, land of Saturn, great mother of earth's fruits, great mother of men! (Georgics 2, 173-174))

How did Roman Republican and Imperial writers, masters of poetry and verse, scholars and creative writers, view the ancient territory of the Liri valley and its environs? How realistic were their reconstructions of time past? How reliable were their conjurations of antique styles and living and morality? Particular authors suggest themselves right away: Cicero, Livy, Varro, and Vergil above all. Vergil's panegyric of Italy and of the Italic farmer's life, and his account of Volsician resistance to Trojan land claims in Italy, are important references for his contemporaries' views on the territory and the people of ancient Latium , the land where the Liri and the Garigliano flow today. Vergil's poetic treatise on agriculture, Dryden's "best poem of the best poet", probably the most difficult, certainly the most controversial poem in Roman literature, can now be consulted in two distinguished new editions by the late Sir Roger Mynors and by Richard F. Thomas . 1 Thomas, in particular, has assessed details of the "Praises of Italy" excursus with exceptional insight, suggesting that many of its details, when scrutinized against the ethnographical tradition from which the laudes derive, indicate that the Italian landscape had been moulded and altered from its wonderful natural state by man's engineering skills and constructive instincts. He has argued that on a suggestive level, and in a complex manner, man's labour in the age of Jupiter does not simply succeed or fail; for Vergil, success normally has a price. 2 Certainly Vergil's topographical outline of productive Italy in the laudes ltaliae ( Georgi cs 2, 136-176) marks a welcome departure from more earthbound topics elsewhere in the collection. The unsuspecting reader is transported abruptly from a didactic world of terrestrial specifics into a hymnic litany of the Italian land and its people. Comparable excitement and patriotic fervour race through the "Catalogue of the Italian Forces" later in Aeneid 7, 641-817. For Vergil, Italy is eternally primaveral, endlessly productive and constantly renewing itself. The landscape painting is expansive with "noble cities, the achievement of man's toil, all the towns his handiwork has piled high on steep crags, and rivers that glide beneath those ancient walls" (Georgics 2, 155-157).3 Equally germane to our topic, Vergil extols the people of Italy: "She has mothered a vigorous breed of men, Marsians and the Sabine stock, the Ligurian, inured to hardship, and Volscians with their short darts" (Georgics 2, 167-169). 4 Elsewhere, the poet waxes rhapsodic over the blessings of 18

the happy farmer's bountiful, lotus-land existence: "Theirs is repose without care, and a life that knows no deceit, but is rich in treasures manifold. The ease of broad acreage, 5 caverns, and living lakes, and cool valleys, the lowing of cattle, and soft slumbers beneath the trees - all are theirs . They have woodland glades and the haunts of game; a youth hardened to toil and inured to scanty fare; worship of gods and reverence for age; among them, as she quieted the earth, Justice planted her last steps" (Georgics 2, 467474). To be sure, the text of the poet's eulogy of the "peaceable kingdom" of yore seems to sour somewhat with the advent of Rome and Caesar and the horror of civil war . But even during the civil strife of recent history (133-30 B.C.), tumultuous times when Vergil composed his Bucolics and Georgi cs, there still lingered the wishfulfilment pattern of bygone days when peace, simplicity and fruitfulness graced the peninsula. Although contemporary Rome had succumbed to luxury and vice and criminality, Vergil (and Augustus) remained alert to the enormous potential ofltaly for good or ill. _For Vergil, the farmer, the ploughman especially, is a marvellous paradigm of commitment to a rigorous existence in order to control, tame and civilize nature, to ensure her bountiful fruits, and to provide a livelihood. Of course, contradictory, subversive forces of disorder, repression and violence were never far distant during these years and they are hardly suppressed in Vergil's panegyrics of Italy and the farmer's life. However, the inherent, natural powers of Italy managed to survive, gamely and persistently, in the hills and valleys of Italy, a troubled but vestigial paradise. Vergil's message is clearly enunciated at Georgics 1, 145-146: labor omnia vincitlimprobus et duris urgens ill rebus egestas ("Unrelenting toil has conquered everything, and need, pressing hard on men through adverse times"). Suffering and need march alongside the vision of Italy as an Eden, where Justice dwelt with farmers before she finally decamped. 6 Ambiguities certain!y abound in Vergil ' s encomia ofltal y, but they are a constant hallmark of his writing career; the oscillation between optimism and pessimism is implicit in all his works. Vergil clearly associated himself with an antiquarian tradition that extolled the virtues of the past. Elsewhere (Aeneid 9, 598-620) , Vergil contrives to have a representative of the defensive Rutulians , Numanus Remulus, Tumus' brother-in-law, voice insults against Aeneas' son, Ascanius-Julus and against the Trojans as a whole .. .digna atque indigna relatu/vociferans (595t),

"things that were both proper and improper to utter." Rutulian (and Italic ) virtues and careers are extolled; tough, frugal, chaste, expert in hunting, riding archery, farming, brigandage and war. The Trojans are their opposite: they are mocked for their decadent way of life, their effeminate clothing and debased morality .7 Ascanius' prayer for assistance against the traducer is answered; Jupiter thunders on the left in a clear sky and Ascanius' arrow pierces Remulus' temples; Apollo congratulates the youth and cautions him against foolhardy ventures (630656). Trojan valour and manliness are acknowledged in heaven and on earth. Here, as earlier in the Georgics, Vergil purports to sketch an account of life, heroic life, in early Italy. Nationalistic antiquarianism was in the wind during his writing career and ethnography was no stranger to Roman intellectuals. The Rutulian folk who served under Italy's defensive hero, Tumus of Ardea, are proud embodiments of a hardy, independent people spawned by an ancient land. But Vergil's Rutulians depart from the ethnographical norm as Remulus describes them in his vituperations against the Trojans; for they are also disciplined folk, who work the land, endure the toil, and are accustomed to do without: at patiens operum parvoque adsueta iuventus/aut rastris terram domat aut quatit oppida hello (Aeneid 9, 607-608). Vergil's picture of the Rutulians conflicts with the conception of the true primitive: their lives are regulated by the agricultural activities demanded by the seasons. The Rutulians of Latium are not unique in Vergil's portrait of rock-ribbed farmer-soldiers. When Vergil provides a catalogue of the Italic forces marshalled against the Trojan "invasion" forces, all are hardy rustics (duri agrestes, Georgics 1, 160). The activities of the Aequi in that context are the most detailed and their regimen is remarkably similar to that of Numanus' Rutulians; neighbours of the Vosci, the Aequi are addicted to hunting and field work; they work their properties under arms, rejoice in booty, and support themselves by their pillaging (Aeneid 7, 746-749). Clearly, Vergil and his contemporaries imagined that this was the modus operandi of the early Italians. His likeliest source of such views and for relevant details was M. Terentius Varro, a native of Sabine Reate and most redoubtable scholar of his time (116-27 B.C.). 8 Quintilian called him eruditissimus (Inst. Orat. 10, 1, 95). His De Aboriginibus et Latinis must have been a major resource for Vergil's inquiry into Italian topographical questions, and present-day historians are generally convinced that Varro was the chief agent in designing and propagating the nostalgic vision of the Roman past that Augustan Rome entertained. Cicero was numbered among those who were impressed by Varro's researches. One might therefore expect that as Rome's first official librarian Varro would also be spokesman for nationalistic antiquarianism. Cato the Eider's Origines and De Agri Cultura probably inspired Varro in his own reconstruction of the past in his De Gente Populi Romani 9• Vergil's regard for Varro and his 19

indebtedness to him is a paramount ingredient of his Georgics and, periodically, of his Aeneid. Horace, his poet friend and soul-mate, shared the same idealized vision of a Roman past where high valour and clean living were a familiar pattern, when Rome's soldiers were drawn from small-scale Sabine farms to which they returned after the necessity of military campaigns. Livy's History also offered a parade of heroes whose lives were models of asceticism and patriotic devotion - Camillus, Cincinnatus, Manius Curius Dentatus, and M. Atilius Regulus, to name only several in the Heldenschau. 10 For writers of the late Republic and Principate, hunting, riding, archery, soldiering and farming were inherited and important aspects of Rome's past. They were stereotype elements in the careers of doughty men of a primitive era and they were certainly part of the "moral rearmament" program of Augustus. 11 Their impact on life in the Liri Valley, among the Volscians and their neighbours, is never in doubt. Toughness, staying power, and military prowess were unquestionably the distinguishing traits of the historical Volscians who embarrassed Rome repeatedly over the years with their military prowess. Livy's comment on the Volscians is grudging token of their pugnacious nature: ferocior ad rebellandum quam ad bellandum gens (7, 27, 7). 12 Vergil's Eleventh Aeneid is by and large a book of funerals, debates, and warfare. Visible space is an important ingredient of the epic as a whole, and Vergil's depiction of landscape in Aeneid 11 is noteworthy. Throughout his account of the cavalry engagement at Latinus' Lauren tine City, Vergil creates a brilliant psychological and atmospheric background by his orchestration of close and remote perspectives, shifting from mass to individual views in a cinematic fashion. It is also the book where Vergil introduces for a second time (earlier at Aeneid 1, 803-817) the greatest personality in the Volscian host, Camilla of Privemum, royal daughter of Metabus and Etruscan (?) Casmilla. 13 Camilla is the central figure in the military action of Aeneid 11, a maiden warrior (bellatrix ... virgo, 7, 805), a veritable Jeanne d'Arc, whose beauty, valour and speed are legendary in her own time (7, 808-811). Her background is as dramatic as her career on· the battlefield. She is daughter of Metabus, arrogant despot of Volscian Privemum, who has been expelled by his indignant subjects (Livy comments on Privemum's love of freedom: 8, 21, 2-10) who are pictured in full pursuit of their deposed ruler and his infant daughter, Camilla. Cato, Origine 62P.2, repeated in Servius' note on the Camilla episode, confirmed that Volscian territory was occupied by Etruscans at some stage but archaeology has not confirmed their presence for long. It is conceivable (but not verifiable) that Metabus was an Etruscan sympathizer, maybe a Quisling, but his expulsion suspiciously also parallels the liberation of Rome from Tarquinius Superbus in 509 B.C. Metabus properly associates with Metapontum, near Taranto, not with Privemum, serving notice, perhaps, that Metabus was Vergil's import to

Privernum; conceivably Camilla was yet another Vergilian "creation". The Amasenus river, swelling to its banks from a protracted rainfall, made crossing by ford, ferry or bridge impossible. Metabus thereupon devotes to his child to Diana, wraps her in cork bark, and ties the precious parcel to his spear which he sends whistling across the turbulent river. Thereafter he swims to safety and recovers his child for a great destiny. Nurtured by a wild brood mare, and raised by her father among mountaindwelling shepherds, Camilla is trained even as a toddler in the Italic art of hunt and, like a junior Amazon, serves as a member of the sacred band or thiasos of Diana (Nemorensis?). Her adolescent charms excite the admiration of Etruscan matrons in the region who long to have her as daughter-in-law despite her devotion to her saviour goddess Diana. When Tumus leads the resistance against Aeneas' Trojans and their allies, Camilla, a Volscian queen, inexplicably restored to her native city of Privemum, offers strategic cavalry support to the Italic cause while Tumus, supreme commander, organises an ambush for the Trojans. Diana regrets her retreat from her service to practice the martial arts and sees her defection as an unhappy prelude to her death on the battlefield. After she has killed twelve of the enemy, Camilla abandons caution in her pursuit of an "ornamental " foe, colourful Chloreus, one-time priest of Cybele and fluttery embodiment of effeminate practices. During her pursuit of her quarry, she falls victim to the attack of Arruns, devotee of Apollo of Soracte. Diana fulfils her promise and sends her nymph Opis as an avenging "angel" to kill Arruns and to convey Camilla's corpse to heroic interment at Privemum. Volscian Camilla and Carthaginian Dido are both conceived along startlingly similar lines and their tragic careers and deaths are equally memorable. For our purposes, Vergil is the premier voice on the Volscian landscape and people whose history is interwoven with the Liri river valley and its environs. Regardless of whether he knew the region from autopsy or not (he certainly traversed some of it, once in company with Horace and Maecenas in 37 B.C.),1 4 Vergil's portrayal of the land of the "people of the marsh" is noteworthy. From the literary standpoint, his realistic description of the Amasenus in high flood during the rainy season, serves both as token of present political uproar involving Metabus and as harbinger of the struggles that will mark Camilla's last hours on the battlefield. The river rises today at S. Lorenzo di Amaseno to the southwest of modem Frosinone, and after a 43 km course debouches into the Tyrrhenian Sea west of Terracina where it forms the eastern boundary of Privemate territory. 15 The reality that greets the traveller in the Liri Valley today, and no doubt in antiquity, aligns remarkably well with the memorials of Italy's past that Vergil extols generally and retrospectively in the epic's laudes Italiae and elsewhere. The ancient hill-towns are impressive even in their ruins: Antium, with its polygonal walls and antique gate; the sanctuary of Jupiter Anxur at Terracina whose

20

defensive system was practically impregnable; Arpinum, home to Cicero, Marius and Agrippa; the acropolis fortifications of Atina at the source of the Melpis (mod. Melfa) river, with primitive rough-cut boulders piled three metres high; Cora's upper and lower town defenses; and Sora's 540 m promontory and its girdle of polygonal walls. The hill-top settlements and fortifications of Latium, piled up by human energy, were and are still an impressive feature of the terrain and Vergil's Volscians "armed with darts" outside their fort were even then recognized as skilled in warfare, a fact that Rome learned to her discomfort through repeated encounters in the field during the 5th and 4th centuries B.C. 16 Vergil had an obvious respect and admiration for the Volscians, who had been stubborn, aggressive foes of Rome in the historic past. But Camilla holds the limelight, like some legendary Amazon, a veritable Harpalyce or Penthesilea, who also embodies the virtues and flaws of her Volscian people, later masters of the river valley, the citadel heights, and the coastal havens. Volscian placenames are notably sparse in Vergil's war-time report save for Privemum, some 40 miles south east of Rome, where the Volscian settlement atop an eminence remains largely unexcavated. After its alleged tyrannical phase under Metabus, a story that Vergil uniquely supplies, perhaps from some commemorative story, perhaps from his poetic imagination after the experience of Tarquinius Superbus at Rome, Privemum was repeatedly inimical to Rome in 358357, 342-341, and 330-329 B.C. Finally victorious, Rome destroyed the fortifications of Privemum and confiscated two-thirds of its territory (Livy 8, 1, 3; 8, 20, 7). Although the original settlement was probably planted on the heights near modem Roccasecca dei Volsci, when Privemum was refounded as a Roman colony with praetores duoviri (ca. 150-100 B.C.) it started its new existence on the plain (Piane di Mezzagosto) and played the role of a way station between the Sacco River valley and the port of Terracina. So, during Vergil's time, "new" Privemum was a reasonably prosperous centre. 17 Its town plan resembled that of Aquinum and Saepinum. Coarelli actually identified a Republic domus with the "villa" property of M. Junius Brutus, aedile in 146 B.C. (Cicero, Pro Cluemio 51, 141). 18 An urban renewal programme undertaken there during Imperial times heightened the visibility and local pretensions of Privemum. Although the chief cult figure among the Volscians was probably Mater Matuta ( = Juno Lucina), with her major sanctuary at Satricum (Boville Emica), Camilla's life of service to Diana as her huntress votary fits in reasonably well with cult practices in Volscian -and adjoining territories where Diana Tifata was commonly worshipped. 19 The Amazonian image, and the mountain retreat of the "Maid" in Vergil's account, may even embody aspects of the cult of Diana of Aricia on Lake Nemi, haunt of the Golden Bough and sinister priest-king rituals. 20 Vergil's consignment of Camilla and her father to the lonely hills where shepherds dwell may be his

sympathetic recreation of early Volscian "economy" with its practice of transhumance and its pastoral life, both features that persisted into his time (and today). From a military standpoint, Camilla commands a Volscian cavalry squadron but infantry are also represented on two occasions. Once, during an assault on Aeneas' camp at the mouth of the Tiber, when the Volscians form an anachronistic testudo and prepare to fill the defensive trenches and demolish the Trojan palisade (Aeneid 9, 505506), and again with Volusus (an eponymous association with the Volsci ?) as commander, during the action at the city of the Laurentines (11, 463). Camilla's story resonates sometimes with Etruscan associations. Certainly the name of her native city, Privemum (Privemo) has an Etruscan ring to it, and her purple mantle and gold fibula (7, 814-816) befit an Etruscan princess. She is a glamorous figure for Etruscan matrons who marvel at her nubile beauty (11, 581-582). Prone to elaborate finery, after the Amazonian, perhaps barbaric fashion, Camilla is ultimately seduced into carelessness for her safety by the outlandish clothes of Chloreus, emasculated priest of Cybele, the exotic Mother Goddess who was patroness of Troy's Mount Ida, but resident in Vergil's time on Rome's Palatine Hill. Vergil's "warrior maid" is clearly a composite figure, of mythology and folklore. Whether or not Vergil meant his readers to associate her name simply with the title camilla, a ritual attendant in cult service of a female deity, or with heroic Camillus, twice Dictator of Rome, victor over the Etruscans at Veii in 396 B.C., exiled to Ardea after his victory over the Gauls at Pesaurum, and recalled finally to be the saviour of Rome from the Gauls in 390 B.C., we shall probably never know for certain. But Camillus did score successes against the Volsci, Hemici and Etruscans in 386 B. C. and again in separate conflict with the Volscians in 381, and he did celebrate a triumph over the Volscians, Aequi and Etruscans (Diod. Sic. 14, 117, 6). He died during a plague in 365 B.C. 22 If a female personage is required for the poet's recreation, perhaps Cloelia or some other rebellious female in Roman history will serve - Fulvia, perhaps, Antony's wife, a keen adversary of Octavian in the Perusine War? The true source or chief association finally remains an enigma. The fulfilled promise of her final inhumation (a Volscian practice) at Privemum no doubt deliberately echoes the tumulus burial of Homer's Sarpedon in his native Lycia (Iliad 16, 667ff.). The tragic story of that beloved son of Zeus certainly colours Camilla's behaviour and destiny, but her faults are uniquely those of Vergilian youthful characters: violentia, audacia, and amor habendi. Her story also exhibits traits that are common to the legend of Romulus and Remus, nursed by a female wolf and raised by shepherds in a woodland setting to be trackers of wild beasts, denizens of a woodland asylum for misfits and revolutionaries; even her corpse, enshrouded by a cloud (nube cava, 11, 593), finds a Romulean parallel (denso nimbo: Livy 1, 16, 1), and both have assurances of

21

immortal fame (dia Camilla, Aeneid 11, 657; Livy 1, 16, 8). The survival of an antique "tumulus" beneath the early Volscian citadel of Privernum might also have been a contributing factor to Vergil's "resurrection" of a warrior princess whose record of gallant, unconventional service to a lost cause, forecasts other struggles, equally brutal and bloody, when Rome brought the Volscians under their control. 23 Vergil's notices of the Volscians (and their Samnite successors are implied) offer an intriguing commentary on their nature and habitat. 24 Their territory was presumably never in doubt for Vergil: it included the valleys of the Liri, Sacco and Melfa, with Sora, Arpinum, Frusinum, Fabrateria Vetus, Ecetra, and Atina, and Privernum in the Amazenus valley; Antium, Satricum, Velletri, Terracina (Volscian Anxur) and the Pomptine marshes, all sites of historic importance, are curiously excluded ·from his roster of Volscian places. The western Volsci on the coast of Latium are a specific reference; the eastern Volsci, in the valley of the middle Liris, for Vergil's uses are a generalised entity contained in the larger tribal unit. Specific sites are omitted; so too soils and crops, cereals, pastures and fodder, animal husbandry, livestock, horticulture and arboriculture, and aviaries; there is no precise allusion to any of them in Vergil's Volscian record. Cato and Varro offer some details along these lines, but their references are few and far between and no major importance can be attached to them. Cato's De Agri Cultura specifies particular products and markets in the Liri valley context; and Varro's De Re Rustica spends time on the description of the Isoletta at Casinum (modem Cassino), on the library and on the celebrated aviary. Cicero gratifyingly offers some surprises with emotional asides on preferred sites and some helpful topographical notes. For his brother's villa at S. Domenico di Sora and for his ancestral home at Arpinum the orator has particular words of praise .25 His remarks on the one-time Volscian territory of Arpinum, where he was born, refers to nearby Atina as well: "Neighbourliness is a quality that demands our commendation, even our love, since it nurtures the bygone spirit of kindliness, it is uncoloured by the sinister hues of petty spite, it lives in no atmosphere of falsehood, it is tricked out by no hypocritical pretensions, it is unschooled in that studied counterfeiting of emotions characteristic of the suburbs and even of the city" (Pro Plancio 8, 22). Granted that Cicero may be resorting to a measure of hyperbole and special pleading for his local client, Gnaeus Plancius, the tenor nevertheless seems genuine: "There was no one at Arpinum , at Sora, at Casinum, or at Aquinum, who was not Plancius' adherent. The thickly-populated district of Venafrum and Allifae, in short, all our rugged countryside, contain in their hills hearts that are loyal and unaffected and staunchly true to the bond of kinship ... " (Pro Plancio, 22). For Mark Antony's rambunctious venture into the Liri Valley, Cicero has only contempt, coupled with a measure of apprehension. "You swept down on the farm at Casinum

of Marcus Varro, a most pious and honest man (Philippic 2, 40, 103). How many days did you most disgracefully carouse in that villa ... What an ill-matched owner! For Varro wished that house to be a retreat for his own studies, not for lust. What discussions formerly took place in that villa, what meditations! what thoughts were committed to writing: the laws of the Roman people, the memorials of antiquity, every system of philosophy and of learning. But in your tenancy - for you were no owner the whole place rang with the voices of drunken men; the pavements swam with wine; even the walls were drenched; free-born boys kept company with male prostitutes, harlots with mothers of families. Men came from Casinum to pay their respects, from Aquinum, from Interamna, but no one was admitted" (ibid., 104-105). Cicero's comments on the villa at Casinum as a writer's retreat suggests that Varro's ethnographical and topographical researches that were so germane to Vergil's epic were actually conducted in situ. 26 Aquinum received Antony with exaggerated deference out of dread: "After setting out for Rome from (Casinum), he approached Aquinum, a well populated community, and quite a large crowd came to meet him. But he was carried through the town in a closed litter like a corpse. Although the people of Aquinum acted foolishly, they did live on the highway" (ibid. 106). 27 · But Cicero's most heartfelt comments on the terrain are reserved for Arpinum, the family habitat, a "lovely and healthful spot", where he frequently refreshed himself after legal and political contents at Rome (De Leg. 2, 1, 3).28 His enthusiasm knows no bounds: "This is really my own fatherland, and that of my brother, for we are descended from a very ancient family of this district; here are our ancestral sacred rites and the origins of our race; here are many memorials of our forefathers .... You see our homestead there in its present condition - rebuilt and extended by my father's care; since he was an invalid, he spent most of his life in study here .... It was on this very spot ... that I was born, while my grandfather was alive and when the homestead, in keeping with the old custom, was small, like that of (Manius) Curius (Dentatus) in the Sabine country" (ibid., 2, 1, 3). For the island in the Liris tributary Cicero has nothing but praise: "That island is a favourite haunt of mine for meditation, writing and reading ... I cannot get enough of this place especially as I have come at this season of the year. I scorn luxurious country-places, marble walks and panelled ceilings. Deliver me from those artificial streams which some of our friends call "Niles" or "Euripi" - after seeing what we have before us, who would not laugh at them?" (De Leg. 2, 1, 2). And again: "Here we are on the island; surely nothing could be more lovely. It cuts the Fibrenus like the beak of a ship, and the stream, divided into two equal parts, bathes these hanks, flows swiftly past, and then comes quickly together again, leaving only enough space for a palaestra of moderate size. Then, immediately after accomplishing this, as if its only duty and function were to provide us with a seat for our discussion, it immediately plunges into the Liris and, as though it had entered a patrician family, loses its less famous name, and makes the water of the

22

Liris much colder. For, though I have visited many, I have never come upon a river which has colder" (ibid., 2, 2, 6). The palaestra, by definition an exercise ground, was probably designed for intellectual rather than gymnastic exercise. Varro' s references and effusions complement Cicero's. His remarks on the aviary at Casinum are noteworthy. "I own, near the town of Casinum, a stream, which runs through my villa, clear and deep, with a stone facing, 57 feet wide, and requiring bridges for passage from one side of the villa to the other; it is 950 feet in a straight line from the island in the lowest part of the stream, where another stream runs into it, to the upper part of the stream, where the Museum is situated." (De Re Rustica 3, 5, 910). The overall effect of Varro's account of his luxury villa with its museum and bibliotheca (3, 5, 9) is impressive (ibid., 3, 5, 11-17). Columella and Pliny the Elder refer less often, certainly less garrulously when they do, to the people and places of the Liri Valley. The most important eyes and ears on the area are undeniably those of Livy, Varro and Vergil, and their paramount importance emerges in the. references and quotations already cited. We have come a long way from Numanus Remulus, the Rutulian brother-in-law of Turnus and bete-noire of Ascanius-Julus. His programmatic description of the Italic way of life may certainly be viewed as Vergil's side-long statement on Augustan designs to reawaken respect for old values and institutions enshrined in Republican history and legendary past behaviour. Volscian Camilla, and Caeretan Mezentius, the exiled Etruscan of Aeneid 10, are two of Vergil's best heroic creations: both are types of archaic heroism whose highest ideal was spontaneous selffulfilment through hunting and warfare. Latinus, the aged, retreatist ruler of the Latins, and Evander, the elderly Arcadian ruler of Pallanteum at the site of future Rome, both suffer personal grief in their commitment to what the new order of things brings to them. The characterisations vary enough to show that Vergil had a profoundly sympathetic insight into the qualities and failings of the Italic people in time past and during more recent tumultuous times of civil war. Although the poet apparently had a firm faith in the moral leadership of Rome and although the Golden Age might be reborn through Augustus' inspiration and effort, nevertheless the Iron Age is always latent too, as the laud.es ltaliae and the "Praise of Country Life" (Georgics 2, 458-474) certainly suggest in his didactic poem. However, the people of the Valley, Volscian and Samnite, its swollen rivers and dark tracts of wild forests, are enduring; the land is deathless and the response of patriotic Italians to its glories and its terrors abides through time. Cicero and his brother at Arpinum and at Sora, and Varro at Casinum, are keenly alert to the productive and restorative aspects of the valley; their country estates and their ancestral pride in the space and its buildings are token of their respect for the countryside which was as impressive then as it remains

today. Our detailed information attaches to notables; but there was, as our field survey in the Liri valley demonstrated, a larger repertoire than landed estates. Salmon's verdict was that the Samnites, successors to the Volscians during the 4th century B. C., were primarily montani atque agrestes as Livy noted (9, 13, 7). 29 Villas have, somewhat contrary to expectation, not appeared to any significant sample in our surface exploration, even in the likeliest spots. But smaller establishments that have surfaced could certainly support grazing herds of respectable size behind their dwellings, and transhumance and short-term settlement must have been a more frequent resort in ancient times than we are told. 30 Cicero in his Pro Cluentio offers a vivid portrayal of life around Larinum during the l' t century B.C., when droveroads (tratturi, ancient calles) still crossed over the town's territory. Perhaps the same was true of the Liri valley. Vergil's reference to shepherd communities in the clearings on the uplands would be a normal complement to mixed farms and to what appears to be a fairly wide range of settlement in the valley. Arable fanning and stock.keeping associated with medium and small-scale farmsteads are not unlikely patterns in the Valley during the late Republic and Early Empire, and the road system supports the notion that markets flourished in the area. Once again, Cicero's Pro Cluentio suggests a specialised rural economy where the wealthier families own and manage land outside but still within the territories of the town. Estates (praedia), business interests (negotia) and stock (res pecuarias) must have been to some extent, as they are today, supporting ingredients and activities of Liri Valley market towns. So much we may derive from Cicero, Varro, Livy and Vergil on the nature and conditions and hoary past of the Liris Valley and its environs. But caution is required. Roman poets and historians are notoriously unreliable assessors and chroniclers of terrain, not only in Ilaly but throughout the empire. 31 Vergil owed his topographical and geographical knowledge largely to Varro's prose, and when that information is transmuted into verse, with the scrim curtain of epic time drawn over it, caveant topographers, archaeologists and historians! There are demonstrably "typical" places and "typical descriptions" in many instances, and illusion and indeterminacy are the hallmarks of poets and often of historiographers. But the land and the people are also part of the continuing system, and both literary artists and scholars may finally be expected to state what they regarded as plausible and verifiable data for their readers, and to record what they regarded as "archaeological" facts in an effort to make their work more plausible and more acceptable for their consumers. The idealisation of ancient Italy, the redressing of landscapes which were familiar, the portraits of "hard primitivism" with the traditional ideology of the Italian peasant and soldiers, and the reconstructions of a remote past in the works of Vergil and Livy and others, testify to the fact that conditions must have remained unchanged, to some degree authentic, into the Augustan

23

age. Vergil was certainly prepared to thrust back into the legendary period Varro' s, perhaps Cato's, reconstructions of archaic Regal and early Republican conditions. At any rate, revivals of the life-styles of antique Italic peoples like the Volscians, one-time masters of the Liri valley, in verse and prose add an important dimension to present-day archaeological ventures.

Notes 1.

2. 3.

4.

5.

6.

7.

8.

The commentaries: R.A.B. Mynors, ed., Virgil: Georgics. With a commentary by R.A.B. Mynors, and a preface by R.G.M. Nisbet. Oxford 1990; Richard F. Thomas, ed., Virgil: Georgics. 2 vols. Cambridge 1988. Thomas, op. cit., 20-21. Thomas appreciates the vignette of the Italian hillcities, with rivers gliding below, but argues for its being assessed in cultural terms. For contrasting estimates of the passages consult: A.G. McKay, "Vergil's Glorification of Italy (Georgics II, 136174)" in Cicero and Virgil: Studies in Honour of Harold Hunt (J. Martyn, ed.). · Sydney 1971, 149-168; and M.C.J. Putnam, "Italian Virgil and the Idea of Rome," in Janus: Essays in Ancient and Modern Studies (Louis L. Orlin, ed.), Ann Arbor, MI 1975, 171-203, with discussion at 203210. verutos: Mynors translates as "armed with javelins", Thomas as "equipped with the short dart". The descriptive epithet follows the ethnographical pattern. latis otia fund.is (468) may refer to extensive ranges and ranches, the so-called latifundia, a term rejected by K.D. White but certainly evoked in Varro's De Re Rustica C 16, 4. Youths engaged in hunting recur in the caustic sermon of Numanus Remulus on primitive Latin virtues in contrast to Trojan decadence (Aeneid 9, 605-607). Vergil's theodicy, "the mysterious ways" of Jupiter in his treatment of mankind, suggests that grim toil may be a necessity but that it does not overcome problems. Thomas (op. cit., 93) interprets the troublesome phase as "insatiable toil occupied all areas of existence." · For analysis and criticism of the Numanus Remulus episode and the diatribe consult: Matthew Dickie, "The Speech of Numanus Remulus (Aeneid 9, 598-620), "Papers of the Liverpool Seminar 5 (1985) 165-221; N. Horsfall, "Numanus Remulus: Ethnography and Propaganda in Aeneid ix, 598f, "Latomus 30 (1971) 1108-1116; and Richard F. Thomas, Lands and Peoples in Roman Poetry; The Ethnographical Tradition (PCPS Suppl. 7, 1982) 98-100. On Numanus Remulus, see N. Horsfall, "Numano Remulo," Enciclopedia Virgiliana 3: 778-779. On M. Terentius Varro consult, inter alios, N. Horsfall, "L'opera varroniana e l'Eneide,"

9.

10.

11.

12.

13.

Enciclopedia Virgiliana 5 (1990) 447-450; P. Fraccara, Studi Varroniani, de gente populi Romani (Padova 1907); B. Riposati, M. Terenti Varronis De vita populi Romani. Milano 1939; and N.M. Horsfall, "Virgil, History and the Roman Tradition," Prudentia 89 (1976) 73-89. On Cato the Censor as source material for Vergil, consult: Nicola Criniti, Enciclopedia Virgiliana 1: s.v. "Catone"; R. Helm, "M. Porcius Cato Censorius," PW 22: 1 ( 1953) 108-165; F . Della Corte, Catone Censore. Firenze 1969 (2nd edition); W. Ritter, De Varrone Vergilii narrandis urbium populorumque ltaliae originibus auctore, Diss. Philol. Halenses 13 (1897) 284-411; P. Fraccaro, "Reminiscenze catoniane in Virgilio," Boll. Fil. Class. 17 (1910-1911) 160-163; and B. Rehm, "Das geographische Bild des alten ltalien in Vergils Aeneis". Philologus Supplb. 24 (1932). Horace extols the antique Sabine virtues in Odes, 3, 6, 37ff. Vergil, Georgics 2, 167ff., and 532542, shares Horace's regard for the high moral standards of earlier folk. Livy ' s views were identical (1, 18, 4). For details on the old-time heroes mentioned, consult R.M. Ogilvie, A Commentary on Livy Books 1-5 (Oxford 1965) s.v. proper names. Camillus' career impinges on Volscian history; Gnaeus Marcius Coriolanus, eponymous hero (?) of Volscian Corioli, and Shakespeare's dramatic hero, is somewhat shadowy although his exploits are fairly reliable. During his exile from Rome, the Marcian sage relates that he led the Volscians against the city and was only deterred from attacking it by the prayers of his mother Veturia and wife Volurnnia (491 B.C.). The threat was no doubt real but the saving peripety seems to be concocted. His career to some degree parallels that of Vergil's Metabus, Camilla's disgraced father and expelled tyrant of Camilla's birthplace, Volscian Privemum. M. Dickie, op.cit., "Ethnography," 178-183; "Antiquarian Reconstruction of Early Rome," 183-188; "Farming, Hunting and Riding" 188194. On primitivism generally in the epic, see M.E. Taylor, "Primitivism in Virgil," AJP 76 (1955) 261-278. Livy's remarks on the Volscians (Books 2-8, passim) are generally less than complimentary. Their rebelliousness, subterfuge, treachery and deceit, addiction to brigandage and piracy, and their readiness to engender secret revolt, were ingrained (2, 22, 3; 2, 37, 4; 3, 7, 1; 6, 2, 9; 7, 27, 7 and 9). On Camilla, consult particularly Giampiera Arrigoni, Camilla: Amazz.one e Sacerdotessa di Diana. Milano 1982; N. Horsfall, "Camilla, o i limiti dell' invenzione," Athenaeum (1988) 31-51; Gilbert Highet, The Speeches in Virgil's Aeneid. Princeton 1972; A.G. McKay, Vergil's Italy. 24

14.

15.

16.

17.

18.

Greenwich, Conn. 1970; Brooks Otis, Virgil: A Study of Civilized Poetry . Oxford 1963; Bertha Tilly, The Story of Camilla from Aeneid VII and XI. Cambridge 1956; H.H. Huxley, "Virgo Bellatrix," Virgil Society Lecture .52 (1960) 1-4; T.J. Rosenmeyer, "Virgil and Heroism, Aeneid XI, " CJ 55 (1968) 195-164; R.G. Tanner, "Some Problems in Aeneid 7-12" PVS 10 (19701971) 37-44; T. Koves-Zulauf, "Camilla", Gymnasium 85 (1978) 182-205; 408-436; R.D. Williams, Virgil: The Aeneid of Virgil. Books 712. London 1973. On Metabus, see T. T. Duke, "Metabus of Privemum," Vergilius 23 (1977) 3436; and T. Koves-Zulauf , "Camilla", Gymnasium 85 (1978) 408-436. On Horace 's Journey to Brundisium (Satires 1, 5) consultG. Radke, "TopographischeBetrachtungen zum lter Brundisinum des Horaz," RhM 132 (1989) 54-72. The Amasenus passage bespeaks eyewitness encounters with seasonal torrents : Ecce Jugae medio summis Amasenus abundans/spumabat rip is, tantus se nubibus imber/ruperat , Aeneid 11, 547-549). The McMaster team studied the problem of bridging such a torrent elsewhere: I.P. Martini, E.M. Wightman, P. Hemphill, "Bridging a Seasonal Torrent: The Melfa River, Southern Lazio, Italy," Geoarchaeology 2:2 (1987) 149159. For detailed study of the Amasenus river in all seasons, consult L. Zaccheo, Amaseno (Frosinone 1979), especially 195-224. The Liris, best known to travellers who travelled by the Via Appia, bas a less sensational "press" than the Amasenus.· Horace, Odes 1, 31, 7-8 refers to the river as taciturnus amnis; Silius Italicus, 4, 43840 offers a comparable vignette; Liris... qui fronte quieta/dissimulat cursum, ac nullo mutabilis imbrilperstringit tacitas geminanti gurgite ripas. Horace, elsewhere, Odes 3, 17, 79, suggests that the mouth of the Liris near Mintumae was liable to flood the sanctuary · shoreline of the local goddess, Marica. On the Volscian urbes and oppida consult E.T. Salmon, Samnium and the Samnites (Cambridge 1967) 189-194; 206-208 ; 229-234; G. Radke, "Volsci", RE 9 A 1 (1961) 773-827; A.J. Toynbee, Hannibal's Legacy (London 1965) passim. For the excavations at Privernum, consult H.H. Armstrong, "I. Privemum: The i>rivemum: the Volscian City;" "II. The Roman City"; "III. Roman Remains in the Territory of the Roman Colony". AJA 15 (1911) 44-59 ; 170-194; 368402; G .M. De Rossi, Lazio Meridionale (Roma 1980) 126-130; F. Coarelli, Lazio (Bari 1982) 278-280; Margherita Cancellieri, "Priverno" , Enciclopedia Virgiliana 4 : 282-283; id., Enea nel Lazio : Archeologia e mito (Roma 1981) 78-79 . Coarelli, Lazio (Bari 1982), 280 ; see also M.L. Moriconi, in Enea nel Lazio; Archeologia e mito

19.

20. 21.

22.

23.

24.

25.

26. 27. 28.

(Roma 1981) 80. For religious details consult Radke, op. cit., and Salmon, op. cit. The known deities in the Volscian Pantheon include Dekluna, Marica, Vesuna, and Mefitis (Mefete, at Aquinum). Arrigoni (op. cit.) 92-115, discusses the Nemorensis traits of Camilla's patroness, Diana. See, recently, Loma Hardwick, "Ancient Amazons - Heroes, Outsiders or Women?", G&R 37 (1990) 14-36; and D. Von Bothmer, Amazans in Greek Art (Oxford 1957). On M. Furius Camillus, consult: F. Munzer, "M. Furius Camillus", RE 7(1910) 324ff; M. Treu, "Das Camillusgebiet bei Livius. Ein Beitrag zur Darstellungskunst des Livius", WJA 2 (1947) 63ff; F.W. Walbank, A Historical Commentary on Polybius I (Oxford 1970, 2nd edition) 184ff.; and R.M. Ogilvie, A Commentary on Livy 1-5 (Oxford 1970, 2nd edition) 669ff. The "heroic" tumulus unearthed recently at Pratica di Mare offers a paradigm for the hypothesis. Cf. Michael Grant, The Visible Past: Recent Archaeological Discoveries of Greek and Roman History and Roman History (New York 1990) 176-180; and G.K. Galinsky, "The Tomb of Aeneas at Lavinium," Vergilius 20 (1974) 211.Italy On Vergil and the Volscians, see Catharine Saunders, "The Volscians in Vergil's Aeneid," TAPA 58 (1927) 92ff.; and Kathleen M. Manneke, Prolegomena to a Social History of the Volscian Territory. M.A. Thesis, McMaster University 1982. unpublished. The Latin colony of Sora (307 B.C.), situated where the Upper Liris ends and the Middle Liris begins, and therefore a strategic site, was four miles from Cicero's ancestral homestead on the Liris. Sora serves as beguiling setting for Cicero's De Legibus: it was an attractive backwater, a community of mountaineers, rough and tough, straightforward and loyal, inclined to be superstitious (Pro Plancio 22; De Div., 1, 105; Brut. 40). On Casinum, consult G.F. Carettoni, Casinum. 1st. di studi Romani. 1940. On Aquinum, see M. Cagiano de Azevedo, Aquinum. 1st. di studi Romani. 1949. On Arpinum, see F. Coarelli, Fausto Zevi, Lazio (Bari 1982) Cicero's reference to a Via Herculanea in De Lege Agraria 2, 36 has been applied to a secondary road linking Aquinum and Interamna Lirenas with the Via Appia at Mintumae. But the Ciceronian reference to a Herculean Highway is more reasonably associated (pace Wightman) with the more celebrated causeway linking Baiae and Puteoli. Cicero's Cumanum and Puteolanum properties made him a familiar user of the Campanian roadway. cf. Angelo Nicosia, "A proposito di una via Herculanea passante per Ausonia," Lazio sud 1:5 25

29.

30.

31.

(1982) 10-11. On the Samnites, E.T. Salmon, Samnium and the Samnites (Cambridge 1967) is indispensable. Consult also his Chapter 14: "The Iron Age: The Peoples of Italy". CAH 4 (Cambridge 1988) 676719, especially 702, where he subscribes to the argument that the name Volsci suggests that they came from a "marshy" district. Varro comments on Apennine shepherds and transhumance (De Re Rustica 2, 3, 9). He stipulates that shepherds require sturdiness for transhumance undertakings, and need to be quick, nimble and able to bring do_wn predators, particularly wolves (ibid., 2, 10, 10-11). For study of peasant life and transhumance, involving sheep and less frequently goats, consult Bertha Tilly, Varro the Farmer (London 1973); Graeme Barker, "The Archaeology of the Italian Shepherd," PCPS 215 (1989) 1-19; E.T. Salmon, Samnium and the Samnites (1967); K.D. White, Roman Farming (London 1970), and "Latifundia: A Critical View of the Evidence", BICS 14 (1967) 62-79; and Marinella Pasquinucci and Emilio Gabba, Strutture agrarie e allevamento transumante nell'Italia Romana, Roma 1979. N. Horsfall, "Illusion and Reality in Latin Topographical Writing", G&R 32 (1985) 197-208.

Chapter 6

THE ROMANIZATION OF THE VALLEY E. M. Wightman

During the period of Volscian domination, the valley bad not fulfilled its geographical potential as a corridor between Latium and Campania, and as the Samnites moved in, tightening control of the upper Melfa and from there advancing on Sora and Arpino in the west, Casinum in the east, the whole area became more of a frontier zone. During the wars between Samnites and Romans, between 328 and 293, this element was to be further heightened, and the valley was the frequent scene of acts of war that are recorded - with sometimes greater, sometimes lesser precision - by Roman historians, especially Livy. Control of the lower Liri was effectively a principal aim of the combatants. 1 The Samnites needed it to strengthen their hold on the metalliferous mountains behind (La Meta and Le Mainarde) and to prevent the Romans enlarging their foothold on Campania. The Romans now had a decided interest in Campania, strengthened by the foundation of a colony at Cales in 334. Communications were however a matter for concern. Even though they now controlled the coast road (a strategic if small colony planted at Tarracina in 329 helped), this could be too easily challenged by an enemy using the passes through the M. Ausoni and M. Aurunci. The danger points were the pass from Fabrateria Vetus (Ceccano) to Priyemum (a Volscian town which was still unreliable), from the area of Arce and Ceprano to Fundi and Lautulae (near Itri) and further east from the Cassino end of the valley through to Formia and Minturna e. Prevention of such disruptions, as well as domination of a second, inland route, was obviously desirable. Romans must have welcomed the readiness of the Volscian communities Fabrateria (Ceccano) and Luca (perhaps M. del Fico) to submit to Rome in return for help against the Samnites (330). 2 Such considerations led to the foundation in 328 of a colony at Fregellae, situated on the left bank of the Liri guarding the approaches to the Lautulae pass .3 It is hard not to see this as an act of deliberate provocation against the Samnites. For the latter, it was bad enough that the Romans should have taken Sora from them in 345 (if Livy is right), 4 since that site, though on the right bank of the river, guarded access to the upper Melfa. Fregellae, besides being on the left bank of the Liri, bespoke Roman intentions far too clearly, and became the immediate casus belli. The actual site of Fregellae is of no little interest. Rome had now entered a new phase of colonization, one which shows a growing confidence. No longer was an impregnable position on some inaccessible hill-top (e.g. Segni, Norba) of prime concern. Fregellae 5 occupied a plateau of some 90 ha, formed by a terrace above the Liri, and defended on the east side by the small but steep valley of a minor tributary. Only on the north side was it readily

26

accessible by a narrow neck of level ground. Yet, although comparatively low lying, it enjoyed surprisingly commanding views, southwards over the junction of Sacco and Liri, northwards to Rocca d' Arce, to a lesser degree eastwards. It is in fact the type of site that was to become normal for colonies, and later for Roman army forts throughout the empire. Guarding a point at which the Liri could be crossed (by ferry if not immediately by bridge) with comparative ease, it controlled communications far more effectively than a site on the hills. Its name was that of an earlier Volscian settlement, most probably that of Rocca d' Arce. Since this had been captured by the Samnites, annoyance at the Roman successor was all the greater. The longest of the various wars b~tween Romans and Samnites (usually called the 2nd Samnite War, 326303) soon erupted. The historians, writing three centuries later, were often unable to recapture the strategic geography of the hostilities. Episodes are often without context, and Roman successes exaggerated to honour individual families. In the first years of the war, Livy implies that the Romans controlled communications and could march deep into enemy territory. While it is not impossible that a Roman army could march down the valley from Fregellae without always encountering organized resistance, it is unlikely that they had the freed.om of movement suggested by the accounts. Rather, the valley must have seen a good deal of skirmishing, as the armies felt out each other's strength. Forces were small, the Roman army at first no more than two legions (perhaps 10,000) which usually operated separately. Ordinary inhabitants must often have taken to the strongholds or to the hills, as at many other periods in the valley's history. A more aggressive Roman policy of entering Samnite territory from Campania led to a severe defeat in 321. A few years of peace ensued, but the Roman act of creating a new rural Roinan tribe, the Oufentina, in the Liri valley 6 (an act that implies additional Roman settlers) shows that consolidation at least went on. This act as much as perception of Roman weakness led to defections at Satricum (possibly M. Nero) and Sora and to a Samnite attack on Fregellae. 7 Although they did not take the colony, they were able to use the Lautulae pass and inflict a further defeat on Rome, which led to further defections among Aurunci and Campani. From this position of weakness Rome fought back hard, defeating the Aurunci (who henceforth ceased to have an independent political existence), winning back Fregellae after another successful Samnite attack on it. 8 More boldly and aggressively, they established the colony of Interamna Lirenas, 9 in what must then have been an exposed and dangerous part of the valley.

Besides generally extending Rome's control further down the valley, Interamna fulfilled more specific purposes. 10 It was a watchdog for Casinum which, however, being still in Samnite hands, was a constant threat to the Roman community (as was the stronghold at S. Vittore). It guarded the southernmost and easiest of the passes to the coastal route, the Ausonia pass through the recently troublesome Aurunci. Furthermore, it could be a base, when conditions allowed, for operations aimed at penetrating Samnite territory by way of Atina or Venafro, while also allowing communications with northern Campania. To begin with, the first two aims must have been the more realistic. The site of the colony, which is said to have had 4000 settlers, was akin to that of Fregellae. It lay on the upper gravel terraces left by the prehistoric lake, overlooking the Liri and protected along its southern flank by the bluff left by an older version of the river bed, on its northern by a tributary. Its name means "between two rivers" (the second is usually thought to be the Gari rather than the tiny Spalla Bassa, though the former is at least 6 km away); the additional names Lirenas or Sucasina distinguished it from another Interamna. Although the number of colonists gave a certain safety, the site must at first have seemed inhospitable, especially if we are right to conclude that much of the surrounding land had not previously been cleared for agriculture (this did however mean few resentful natives). The foundation of Interamna may well have been crucial. Although the Samnites managed to take Sora in 306 (and this caused the defection of Anagnia and Ferentinum in the Sacco valley), the Romans shortly thereafter mounted their final successful offensive. 11 One part of the double army making the assault on the heartland of Samnium may have made its way up the Gari under the colony's immediate cover. When peace came in 304 the vulnerable and essential Sora was shortly thereafter made a colony, while its neighbour Arpino was given a form of Roman citizenship (without suffrage to Rome). 12 Other arrangements are not recorded. Atina and the upper Melfa seem to have remained in Samnite hands. Aquinum may have been recognized as an independent ally of Rome. Roman control of the lower Liri was not assured, although the fate of the strategic Casinum and Venafrum (just across the hills, with access to both Samnium and Campania) is unclear. The Romans shortly went on to organize another new rural tribe, Teretina, among the Aurunci and northern Campani. 13 It took further hard fighting, during which Rome was menaced from the north by the Etruscans, Umbrians and even Gauls, as well as the Samnites, before a lasting peace and a satisfactory boundary between Roman and Samnite spheres and a lasting peace was achieved. The conclusive battle was fought further north at Sentinum, but both before and after this the Liri valley was again the site of operations. This time the Romans were able to use Sora as a springboard for the upper Mel fa valley. 14 The Samnites retaliated by raiding the lower Liri; it was

27

probably here that the Samnites entered a Roman camp under the cover of dense fog, and Interamna was almost taken over in a course of a pillaging expedition. The colony also served as a base for the final campaign into Samnium, which proceeded from there to Atina passing an Amitemum which might be either S. Vittore or a site at S. Elia (of which there are now no visible remains). 15 Thereafter the armies split up to advance on Cominium and Aquilonia, two Samnite strongholds indicated by Livy to be approximately 20 miles apart. 16 If Cominium is correctly identified with Vicalvi, or a site just to the northeast in the Val di Comino, then Aquilonia is probably on the eastern side of the Mainarde (which can be reached by a pass), perhaps near Montaquila. There are also arguments for seeing Aquilonia as being nearer the heart of Samnite territory, at M. Vairano not far from Campobasso (in which case another Cominium must be found). Both Atina and Venafrum now lay clearly~ Roman hands, and became praefecturae administered by a Roman official, as probably also Casinum. The frontier between greater Latium and Samnium now lay along the upper Voltumus, and the Liri valley was now free to act as a corridor giving access both to Samnium and Campania. There was a risk of it again becoming a battleground as the Samnites joined with Pyrrhus of Epirus (acting on behalf of Tarentum) to threaten Latium. After Pyrrhus' defeat, the Romans saw fit to plant another strategic colony, this time at Isemia, further into Samnium than was Venafro (263). 17 It may have been after the founding of lsemia that the Via Latina was completed. After linking Fregellae, Aquinum and Interamna and crossing the Gari near S. Angelo in Teodice, it did not take the shortest route to Teanum Sidicinum (the route of the modem autostrada, skirting the northern flank of M. Camino and M. S. Croce). Instead, it climbed over the saddle of M. Cesima and headed for Venafro. This makes it clear that the Via Latina was originally seen as a road to Samnium as much as to Campania. The latter could be reached by going down the Voltumus from Venafro, while the contrary direction led to Isemia and the heartland of Samnium beyond the M. di Matese. The Roman framework, consisting of colonies, prefectur es, cities with half-citizenship (which Casinum may have gained at some point) and federate allies, was now established, and with it came peace. Not all of the previously independent local communities of Aurunci and Volsci had survived. Satricum and Luca disappear, as do Plistica and Amitemum. It is to be noted that both M. Nero and S. Vittore shows some signs of subsequent use, consistent with the existence of sanctuaries that could survive in places which were no longer the centres of political entities. Whether or not they are to be identified in Satricum or Amitemum, their fate is typical for places that had defected or resisted, and even for some (Plistica) whose only fault was to have been taken by the Samnites. Possibly, however, some of these places had an insufficient population to make them viable as independent units under

a permanent Roman system, so that incorporation within the nearest larger centre (such as Casinum for S. Vittore) was the most reasonable solution. Since Roman historical writers were pre-occupied with wars and politics, it is, typical, in the context of Rome's second great war with Carthage that the valley is heard of again. Hannibal evidently recognized the strategic value of Casinum when he was attempting to rally Rome's Italian allies to himself but, according to the tale, was led instead by a guide who had misunderstood his purpose to Casilinum, on the Campanian side of the mountain barrier. 18 Despite his base at Capua he thus failed to gain a foothold further north (though he gained some Samnite allies) and traversed the valley only later, when he led his army towards Rome (mainly to divert the Roman's attention) in 211. He laid waste the territory of Casinum (which did not surrender to him) and of Fregellae (where the bridge over the Liri had been cut), but passed by Interamna and Aquinum. 19 Livy's text may imply that he took the old road under the mountains, and turned south to Fregellae only after crossing the Melfa near Roccasecca. But while most of the campaigns affected the area only indirectly, Interamna was one of a dozen colonies claiming in 209 that they no longer had either men or money to contribute to the war - a reminder that the Liri valley, like other areas of Italy, contributed now and later to Rome's military exploits, and that this could readily affect population levels. 20 After taking no action at the time, Rome demanded an increased number of soldiers from the defectors only a few years later, 21 and it must be supposed that Rome's almost continuous warfare from t\}e time of Hannibal onwards constituted a steady drain. 22 Since not all veterans would survive, or return even if they did, this is also one of the means by which some families must have failed while others grew richer at their expense. It would also be in keeping with the conditions of the 2nd century if a number of families from Interamna left the area to seek their fortunes in Rome, where they had theoretically the possibility of gaining full Roman citizenship, an option not open to allies. As the weight of Rome's controlling hand over the peninsula was more keenly felt, this status became increasingly attractive. Eventually Rome's refusal to consider giving full franchise to Latin citizens was the cause of the one event in the valley to reach the history books in the 2nd century - the revolt of Fregellae. 23 The reprisals saw the city razed to the ground, but it was replaced shortly thereafter by the new colony of Fabrateria Nova, placed on a non-defensible site on the south bank of the Sacco. 24 The territory of Fregellae henceforth belonged to the new city, though Aquinum may have gained some additional land. The final period of warfare between Rome and her Italian allies - the Social War in 91-87 B. C. - also hinged on the question of citizenship, which Rome had refused to extend. Nothing about attitudes or even specific to the valley is known (the Samnites were again among the most resolute of Rome's enemies). The results were a change of status,

28

as Roman citizenship was granted to those who ceased active opposition (eventually to all) and the cites were reorganized as municipia with boards of magistrates (duoviri) of a uniform type replacing previous arrangements, which were now phased out. A decade later, it is possible that some fresh settlers arrived in the form of veterans of the dictator Sulla. A great deal of land throughout Italy changed hands as a result of his confiscation and the opportunities enjoyed by his favourites; we hear of one such, C. Quintius Valgus, who had lands at Casinum as well as in Samnite territory. 25 Fresh settlers certainly reached the area after the death of Caesar. Aquinum and Casinum formally became colonies (with duoviri) and the centuriation visible in the territory of the former (especially on the plains) most likely belongs to this time. Around Interamna, the work of official surveyors seems to have taken a less rigorous form that has left no clear traces. The corollary of the new colonization is as interesting as the fact itself - that the population of the area tended not to maintain itself unless supported by a fresh influx from time to time, a question that will receive further discussion. Under the pax Romana of the early empire, the area gains no further mention from writers of history. Fortunately, there are occasional references in the agrarian writers from Cato the Elder onwards. There is also the·stray comment in the Younger Pliny to confirm what the epigraphical record (which increases just at the time when the writers fall oft) tells us - that Casinum in particular was the home of a few powerful families, some of whom played their part in the government of the empire. Aquinum was very likely the home town of the poet Juvenal. Some figures, and the more numerous local aristocracy that lay beneath them, will be further discussed after the rural settlement patterns have been considered. The epigraphical evidence unfortunately shares one feature with the literary, in that (the agrarian authors apart) it tells us more about the life of the towns than about the countryside.

Notes 1.

2. 3. 4. 5. 6. 7. 8. 9.

10.

For the background picture, and discussion of the events noted below, see Salmon, Samnium passim. Livy VIIl.19.1. ibid. VIIl.22.2; Salmon, Samnium 212; Coarelli, Fregellae, with bibl. cited. ibid. VII.28.6. Coarelli, Fregellae 26-29; Crawford and Keppie, PBSR 52 (1984) 21-22. Livy IX.20.6; Diod. Sic. XIX.10.2. Livy IX.12.5. ibid. IX.28.3. ibid. IX.28.8. Most of the ancient literary sources on Interamna, and some inscriptions, are presented by Cagiano de Azevedo, Interamna Lirenas 47-52. For the Aquinum Casinum texts, see id., Aquinum 73-78, and Carettoni, Casinum. Cagiano de Azevedo, Interamna 8.

11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22

23. 24.

25.

Livy IX.42.11-43.7. ibid. X.1.1-2. ibid. X.9.14. ibid. X.14.4. ibid. X.39.1-5. ibid. X.39.7; discussed by Salmon, Samnium 270-2 Livy, Epit. XVI; Vell. Pat. I.14.8. Livy XXII.13.5-9. ibid. XXVI.9.2-3. ibid. XXVII.9-10 passim. ibid. XXIX.15. Note that Samnites and Paeligni were settled at Fregellae in 177 (Livy 41.8.8, Salmon, Samnium 318). This suggests that many of the original inhabitants had been lost. The same may have happened elsewhere. Livy, Epit. LX. Vell. Pat. I.15.4; A. Nicosia, Fabrateria Nova (Pontecorvo, 1977); M.A. De Lucia Broli, Arch. Laziale 5 (1983) 104-11, with bibl. cited. Cicero, De lege agr. III.1.3; 3.13; 4.14.

29

Chapter 7

COMMUNICATIONS E. M. Wightman

Seen from a satellite, or viewed on any simplified map, the whole Sacco-Liri river valley has the appearance of a natural corridor. No insuperable problems present themselves at first sight, though some parts of the landscape--for instance, the rolling hills and steep valleys east of Frosinone, or the narrow pass over to Campania in the direction of Teano - are clearly trickier than others. The smoothness of the plain below the M. Cairo massif (Figs. SA, 13B) intensifies the general impression of easy communications. In fact, within the 30 km. stretch between the southward course of the Liri River as it issues from the hills and the parallel course of its tributary the Gari River two kinds of problems arose. The first of these is timeless, and can at best be partially solved by advancing technology, for it springs from the underlying geology. Liri, Melfa and Gari rivers must all be crossed in tum. The Liri River, with its strong perennial flow (Fig. SB), regularly increased by spring floods, might seem the worst, but at least its banks are relatively stable and its flow predictable. Though fording points may be rarer, it is less difficult to bridge or to cross by boat. The Melfa River, while easy to ford during the summer months, is certainly the trickier if permanent crossing points are desired, because of its steeper downward course and gravelly bed (Fig. 15). The Gari River is like the Liri River on a smaller scale, and also liable to spring floods and marshy stretches.

In addition to the specific obstacles of the streams, the valley's natural vegetation - some idea of which can be gained today from the oakwoods west of the Melfa River - impeded progress, and the Iron Age settlement patterns, as seen, strongly indicate that clearance on a more than localized scale was not undertaken until later. Settlement patterns echo the second problem, which is that changing political considerations can greatly influence the desirability and the feasibility of using the valley as a through route. Conditions during the last two centuries before Roman conquest, it has been argued ~hove, were inimical to traffic. The valley served as frontier rather than corridor, and while not of course hermetically sealed, was no highway. Such travellers as passed through would in any case not have gone down the middle of the valley as did the Roman Via Latina and its medieval or modem successors. The available route would have been a less direct and more northerly one, skirting the hills from a starting point at Praeneste. From Ferentinum or Frusino (Frosinone) it may have been normal to head first east towards Sora and Arpino, crossing the Liri River in this stretch, and only then southwards towards the lower Liri River valley. 30

Rocca d' Arce could have been by-passed by a hill-road from Arpino which emerges from a minor valley (Rio Moscosa) to meet the Melfa River just west of Roccasecca. Here lay the oldest and most permanent crossing-point of the Melfa River, where it can be forded (at least for part of the year) just below its exit from the steep and virtually impassable gorge which deeply furrows the M. Cairo massif. Discovery of Iron Age pottery just east of the crossing (the modem Ponte Vecchio) confirms an early date for the route and suggests that the road then forked into two, perhaps three (Fig. 31). The most northerly and minor branch led up through modem Roccasecca village into the hills beyond Colle S. Magno. While our very partial sampling of the uplands did not confirm Iron Age settlement, it is not thereby to be ruled out, especially as some Bronze/Iron Age pottery was found in the upper end of Roccasecca. The main branch pursued a route along the lower edge of the foothills, passing several major Iron Age sites - essentially the road known to-day as the Pedemontana. A more southerly branch probably headed at an angle across the plain to Aquinum (Fig 20, broken line); this, like the others, thereafter remained in use through Roman and medieval times and can be seen today as a marked oblique line on maps, aerial photographs or from such vantage points as Roccasecca and Castello. Was there a corresponding road along the south side of the valley? Certain Iy, further north nearer Rome, there were tracks along the lower hillsides in the vicinity of modem Artena (probably not the ancient site of that name), across the gap from Praeneste/Palestrina. The site of Fabrateria Vetus was probably at Ceccano, and a fortification is preserved at M. Nero near Castro dei Volsci. Within the lower Liri Valley, the sites at M. Leucio, S. Giorgio and S. Apollinare raise the possibility of a continuation, though the chain is less strong than on the other side and the goals less evident. There is also little definite evidence for north-south links across the valley, though tracks from Aquinum to M. Leucio and S. Giorgio are not implausible. Beyond Casinum, viable routes presumably led up the Rapido River valley to Atina, and by way of modem S. Vittore del Lazio and S. Pietro Infine (both sites of fortifications) to Venafro. In short, the major centres were linked, and those links to some extent also served local needs. About other, minor tracks, little can be said. The physical aspect of such routes is not directly attested. At the very least, clearance of vegetation allowed the passage of tracks which doubtless varied. in width, and which would be worn more by hooves and feet than by wheels. The techniques for the rough paving of roads and fords clearly existed, but there is not evidence for their use

for this purpose. Most strangers presumably came prepared with something in the way of tokens of guest friendship or safe-conduct, or else travelled with an armed escort at their own risk. New needs in the sphere of communications came into being with the founding of Fregellae and later that of Interamna: new solutions were found, though doubtless not all overnight. Faster, more direct routes were called for, so that the new colonies could be conveniently linked to each other and to the towns further north, and, in addition, access to Samnium and Campania be provided by a continuation of the route. The exigencies of the late 4th century - wars first with Latins, then Samnites, the need to reach Campania, intensified by the founding of colonies there, saw the beginnings of Rome's great network of strategic roads. The Via Appia, giving access to Campania by the coastal route, and the Via Valeria going eastwards across the peninsula by the Fucine Lake, are the other obvious examples, both begun before 300 B.C. The Via Latina differs from them in not bearing the name of a consular builder, but answers the same need to provide access to distant colonies and potential theatres of unrest. From now on, political barriers disappeared as a result of Rome's victories, and three types of roads can be distinguished: strategic trunk routes, roads providing links between them and roads of purely local importance. The first, the viae publicae, were the concern of the Roman state, the others of the local communities or even of private individuals. The name of the Via Latina' shows that it was considered the extension or rationalisation of an old route . The name is derived either from an original goal, the federal shrine of Jupiter Latiaris in the Alban Hills, or from the fact that it linked the originally Hemican or Volscian cities of the area which, since the Latin War, was thought of as "Greater Latium" (Latium Adiectum , in Pliny's later phrase). 2 Although the general context, purpose, and line are clear enough (Fig. 20), there are some problems over its exact route and the dating of the various stages which are attested from milestones and other late inscriptions, and which answer to modifications of function. To start with, a direct line from Frusino (Frosinone) to Ceprano had to be sought: at the latter place, a convenient crossing point over the Liri could be found between the bluffs of Ceprano and Fregellae on either side, and if not immediately, certainly within a century of the colony's foundation it was provided with a bridge (which was destroyed by Fregellans to impede Hannibal in 211). From Fregellae eastwards, progress was not entirely simple, being first barred by a series of small yet steep and awkward valleys, heavily wooded at the present day. However, the Melfa River could be reached by skirting these valleys to the south, and this route, shorter and easier than a lengthy northward deviation to the Pedemontana or a double crossing of the Liri River to the south, seems to have been followed. At the chosen point, between bluffs and narrows upstream and an open,

31

shifting, braided stretch downstream, the Melfa River could be forded with ease during the dry seasons, until (just when will be discussed), a bridge was built. Once up on the plain of Aquino, the roadmakers could simply choose their path, and the original line may have lain a little north of the subsequent one which is still visible in the countryside today. Immediately east of Aquinum, a small stream broadened into lakes and marshes, and though one at least of the narrow points was in due course bridged an alternative solution may have originally have been preferred. This was to proceed southwards from the town, down a long headland to a point some 2 km north of Interamna, where the stream could be more easily crossed (modem Selva di Troccheto and Ponte Fraiola) . At Interamna it did not enter the city itself, but ran along a parallel ridge to the north. Later, the Via Latina went from here to Casinum, crossing the Spalla Bassa on the little arched bridge (Fig. 41C) which is still preserved and heading straight for the latter town, a line still largely followed by tracks and roads. This may not however have been the original route, for Campania and Samnium could be more directly gained by passing to the south of Casinum. The travertine bluff at S. Angelo in Theodice, and the gentler headlands just opposite may in any case have offered a better place for crossing the Gari than the flatter, possibly marshy stretches above and below. From Interamna, S. Angelo in Theodice was reached by descending for a short stretch into the Liri valley, while avoiding the floodplain. An ascent, marked today by a deeply gouged hollow way, then gave access to ridges which led to the crossing-point. From here the road could pursue its course without difficulty, passing to the south of M. Trocchio and M. Porchio, but to the north of M. Lungo. It then continued on a course which took it north of the modem roads and the direct route to Teano. Instead, passing below the modem S. Pietro Infine (traditionally the boundary point between Lazio and Campania) it negotiated a much higher, more difficult pass, descending into the Voltumo River valley at Venafro before turning south. This apparent deviation had however a good strategic reason behind it. Venafro was a key site and acted as a tum-table , for from here advance could be made either into Campania or into Samnium, and the route, while not the most direct to Teanum, made excellent sense in the aftermath of the Samnite wars, Subsequent modifications to the Via Latina (Fig . 20) in fact mirror the fortunes of the various cities (described briefly in the next section). When Fregellae revolted, was destroyed and replaced by Fabrateria Nova (124 B.C.), the road crossed the Sacco River between Ceprano and Fabrateria (thus avoiding the site of Fregella e) and only thereafter the Liri, to join , after this short detour , the old line towards the Melfa River. Already before this time, the attraction of the trunk road to Casinum appears to be confirm ed by the distance given on a milestone (near Venafro) of 127 B.C ., a year when the road clearly received attention from one of the consuls . By 39 B.C. the Via Latina went straight from Aquinum to Casinum, by-passing Interamna (there may have been an intermediate

stage when it left Aquino for Interamna over the lower of two bridging points). These changes reflect the disappearance of a strategic role for both colonies and road, and the increase of economic considerations. Yet later itineraries show that the road still went to Venafro, even if the geographer Strabo might erroneously suppose that it took the shorter course to Teanum, along which another road certainly did run. The by-passing of Interamna was only one facet of the work undertaken on the Via Latina in 39 B.C. Aquinum became a colony about this time and its territory was surveyed and centuriated into regular squares for the settlement of veterans from the Civil Wars. That of Interamna was also surveyed, but according to a less strict system that has left no permanent trace on the landscape. It was probably now that the Via Latina received its final course across the plain, a course which may have been determined by the choice of some mountain (e.g. the hill behind Rocca d'Evandro) as a sighting point when the grid was being laid out. This work may mark the inception of the Via Latina Nova mentioned in later inscriptions. 3 The Via Latina Vetus retained its separate identity , however, and its status as a public road with curators to look after it, being normally bracketed with the Via Labicana (which left Rome on a different path but joined the Via Latina near Anagni). Now too, if not indeed before, the old hill towns such as Anagnia were by-passed by the newer version, creating new road-side stations that were the ancient equivalent of the more modem "Scalo", situated at the nearest point on main road (or later railway) to the inconveniently placed town. The traditional appearance of a Roman trunk road - well engineered bed, a paving of stone blocks, bridges at least partly in stone, would be reached only with time. In the initial stages the main task was to delimit a strip of public land and clear it of vegetation: at the most, gravel or stones would be laid along the track (no problem in the lower Liri valley), and wooden bridges provided when the need could not be avoided. Paving and the construction of more solid bridges could come later, most likely in the 2rd or the late 1st century, periods when much attention to roads in general is attested. As noted, there was a bridge at Fregellae before the Hannibalic invasion, though its type is not known. The Melfa crossing could have remained longer as a ford, for the gravels were easier to cross, and even a strategic road was required mainly in the summer. It was also more difficult to bridge; the combination of gravelly bed and intense floods have destroyed a whole series of ancient and modem bridges (Figs. ISA, B). Remains of some at least of those which carried the Via Latina across were uncovered in the course of heavy floods following on gravel extraction in 1976 (Fig. 16). A series of wooden piles at first sight suggests an initial wooden bridge, but may rather have served to break the current for the first of two stone-piered bridges. A Cl4 date gives a date in the 3rd or 2rd century B.C. for the piles, and the first stone bridge (which almost certainly was not arched, but had a flat wooden superstructure) could as a result be

32

as early as the 3rd century or as late as 127 B.C. The second bridge (also flat-topped) incorporated a number of re-used decorated blocks. The possibility that these came from the destruction of Fregellae arises, but this bridge could equally well belong to the Augustan period or later-Vesp[ asian] re-cut · milestones on the road, Hadrian repaired the bridge at Ceprano (For further discussion of Melfa River area, with presentation of survey results, see Chapter 12, this volume). The paving of the Via Latina may, but need not, have belonged to one of the periods of bridge-building: since it was a via publica, it would in any case be undertaken at state expense, whether or not undertaken as a single programme. So far as can be told, white limestone blocks, such as can still be seen on the stretch of roads near the amphitheatre at Cassino, were the normal material. Although a few black basalt blocks can be seen built into houses at Aquino, no regular trail of these has been observed, and they must have been confined to certain stretches, perhaps within the city of Aquinum itself. Throughout its varied forms, the Via Latina was obviously the main artery of the region. Like some other strategic trunk roads, it did not however either follow or attract settlement in the same way as roads serving more local needs. There are only a few republican sites along its course, though these include a small straggly village on a tongue of higher ground east of ·S. Angelo. Even as the road's role became increasingly socio-economic rather than strategic, linking the main market centres and allowing a fast route to civilian travellers, increasingly landowners, its impact on the countryside remained quite low. This, combined with the use of local limestone for paving, sometimes makes its detailed route the more difficult to trace. Of the secondary roads (Fig. 20) that provided links between the main public arteries, the one which linked Aquinum and Interamna with the Via Appia at Mintumae by way of the Ausonia gap also very obviously passed from a strategic to an ordinary civilian role. Roman colonies, like many of their predecessors, were obviously sited with an eye on the passes which led from the Sacco-Liri valley over the mountains to the coast . Control of the passes offered to the Romans alternatives in the event that stretches of the principal routes were unsafe, and prevented any potentially hostile inhabitants of the mountains from uniting. The existence of the Ausonia pass (the name is modem, not ancient) was clearly a factor in the siting of Interamna. Whether a road was ever engineered for essentially military purposes is unknown . When we meet this road in the writings of Cicero, it is the Via Herculanea, taking its name from the sanctuary of Hercules situated near Ausonia and affording a convenient way of reaching the orator ' s pative Arpino from the coast. 4 It diverged from the Garigliano River just outside Mintumae, and followed a small valley up to the pass, which is relatively low and easy, remaining below 200m. The remains of an old bridge (probably medieval but on an

ancient line) over a small but awkward stream at the northern end of the pass (Toricelli) are visible. The last part of the path to Interamna was over the Liri gravels. A bridge would then be needed to carry the road across the Liri River (an elderly native of S. Giorgio claims to have seen the remains of piers at low water as a youth). The Via Latina then took the traveller to Aquinum, and from there the fastest way to Arpino was the old road to the Roccasecca crossing followed by the hill-road or the road round by Arce depending on the precise destination and the desired degree of comfort. Another, slower route from the coast was available, for a road almost certainly ran up the Garigliano from Minturnae to Suio, where the sulphurous waters were exploited for medicinal baths (remains still visible) by late Republic-an times. From there, the way to S. Apollinare passed over the low saddle once guarded by the Iron Age site Morroni. An extremely low-lying ·site (within the modem flood-plain) at S. Apollinare could well mark the site of a ferry, the modem successor of which was only recently replaced by a bridge over the Liri. Once on the north bank, it was easy to reach an old ridge-road which is bordered by Roman sites for part of its length and which heads for Aquino (Figs. 44, 45). Another branch of the same ridge road probably crossed the Gari immediately above its confluence with the Liri River at Giuntura, forming a link to the Via Latina further east, or simply a route down the other bank of the Garigliano. Further west, Fregellae guarded the much higher and more difficult pass which debouched above Gaeta and Formia (probably near the ancient Lautulae), and a road from Aquinum (Fig. 20) is thought to have crossed the Liri River (most likely at Pontecorvo, though no remains are certainly earlier than medieval) and headed in that direction, passing just below M. Leucio. An alternative may have been to head for the confluence of Liri and Melfa rivers (a string of sites leads in this direction) and cross the Liri River just below (by ferry, since it is not a suitable point for a bridge). These roads to the coast were not Roman discoveries, though Roman technology and care made them more negotiable. The same is true of the roads on the north side of the valley, leading to Arce, Arpino (and thence Sora), and Atina. Only the Melfa River gorge in all probability remained a forested and trackless defile, while the uplands would be reached by mere mule-tracks, not very different from their later successors, since they are essentially dictated by the terrain. Roads of purely local importance doubtless included several that criss-crossed the valley from north side to south. Some are suggested by strings of sites, or by topography. Two such possible routes can be seen just east of the Melfa River (Figs. 22-24), and Aquinum must have been approachable from north as well as south. Another natural line is north of Interamna to the foothills near Piedimonte, followed by a road at least since medieval times. The Roman situation may have been a little more complex. A road north from lnteramna would

33

meet up with one of the Aquinum centuriation grid-lines (Figs. 34, 36), and since such grid-lines were normally marked by roads, this offered an obvious route to Aquinum (and indeed a version of it may have become the Via Latina). The sites north of Interamna in addition show a strong south-north linearity which does not fit, probably because earlier than, the centuriation grid. There was in all likelihood a north-south track, with private roads (diverticula) to individual sites leading off it. Since all farms must have had a track or pathway leading to them, there must have been a relatively dense pattern of such tracks, most of which are not recoverable. Only one possible instance of a private track to a villa has been recorded: below the opulent villa Parito the farmer reported an oblique stony strip to be visible after ploughing. Other natural lines for minor roads are the Gari valley south of Casinum and the various headlands in the dissected area. Occasionally, paving which could conceivably be ancient can be glimpsed at the edges of a modem side-road (for instance, on part of the ridge-road between Panaccioni (site 462, Figs. 20, 40) and Piumarola (site 240, Fig. 20, 32), but the continued use of limestone blocks for road-bottoming in the area precludes certainty. The Roman period thus saw the transformation of the valley from a largely uncleared area of natural vegetation, with established tracks (linking major centres and accompanied by minor settlements) only along the lower flanks of the hills, to an open stretch of country with easy communications in many directions thanks to a hierarchy of major, secondary and private roads, a number of them paved. Bridge-building technology had overcome some of the natural obstacles, peaceful unification under the hegemonical rule of Rome the human problems. Various levels of upkeep reflected the new and more complex system. Secondary roads were the responsibility of the various local communities, service roads to farms or villas of the private individuals concerned. Landowners might also be expected to keep stretches of public road free from encumbrances, though repairs, sometimes attested by milestones or other inscriptions, were at the expense of the state. The optimum was reached by the late republic and early empire: it will be seen that later settlement patterns suggest that many sites, and with them some of the tracks, had fallen into disuse long before early medieval times.

Notes 1.

2. 3. 4.

Via Latina: G.Radke, P.-W Suppl. XIII (1973) cols. 1487-94. N.H. III.59. C/L X.5398 (=/LS 11159). de Lege Agr. 11.14.36.

Chapter 8

SETTLEMENT PATTERNS AND SOCIETY E.M. Wightman and J.W. Hayes

The communications network, it has been seen, was a function of settlement rather than a determinant of it, with a hierarchy of roads corresponding to different types or levels of settlement. It is to the pattern of settlement in the Roman period that we must now tum, and the extent to which it was a function of topographical or other factors, natural or cultural. At the top of the ladder were of course the towns of Fregellae and its later replacement Fabrateria Nova at the western edge of the area, Aquinum, Interamna and Casinum more central to it. Their different status and history is of importance for settlement here only inasmuch as the Roman colonies brought an influx of settlers from outside to sites which, though of strategic importance to the Romans in their quest for domination within Italy, had not naturally grown into centres of population. Additional settlers might also be added after the initial foundation; we hear specifically of Samnites and Paelignians coming to Fregellae in the early 2 nd century. 1 The artificial nature of the Roman colonies is most apparent at Interamna, though the apparent blank may be exaggerated by our ignorance of the full extent of the earlier community across the river at S. Giorgio. Fregellae at least took its name from the Volscian site which it effectively supplanted, even though this may have been some distance away at Rocca d'Arce. Casinum and Aquinum were wholly indigenous growths, and the original nature of the latter - assuming that Mefete was a sanctuary of Mefitis - may be reflected in the striking preponderance of female deities (Isis, Fortuna, Ceres, Bona Dea)2 later found in the town and its territory. In the 111 century B.C. distinctions became more blurred. Aquinum, Casinum and Interamna all were granted municipal status (with appropriate constitutions) after the Social War: 3 moreover, the first certain Iy, the second probably and the last perhaps received an influx of veteran soldiers from the Roman civil wars around 40 B.C. 4 All alike had long been the principal market and administrative centre for their regions. In addition, all came (perhaps by stages) to adopt the Graeco-Roman tradition according to which the territory was viewed as the subordinate adjunct of the town, and the latter was the major focus of civic pride even although its development was supported by the agricultural wealth of the territory. The extent of the individual territories can be established approximately, but not with precision or certainty. There is no convenient correspondence between towns and later bishoprics, such as allows the medieval diocesan boundaries of France to be used as a basis for the territories of the Roman period. 5 The boundaries of the territorium of St. Benedict, centred on the abbey of Montecassino, would be of greater help if the earliest (lOllt

34

century A.O.) form was more fully comprehensible, but it is clear that parts of the original territories of Aquinum and Interamna have been incorporated into it, perhaps also smaller areas from Atina, Teanum and Suessa Aurunca. Medieval boundary documents do at least confirm what topographical considerations suggest: they often ran along mountain ridges and summits (M. Cairo appears consistently, already under that name) and the major rivers, while smaller streams might also be used but were as often as not crossed. At some points, boundaries probably did remain relatively stable through the centuries. This is most likely in the mountains, but may also be true, for instance, of the western border of castellum Teramense, successor to Interamna, against Aquino and Pontecorvo; in both periods delimitation had to be made through the same area of difficult terrain, awkwardly dissected by small valleys running in different directions, lying to the east of Pontecorvo (even if the last place had no independent existence in Roman times). Specifically Roman evidence on boundaries is of two kinds. The first is that formed by the find-spots of rural inscriptions attributable (often because of the mention of a magistrate or tribe) to one or other of the towns. Thus S. Angelo in Theodice has produced inscriptions appropriate to Casinum, 6 at Pontecorvo inscriptions of Aquinum 7 predominate over those of Interamna, 8 and at Piumarola inscriptions of all three towns are found. 9 Problems however arise, since it is not clear whether the inscriptions originated in these places or were carried there in the course of medieval building activities - one inscription from Piumarola 10 is a list of the magistrates of Interamna which would naturally belong to the city itself rather than to a point on the periphery, even if we suppose that there was a village or important sanctuary. Nonetheless, since the boundaries must have run near Piumarola simply because of the distances, the indications from inscriptions may effectively be correct, if not necessarily for the right reasons. As for Pontecorvo, its dependence on Aquino in medieval times may perpetuate older boundaries, in which case the inscription of an Interamna magistrate 11 must have wandered there at a later date.

In one area the results of survey may indicate a boundary zone - and it must be remembered that wooded, infertile or mountainous areas could function as broad boundary zones within which precision would only be important within particular legal contexts. Between 1km and 2km north-east of modem Pignataro, south of the Rio Poppeto, lies an area curiously void of remains of Roman settlement, yet no different in topography and soils from neighbouring areas. In all probability this area was never cleared of woodland, and was effectively a boundary zone between Interamna

and Casinum, even if technically it may, for instance, have been ager publicus of Interamna, with rights of forage that could have been rented out (the discovery of an inscription mentioning a magistrate of Interamna built into a farm on the north side of this blank area suggests that the actual boundary was close to the Rio Poppeto). This area serves as a reminder that topography alone did not always determine settlement, and that the existence of territorial boundaries could also, at least to a mild degree, affect marketing habits, in that a person living just south of the Rio Poppeto might have gone to Casinum as against the slightly closer Interamna. Although there is no need here for a detailed description of monuments, the physical development of the towns must be traced a little further in their broad lines, so as to allow meaningful comparisons among them, as also between urban and rural developments. Although archaeological knowledge of the towns is inevitably imperfect, various measuring criteria can be applied, even if not wholly systematically - simple area, facets of town-planning, complement of major monuments, number of inscriptions, testimony for municipal organization and commercial life. To these can be added data from recent excavations at Fregellae, 12 and survey at Aquinum and Interamna (only at Casinum is this rendered wholly impracticable by modem building). Inequalities between the cities rapidly emerge. Interamna, despite its recorded original 4000 colonists, was never as large as the others. Survey shows that the town was confined to the more southerly of the two parallel spurs that were available, with only a narrow ribbon development accompanying part of the Via Latina as it ran along the northerly one. The total inhabited area never exceeded 30ha. or at most 40ha. By contrast, Fregellae, despite its shorter life, covered some 80-90ha. Aquinum may originally have been more modest: survey here (Fig. 39B, Appendix D) shows that pottery of the 4th and 3rd centuries B.C. is commonest in the northern part of the city, 13 which implies that it spread south from the original site of Mefete. Its growth may have been assisted by the arrival of newcomers after the sack of Fregellae, for epigraphic evidence points in this direction, 14 and Fregellae's successor, Fabrateria, was a mere 25-30ha. in extent. Since it is normally the cities of the late Republic and early empire that are most easily delineated, it is worth noting that Fregellae had vaulted drains below some of the streets. Other normal monuments of the republican period included temple complexes, attested by excavation at Fregellae and suggested by survey outside as well as inside Interamna 15 and in the northern sector of Aquinum, near the later capitolium. 16 Nothing precise is known of fora: at Casinum the construction of such a complex would have posed a real problem unless the town expanded downwards on to the lower ground. Such a spread, attested by local tradition but not yet securely by archaeology, would imply that some level ground had been drained by the containment of the watercourses. This is however not only possible but likely, the more so since

35

there are remains of aqueducts (though of uncertain date) leading to all the towns, perhaps largely in order to supply public baths. The amphitheatre and theatre of Casinum were however constructed on the flank of the hill, the former perhaps in the 1st century B.C. (with later additions by a member of a prominent local family), 17 the latter a century later. The capitolium, theatre and amphitheatre of Aquinum belong to the period after its establishment as a colony, as do the gates, walls (though an earlier phase of these may have existed) and an honorific archway spanning the Via Latina Nova after it crossed the stream to the east of the city. By contrast, city walls are not definitely attested for Interamna, which also lacks theatre and amphitheatre (though part of the valley between the two spurs could easily have been adapted to such functions). The same contrast shows up in the inscriptions, where Interamna has only one quarter to one fifth of the number found in the towns and territories of Aquinum and Casinum, 18 and major families with links to the city of Rome and the political or administrative life of the empire are also commoner in the latter two places·. The general picture is thus wholly consistent with the by-passing of Interamna by the Via Latina Nova, and with the conduct of M. Antonius the future triumvir when he spent some time in the area in 44 B.C. He stayed principally at Casinum, partly because he could occupy the rich suburban villa of his political enemy M. Terentius Varro (the confiscation of whose property was definitely planned, if not certainly perpetrated). On the way back to Rome he did however (Cicero tells us) stop at Aquinum, where he was warmly welcomed by what is described as a large population. 19 The inhabitants of Interamna had to travel to one of the other cities to greet him. The growth of the towns or cities, and the increasingly monumental appearance which they assumed in the later Republic, is itself one aspect of the transformation of the landscape. It had of course implications for the rural hinterland, since a growing urban population (even if still small by any modem standards) could only be supported by increased agricultural output. Recapturing the evidence for such rural change was an important aim of the archaeological surface survey undertaken in the region. In view of its general history it could be anticipated from the start that change would at some times and places be relatively slow and natural, at others more rapid and forced, depending on the proximity of an indigenous town or a Roman colony. Both patterns are indeed found. On both sides of the Melfa near Roccasecca, more Iron Age sites than not show some measure of continuity into at least the early Roman period, though new Roman sites are almost as many as the Iron Age total. If the area to the west of the Melfa did, as is usually supposed, originally belong to Fregellae, then it was sufficiently remote from the colony for the established pattern of settlement not to be wholly disrupted. Further east, in the foothills between Castrocielo and Piedimonte, some sites likewise suggest continuity. Problems of precise dating inevitably arise, because without excavation it is not possible to define

when wheel-turned coarse pottery was first used, or just how long hand-made coarse pottery in Iron Age tradition persisted. Sites producing fragments of Iron Age type could in fact be founded in the period after the Samnite wars, although equally a measure of continuity might be anticipated in the territory of Aquinum. In the plains and the dissected area, regardless of adherence to city, nearly all rural sites were inevitably new. Even here, however, there is the occasional possible exception, such as the villa site 20 above the confluence of Melfa and Liri rivers, in Aquinate territory. East of Interamna, a site near Giuntura (the confluence of Liri and Gari rivers)2 1 produced some pottery of Iron Age type followed by late rather than early republican wares. The sizeable community on the Liri gravels at S. Giorgio 2th ia came to an end (not without using some wheel-made coarse wares), understandably in view of its proximity to the colony. Thanks to its colonial foundation and to the focus of the survey project, Interamna and its territory offer the best opportunity to study the population of related town and countryside, and to encounter the various historical and archaeological problems which arise. Since much of the area had to be cleared before agriculture could be undertaken, the original colonists must be imagined as living in the town and farming their plots from there - a time-honoured practice which the uncertainties of war-time would render the more advisable. The site offers space enough for the recorded 4000, at least on a temporary basis (if families are included, the ensuing density of 100 households per ha. is rather high). If the land allotments were of the reasonable order of 6 iugera (l.5ha., close to the minimum on which a family can comfortably live), then 24,000 iugera or 6000ha. of workable land was required, an amount which can be imagined as a block of land 10 km. by 6 km. While the total territory appropriated (much of it from Casinum) was at least five times this size, much of it was mountainous, or remote from the city. The land on either side of the Liri from the supposed boundary with Aquinum to the confluence with the Gari in fact supplies about the required amount, if the floodable river flats are deducted. If this was indeed distributed, then some colonists would have a journey of 6 or 8 km. to their fields. If, on the other hand, the allotments were smaller, access to common pasture for animals would be needed by way of compensation, and this would again involve journeying by some member of the household. Life can scarcely have been easy: we may marvel at the courage and constancy of the colonists, or suspect, probably with reason, that the whole 4000 never arrived and if they did, failed to survive the initial rigours. Livy's record of the figures could of course be at fault, as has sometimes been suspected. Archaeologically, this early phase is in one sense a blank. No even scatter of small sites with late 4 th or 3 rd century pottery has been found in the countryside: rural sites with pottery datable earlier than the 2 nd century B.C. are very rare, fewer than along the foothills. Moreover they form no particular pattern (e.g. occurrence at a regular interval

36

from the town), and some that are near S. Angelo in Theodice 22 may represent expansion from Casinum, not Interamna. Pottery of this early period is however found in some quantity in the city itself.n The alarms of warfare once over, the colony may have settled down to a phase in which land conveniently accessible from the town was farmed from there for several generations, while control over the rest may have slackened. That there was no inexhaustible supply of manpower is shown by the predicament reached in 209 (admittedly after a decade of war that had included Hannibal's march through the Liri Valley) when Interamna (like Sora and Suessa) was one of a group of cities which declared themselves unable to raise the required quota of men for the army. 24 Since the area of the city suggests a population no higher than 3000-5000 souls (assuming a density of 100 to the ha., or just over), support is gained for the idea that the original numbers had not been maintained. An alternative view is to suppose that the rural settlements of this period have simply escaped definition, because most were flimsy, many were short-lived, and associated pottery is too small in amount or too ill-known for recognition - or conceivably does not show up on sites which continued. Against this, it can be pointed out that at least one well-defined site of the later republic 25 was a farmhouse constructed largely in non-durable materials. The possibility that the earliest, hence lowest, strata of a site are under-represented in surface collection is however real. Certainly, it is the rural settlement pattern of the last two centuries B.C. that is most readily grasped, and it is hard not to see the record as testifying to an enormous amount of building in the countryside. One reason for this is that peace in Italy was now established beyond apy doubt, since after the Hannibalic war military activity took place on distant overseas fronts. External wars were however virtually continuous, and contributed to changes in Italy itself. A good 10 % of adult male Roman citizens were often absent on military service, and although the figures for Latins and Italians cannot be calculated, contributions to the total Roman army also fell hard on them. Peasant small-holders were the most vulnerable, and for every soldier who returned safely with booty for a prize, some family would be ruined to the point of losing their land and possibly leaving the region for the larger cities, especially Rome. Meanwhile the richer land-owning class of Roman citizens, serving as officers or even commanders, was well-placed to profit from the wars; they received an unfairly large share of booty and had the possibility of purchasing some of the almost countless thousands of slaves who came on the market. From the writings of Cato ("the Censor") it is clear that wealthy Romans might buy land in a variety of places (Cato himself had estates in Campania, but shows a close acquaintance with the Casinum region, 26 and may also have owned property there). They were also increasingly interested in exploiting estates in a profitable way, an · interest which Cato shared and to encourage which he wrote his treatise "On Agriculture", in which he advises on the profitable use of slaves in farming. The cessation in 168 B.C. of

tribute on Italian land may have been a further incentive. It was, then, a time when land was changing hands, when the richer was buying out the poorer, who might equally be threatened by incoming aristocrat or by a more fortunate neighbour from bis own class. Against this background, it is arguably not surprising that survey reveals something in the order of two to four probable farm-sites per km 2 • (with a maximum of 8-10), rather than the 50 or more that would seemingly be required to represent the colonial allotments of Interamna. In simple numbers of sites there are no glaring differences between the territories of Interamna and its neighbours. The picture changes considerably, however, if size and richness of sites is considered. Here a clear pattern emerged, and if we cannot grasp all the stages we can at least review the developed late republican form against its topographical background. Four basic types of landscape were available in the valley - the lower terraces and flats of the river valley narrowly considered, the so-called dissected area, sometimes gently undulating, sometimes awkwardly carved up by the steep valleys of the small streams draining into the Liri, then the flat plain, and finally the foothill slopes where hills and valley meet. A fifth, the usable uplands, was patchily present in the less steep middle slopes of the M. Cairo massif. The first was little favoured, even where (as happens especially near the Melfa) considerable areas were safely elevated against any risk of floods. The dead-flat areas of the plain were also not popular, though here a distinction should be drawn between the area near the Melfa River west of Aquinum, which showed little evidence of settlement, and the area east of Aquinum, between that city and Interamna, where a pattern close to average appeared, and the few blanks may be the result of modem development. West of Aquinum, settlement tended to gravitate towards the ridges and headlands of the rolling area south of the plain, rather less strongly dissected here than further east. Good drainage and outlook were clearly among the determining factors, and it was easy for north-south access tracks to run down the ridges. One particularly large and rich republican site sat astride a headland overlooking the confluence of Liri and Melfa: 27 a row of sites to the east of it (Figs. 22-23) hints at a road which crossed the combined rivers (by ferry, since this is not a suitable bridging point) to link up with another running south of the Liri. The same preference for ridges and (rather less frequently) headlands shows up around and east of Interamna. Variations include gentle slopes (especially south-facing) or the heads of small valleys, and there are one or two sites very near the Gari and Liri rivers, . perhaps again connected with crossing-points. Distribution was not however even throughout, but included the large blank area north-east of Interamna (Figs. 43-45), hard to interpret unless as uncleared land. This divided the settlement of this whole zone into two groups, one in the immediate confines of Interamna itself and another which looked to the Gari River as much as the Liri River, as if it had been peopled by settlers coming south

37

from Casinum rather than east from Interamna. Most sites of dissected area and plain give the impression of being relatively modest establishments, as do the settlements discovered in those areas of the uplands which were surveyed. Observations based on examples cut by the foundation trenches for modem houses suggest that many were simple farmsteads built largely of wood and clay, with only a little stone employed in the foundations; some smaller or less distinct sites may represent subsidiary buildings or huts, or have served some specific function such as that of well-house or small rural shrine. Perhaps one in five (or less, if all the indistinct scatters of pottery are taken to be sites) gave evidence from overall size, from the amount of fine pottery, from upstanding remains or at least worked stone blocks, of something rather more substantial, with greater use of permanent building materials. One upland site 28 is on terraces that could be ancient in origin. Only two - the one overlooking the LiriMelfa rivers confluence and another on a terrace east of the Gari River 29 - testified to rich decor in the form of mosaics, thus indicating the type of comfortable country establishment that a Roman noble might have considered appropriate if he were to spend time on his estates, whether to enjoy a quiet retreat from city life or to attend to their administration. This picture changes quite dramatically as the foothill slopes are reached, for here we find a row of richer establishments, 30 and the word villa rather than farm seems called for (Figs. 34-36). This is the word used by the late republican writers on agriculture for the country houses on their estates, though a modem audience req~ires cautioning on something that the ancients took for granted - that the villa was no mere pleasure-house built on a small plot of land purchased for the view, but the working centre of a surrounding estate. Provision for agricultural activities as well as for the refined tastes of its master could be combined with a degree of proximity surprising to the modem mind, which does not expect a smelly wine-press next door to a well-decorated dining-room. 31 From surface survey alone, it can of course be hard to distinguish a building which was luxurious early in its life from one that grew and attained to luxury at a later stage. As a general rule, rich country establishments did tend to grow both in size and ostentation. But the foothills offer two examples of sites32 which did not survive long into the imperial period, and were of relatively modest dimensions (some 500 and 1250 m 2 respectively) yet nonetheless were decorated with mosaics and had other types of solid pavement either for leisure (baths) or agricultural activities. Here then lived the wealthy, usually on the slopes just a little above the Pedemontana, but occasionally going to some distance to separate themselves from the inhabitants and occupations of the valley floor - witness the villa which later became the church of S. Angelo in Fortunula, 33 enjoying a truly magnificent view but perched so high that it could barely if at all be reached by wheeled transport, and a great terrace wall was required to give a

wide enough building platform. It is to be noted that the position of these villas ts m accordance with the advice given by Cato and later Varro on the siting of one's country home - preferably on a slope, facing south or south-west, and raised high enough above the valley floor to avoid noxious mists. 34 Indeed, it might be asked whether they had not this tract of country in mind, since it was known to both of them - even if Varro's villa just below Casinum was not actually in accord with his own best advice. The probable economic corollaries to this choice of site will be further explored below. Meanwhile, it is also to be noted that this greater rural display of wealth corresponds with the greater general prosperity noted for Aquinum and Casinum as against Interamna. This leads to further questions on the interrelation of towns and countryside, for it may be asked if the great upsurge in dispersed rural habitation took place at the expense of the urban population, or whether rural and urban prosperity in fact went hand in hand. The best opportunity to answer these questions is again provided by lnteramna, where the city itself as well as part of the territory was the object of systematic survey. 35 In the century between 150 B.C. and 50 B.C. the city area was occupied to its fullest extent, but a certain differentiation between one part of it and another can be observed. In the north-western half, evidence for a zone of industrial activity was found in the shape of pottery wasters and iron slag. While the artisans may have lived close to their kilns or forge, this area was probably deliberately kept separate from better quality habitations and public buildings sited further east. It is noteworthy that by the late republic there was a ring of sites around Interamna at no great distance from it36 (Fig. 43). Some landowners and farmers may have preferred to live there rather than in the town itself, and even this suburban area may no longer to any extent have been farmed from the city. Conceivably then the actual population of the town declined a little, or at best stayed steady, while urban-based economic activity increased in the course of a livelier interchange between the town and country. The wholly rural-based population presumably rose, though ignorance of the preceding phase makes comparison difficult. After the middle of the l81 century B.C., indications become curiously and interestingly diverse, even, at first sight, contradictory. On the one hand settlement of veterans is recorded, for all three towns, and the population of all three territories was thus boosted (lnteramna, where allotments seem to have been on an individual basis only, perhaps less than the others). The first two centuries A.O . is the period to which the bulk of the inscriptions belong, as well as a number of the major monuments of the cities: the ostentatious display of civic pride in certain conventional manners reached a peak. Yet the survey of lnteramna shows that the occupied area of the city was shrinking by the late 1•t century, and this shrinkage only became more marked as time went by.

38

This may denote the start of a trend which was to affect the other cities a little later, for Aquinum also shows considerable reduction in the imperial period, though neither so soon nor quite so drastically. Moreover, we tum in vain to the countryside for any obvious compensatory rise in population there. True, around Interamna there appears to be a slight increase in the number of smaller sites in the earliest imperial period, and this could perhaps be connected with the assignation of lands to individual veterans. In the Gari River valley, however, there is an overall decrease, especially of smaller sites. Near the Melfa there is also a slight ~ecrease overall (from 19 to 16), with a few new sites showing but rather more given up (these figures are of course subject to the bias inherent in a single survey, and cannot be taken as wholly precise). This is the more surprising in that the centuriation at Aquinum, and the official creation of a colony, suggest land re-allotment on a considerable scale. In the plain east of Aquinum, there is in fact a very slight increase in the number of minor sites (the probable modest farmsteads). Their location however bears no obvious relationship to the centuriation grid, unless Site 209, at the junction of two major axes (Figs. 34-36), is to be seen as subsequent to the centuriation. Along the foothills there is a slight decline in sites both small and large, though also one apparently new villa, S. Pietro in Campea, 37 overlooking the Melfa valley from a vantage point just west of Roccasecca. The next chapter of the story reads differently from area to area. The numbers of rural sites of 2nd-3rd century date remains steady around lnteramna, but their total drops tangibly on the plain and slightly in the foothills. In the Melfa area there is a slight rise, with some noteworthy new low-lying sites between the rivers and Fregellae. 38 The dissected area east of Interamna shows the same tendency much more dramatically, with the number of sites in the Gari area actually doubling, and again including some new low-lying ones not far above the flood-plain. The city of Interamna meanwhile had shrunk to less than half its original size, so that there is no sign here of any concomitant urban renewal. Thereafter, in the course of the 3rd century, the number of detectable sites drops dramatically everywhere. Only a mere handful of sites produce pottery recognizable as later than early 3 rd century--2 near Interamna, 5 near the Gari, 3 in the plains, 2 in the foothills and 3 near the Rio Moscosa. The surviving sites are mostly large ones, very few are new and some (e.g. in the Rio Moscosa 39) are notably long-lived, having persisted since Iron Age or early Republican times. While doubtless there are other sites unrecognized because of their lack of fine wares or other known pottery types, this evidence, especially when put alongside of the continued decline at Interamna and less dramatic decline at Aquinum, tends to indicate a drop in overall population, even if larger numbers of people were perhaps now grouped into the fewer, bigger rural sites. Until now, the types of settlement analysed have been at one or other end of a spectrum, with the towns or cities

contrasted with dispersed rural habitations which, it has been tacitly implied, were either single family homesteads or the property of a rich family, controlling a number of servants or slaves. Various nuances must be introduced. To start with, for the rich, rural life was not an alternative but an adjunct to city life, for the families who built the villas would be those who also dominated city affairs, and a town residence would claim part if not most of their time. For the peasant, however, whether owner or tenant of his farmland, the city would serve mainly as a market, to be visited as occasion demanded (except for any who still lived in the city and farmed nearby). Homely dwellings for the single family (perhaps with a slave or two) are usually considered the norm, with the more substantial farmstead of the wealthier peasant or local minor gentry differing only in scale, falling between the peasant's house and the rich man's villa. In fact, relevant evidence is largely lacking. Literary references tend to imply families living separately, but need not give a fair picture. Even with excavation, it is hard to determine the number of persons living in a building or building complex. It is certainly possible that farmsteads sometimes had multiple occupancy just as the large old farmsteads in the modem countryside often do. Moreover, one or two large sites without any signs of richness could be small hamlets; such are found especially in the Rio Moscosa area. 40 Readily distinguishable from the farmstead which perhaps housed three or four families are larger villages connected with roads, river crossings or simply occurring in the interstices between the towns. Survey recovered two clear examples. The first 41 lies on a ridge east of the Gari, across from S. Angelo in Theodice, on the line of a road which, it was argued above, was the original Via Latina. Some 750m. long and 100-150m. wide, it covered some l0ha. at its maximum extent, and began life in the 3"' century B.C. Just east of the Roccasecca Melfa River crossing, a larger, though perhaps rather sttaggling village 42 (Fig. 31) developed from a small Iron Age settlement or settlements. Extending in due course to some 25ha., it was almost as large in area as Interamna, but did not bequeath the same overall density of surface material. As seen already, this was the meeting place of three roads as well as a preferred crossing point of the river. The village here was dominated by a villa 43 which occupied a vantage point at the southern end, where massive terrace walls retained a squarish building-platform, some 50m x 50m., yet the occupation spread is far too large merely to represent an accompanying farmstead. The first village persisted beyond the 3ro century A.D. while the second, despite its favourable location did not, but rather shared the fate of so many other sites - any continuing occupation was of low order which has not left recognizable traces. Other places may have seen Roman villages, but all present problems. All have produced Roman inscriptions (as has Roccasecca) , but all are also the site of medieval and modem occup ation, so that the inscriptions could have

39

been brought from elsewhere as building material, while survey evidence either cannot be collected or tends to be inconclusive. S. Angelo in Theodice enjoys a favoured topographical position on an outcrop of travertine which forms a bluff and facilitates the crossing of the Gari River. The apparent lack of Roman material from immediately around the medieval castellum need not be conclusive, though whether villages could arise on both sides of the river remains a moot point. Pontecorvo, on its even more obviously commanding bluff, is another natural river-crossing point. Inscriptions found there, mostly re-used and some clearly prized for decorative purposes, could have come largely from Aquinum (with at least one from Interamna 44), though the presence of two dedications by a single lady 45 suggests local origin, not hapbaz.ard cartage. By contrast, the inscriptions from SS. Cosmas and Damian 46 are truly rural; they could simply reflect the estates of rich landowners, but the road to Fundi passes by, and the spot is remote enough from Aquinum - and even Pontecorvo - for a village to make sense. Likewise, S. Elia Fiumerapido, between Casinum and Atina, was most likely a minor centre, perhaps also S. Vittore. Most tantalizing of all is Piumarola, where at least one of the inscriptions found - a record of the magistrates of Interamna 47 - should logically come from the city. Survey (Fig. 39A) produced ample testimony of the medieval nunnery and village, but only a few sherds of Roman pottery. 48 If a Roman centre did exist here, it must either be below the heart of the medieval village, or on some other site to the east. Yet topographically the most attractive site is that of the present village, slightly raised above the plain and containing an ample sorgente, harnessed to the mill-wheels of later centuries and unlikely to have been ignored in ancient times. The discovery of the statue of a female divinity also points to an ancient presence, suggesting a sanctuary. On. balance it seems likely that Roman occupation at Piumarola was masked by medieval and modern, and details are not attainable. A further rural centre or centres, perhaps both village and sanctuary, was to hand south of Interamna near Ausonia, where there was an attested temple of Hercules. 49 Villages, hamlets and rural shrines, while not predominating, were thus all part of the more peopled landscape of Roman times. Dispersed rural settlement, however, became the norm, reaching a peak in the late republic and early empire, then fluctuating according to area until the overall general change of the 3rd century. This change may have involved more nucleation, in the sense that the few large remaining sites may have grouped more people together, but there is no corresponding growth in hamlets or villages - indeed, as seen, the Roccasecca village seems to fade out. Nor is any pattern of small, poor "satellite" sites around the main ones in evidence, but this could be the result of an absence of recognizable fine wares, and the possibility of their existence must be allowed. Nonetheless, it is bard not to see a drop in the overall population and a corresponding change in the landscape, in which ruined dwellings, overgrown tracks and uncultivated fields must have been

more numerous.

27. 28. 29. 30.

Notes 1. 2.

3.

4.

5.

6.

7. 8. 9. 10. 11. 12.

13. 14.

15.

16. 17. 18.

19. 20. 21. 21bis. 22. 23. 24. 25. 26.

Livy XLI.8.8. See C/L X.5387, 5384, 5382, 5383; G.I, Aquinum nos. 27 (pl. XV .1), 41 (pl. XX.3). Add Diana: ibid. no. 2 (pl. VI.3). [Note: coding for Giannetti refs. follows that employed by Giannetti in his 1973 art. (which becomes "G.IV"; "G.III" = 1971, 787ff.) .] Aquinum: see Cicero, Philippics II.41.106. Interamna: noted thus in undated inscriptions, CIL X.5332, 5338, 5339, 5411. Discussion in C/L X, 510. For Aquinum, see C/L X.5393, and Liber coloniarum, 229. The difference is partly one of scale--four towns in the Liri Valley share an area which is far smaller than that of an average Gallo-Roman civitas. On this, see E. M. Wightman, Gallia Belgica, p. 97f. B. T. Batsford Ltd., London, 1985. C/L X. 5176, 5196. Of Aquinum: CIL X.5415, 5481, 5502, 5539. Of Interamna: C/L X.5411 Aquinum: C/L X.525. C/L X.5405. C/L X.5399. Coarelli, F., "Note sulla topografia extraurbana di Aquino", Quaderni dell 'Istituto di Topografia antica della Universita di Roma 1, Saggi di fotointerpretazione archeologica (Roma, 1964) 5154. And, Crawford, M. H., L. Keppie, J. Paterson and M. L. Vercnocke, "Excavations at Fregellae, 1978-1984, Part III. VI. The Territory", PBSR 54 (1986) 40-68. See Appendix D. Crawford, M. H., L. Keppie, J. Paterson and M. L. Vercnocke, "Excavations at Fregellae, 19781984, Part III. VI. The Territory", PBSR 54 (1986) 40-68. Sites 363 (Appendix E), and INT. IE (see J.W. Hayes and E.M. Wightman, Arch. Laz. 6 (1984) 43, and figs. 1-2 on pp. 137-38). See Appendix D, Aq.Ua, north. C/L X 5183. 57, as against 245 for Aquinum, 302 for Casinum. (Count based on C/L, Giannetti, and Pantoni & Giannetti [1971].) Cicero, Philippics II.41. 106. Site 20. Site 474 . Site 494. Sites 296, 297, 341, 342. Appendix A. Livy XXVII.9.7. Site 228. Cato, De agri cultura 135.3; 136 [ = CXLIV, CXLV]. 40

31.

32. 33. 34.

35. 36. 37. 38.

39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49.

Site 20. Site 146. Sites 20, 522. Sites 80, 98, 119, 135, 142, 143, 153?, 162, 165?, 169?, 180, 193. . The example of Settefinestre in coastal Etruria comes readily to mind--see A. Carandini et al., Settejinestre: una villa schiavistica nell 'Etruria romana (Modena , 1985) 1.1, pp. 101-85 passim, esp. figs. 149-51. Sites 162. Site 193. Cato, De agri cultura 1.3-4; Varro, Res Rusticae I.12.1-2. (Cf., in more detail, Columella, I.iv .9-v passim.) For more details , see Hayes and Wightman , in Arch .Laz . VI (1984) passim. Sites 317, 354, 355, 364, 373, 376, 486. Site 98. Note also a paucity of sites of this period W. of Fregellae: Crawford, M . H., L. Keppie, J. Paterson and M. L. Vercnocke, "Excavations at Fregellae, 1978-1984, Part III. VI. The Territory", PBSR 54 (1986), table I on p. 51. Sites 86, 90, 91. Sites 86, 90 + 102, 92, 94, 100. Site 526 (see Fig. 43). Sites 113-18, 120. Site 119. C/L X.5411. CIL X.5383, 5384. C/L X.5370-72, 5378 . C/L X.5405. See Appen'dix E, Site 240, and Fig. 66. C/L X.5366.

Chapter 9

SOCIETY, ECONOMY AND THE ENVIRONMENT E.M. Wightman and J.W. Hayes

Any attempt to describe the patterns of settlement revealed by survey inevitably leads to questions of population and land use, some of which have already entered the discussion. Society, settlement, land use and the interaction between them can be more fully delineated if further sources of information are tapped. First, local inscriptions as well as the occasional literary mention allow some of the valley landowners to emerge from anonymous obscurity. Secondly, if modern analysis of the land forms and soils of the valley is coupled with the Romans' own attitudes to such matters, we can move towards a reconstruction of the likely land-use patterns, including crops (unfortunately, no pollen analysis from archaeological contexts is available to confirm or modify the hypothetical picture). We can also go a little further, if tentatively, to reconstructing estates and the pattern of land-holding. Moreover, further impressions of the general regional economy, and of the interaction between town and country, can be gained from a study of the pottery recovered from the survey. We cannot put names to the early colonists of Interamna, nor, for that matter, to the surrounding indigenous land-holders. Reasons for supposing that the recorded original figure of 4000 is probably exaggerated (and certainly so if it is taken to represent the permanent early population) have been discussed. The suggestion that normal patterns of buying, selling and consolidation of plots into larger holdings took place is further supported by the nomenclature found in the later inscriptions (which peak around the pt century A.D.). Non-Latin, Oscan names (more precisely, names that are less common among Latins than among Volscians, Samnites and Campanians) predominate over Latin, just as they do in neighbouring territories, and the most that can be said is that the predominance is a little less marked. Neighbouring local land-holding families were doubtless well-placed to acquire any abandoned colonial allotments, and the record suggests that their competition was too strong for the descendants of the colonists. The simple distribution of sites and blank areas slowly indicate, as has been seen, that the Gari valley was settled from Casinum rather than Interamna. The process cannot be followed in detail, any more than we can document the probable purchase of some of the more favoured land under the mountains by wealthy Roman citizens from Cato's generation onwards. It is evident from remarks in Cato and later writers that the land had a distinct appeal, and significantly Cato's local knowledge extends to the naming of a recommended rope-maker, L. Tunnius (a name not attested epigraphically) at Casinum. 1 Furthermore, in the 1•t century, a supporter of the dictator Sulla, at home in northern Campania, acquired (by fair means or foul) land

41

in the Rapido valley near S. Elia, where. an inscription records him and a freedman. 2 The l' t century B.C. polymath and writer on agriculture, M. Terentius Varro, was another incomer. The villa just outside Casinum, south of the modem railway station and close to the Gari, may well be correctly identified as his, for his property lay on either side of a river, the course of which was regulated and spanned by a bridge. 3 His reference to large herds of goats around Casinum" will have stemmed from close first-hand observation. We cannot tell the full extent of his land-holding in the area: it could have included other estates in addition to the one with his preferred villa, for ownership of multiple estates was not uncommon by this time. He is certainly unlikely to have held less than the 25-60ha. estates recommended already by Cato, 5 and could have had much more. Knowledge remains imprecise on whether a rich landowner might buy up parcels of land which were not in themselves farms provided with farm buildings, and which might be discontinuous with each other and with the main estate. Certainly, the writers normally seem to envisage the purchase of an entire estate complete with buildings, though a few indications may point in the other direction. For the purposes of organii.ation such plots could present a problem unless they were relatively close, or else large enough to be set up as independent farms in their own right, run by an overseer or tenant and producing what was needed to feed the staff. For this reason the acquisition of small unconnected plots by a big landowner may not have been the norm, but for the medium landowner or the aspiring peasant it may have been the only way to increase holdings, and thus a frequent practice at a social level which could not always aspire to unified, consolidated estates. Archaeological evidence, unless supported by documentary, can rarely, if ever, hope to recover the pattern of estates as opposed to that of settlement, although the latter may favour certain hypotheses over others. The Roman system of self-government by local communities essentially involved the major land-owners in any given area, for status and wealth were intimately connected and wealth was normally derived from (or invested in) land. Inscriptions recording magistrates thus form one obvious source of information, and the number known in the Liri River valley is conveniently increased by the partial list of the quattuorviri of Interamna, preserved at Piumarola. 6 At Interamna the Avidii, Barronii, Cavarii, 7 Caprii 8 and Pacuvii 9 stand out, and almost the same families, 10 minus the Cavarii but with the addition of the Fufidii (who married into the Avidii) 11 at Aquinum (the Barronii are shown by their tribe to be Aquinate). To

their number must also be added Decius, Iuvenalis 12 , almost certainly the satirical poet Juvenal who, while perhaps not a native of the region, held land and office at Aquinum. The most important families at Casinum were the Futii, 13 Luccii, 14 Paccii 15 and above all the U mmidii 16 who, in the early empire, must have towered above all others in the valley through their holding of consular offices, their high connections (including the younger Pliny 17) and their benefactions to the local community. Since a number of men held office in more than one city, confirmation is to band for the holding of multiple estates in neighbouring territories: the overlap is most obvious between Aquinum and Interamna, where the same families seem to have dominated both communities. Major figures from senatorial circles who became patrons of municipalities though offices such as the curatorship of the Via Latina 18 may also have owned land locally, though this cannot be proven. How far can individuals or families be further pinned down to localities or even to individual estates? The best example here is Juvenal, for the inscription, 19 a dedication to the goddess Ceres (mentioned as an Aquinate goddess in an apparently autobiographical passage of one of his satires'°) was found at S. Pietro in Campea, site of a medium-sized early imperial villa with adjacent terracing: set on the summit of a foothill, the villa had ample land suitable for olives both above and below, limited agricultural land immediately around but any amount below in the valley of the Rio Moscosa, which it surveys. Few inscriptions are so straightforward. In fact, if stones of somewhat dubious origin, such as those of Piumarola and Pontecorvo, be discarded, the number of rural inscriptions is quite small. Moreover, many of them record freedmen 21 rather than obvious members of the aristocracy. While such freedmen may have been buried on the estates of their patrons, they may also themselves have become landowners, in which case they testify, not to the ancestral estates of the patron families, but to the continual replacement of one population group by another in the countryside (as well as the towns). Although such factors introduce an element of uncertainty, it is still worth looking more closely at the families mentioned on rural inscriptions, and simply assuming that freedmen reflect fairly closely the families of their patrons, and that homonyms basically represent single families, or at least collateral branches. In fact, interesting patterns sometimes emerge. A magistrate of Interamna, a member of the Pacuvii, was at home in the south-eastern tip of the territory, on the way to Suio, near S. Andrea (south of S. Apollinare, and outside the zone of survey) .22 The name recurs in a rural context at SS. Cosmas e Damiano, 23 also south of the Liri River but in Aquinate territory. Another magistrate, M. Procilius, set up an inscription near Ausonia, 24 but since this was a dedication, it does not prove (while it may suggest) that he lived near at hand. Both these names could be of old indigenous families. The opposite is true of the Cornelii and Valerii recorded at S. Apollinare, 25

42

who must reflect either colonists (there are other Valerii at Interamna, 26 and the Comelii there include a duovir and a sevir Augustalis 21) or other later incomers (one Cornelius, known from a dedication from Pontecorvo, 28 is seen by his tribe and his Graeco-Roman bilingualism to be such). Some names, such as Cerrinius from Acquasanta just east of S. Giorgio, 29 are probably of local families (this one marrying into the Caprii), some otherwise unknown. The Poppaeus attested at Collenuovo:n in the Gari River valley could be local in the wider sense but the tribe suggests Aquinate origin: a connection with Site 457 is likely. That estates were sometimes named after their owners is shown by an inscription from SS. Cosmas e Damiano 31 recording that the Fundus Flamininanus was reached by a road which had to pass through the Fundus Babuleianus (the second name possibly local, the first clearly not). Few securely rural inscriptions come from the plains or dissected area, as opposed to the south bank of the Liri, the northern foothills or the Rapido River valley. An exception records an unfortunately nameless magistrate of Interamna(?), apparently at home just south of the Rio Poppeto (though the inscription must have been moved, since there is no site in the immediate vicinity of the farm into which it was built). One family, the Caprii, must have had holdings which spanned the valley from south to north. At one point the most outstanding family of Interamna, where they were magistrates and patrons, 32 they may have contributed a magistrate also to Casinum (where a branch of the family was certainly represented) in late republican times in the form of one who married a Cerrinia 33 (the name known otherwise only near S. Giorgio 34). A Caprius is also found at Piedimonte, 35 and if this inscription is-truly local it is tempting to connect it with the rich foothills villa Parito immediately to the east. Although the U mmidii and Luccii of Casinum escape location (perhaps their estates lay close to the city) the Futii turn up at S. Elia 36 and the Paccii at S. Vittore.37 The Blaii, another local name, are known from Casinum 38 and S. Lucia (in the form of a long list of their freedmen), 39 and are other potential candidates for ownership of the villa Parito. Yet others .are the Vicrii, also known from their freedmen, found at Piedimonte (or S. Lucia) 40 and Roccasecca, 41 while the Rasinii might be connected with the big villa Eucheria near Castrocielo. 42

A little information can thus be gleaned from literary and epigraphic sources about the valley landowners, but it has obvious limitations, even for the well-to-do whom it mostly illumines. Although we can see that many of the families mentioned were prominent in the l 1 1 century A.D., while a few are (from epigraphic criteria) probably or certainly later, we cannot follow their various fluctuations in fortunes. Nor can we be sure that the apparently diminishing number faithfully reflects processes whereby some families died out or became impoverished while a smaller remaining number grew richer, controlling more land than ever - though it is certainly tempting to suppose this, the more so since it is consistent with the

archaeological record. Inscriptions become markedly fewer everywhere after the early 3rd century, for reasons at least partly connected to questions of taste and priorities, but a connection between this and a change in settlement patterns is by no means to be ruled out. It would imply, essentially, that the type of competition for land and for local honours that had also expressed itself in the ostentatious use of stone inscriptions no longer took the same form: a few families were established beyond doubt, while the rest had fallen by the wayside to a definite position of inferiority. The successful families might not be local: the dominating family at Interamna in the early 5th century is known to have been the Sentii, 43 a branch of the distinguished consular Sentii, who were probably from Atina. To tum from documentary evidence to the natural resources of the valley environment may seem a great leap, but of course the latter contributed to the former. The various landforms and accompanying soils have already been described, and the implications partly seen. To the ancients, as to-day, the soils of valley were neither the best nor the worst available in central Italy. They were better than those of Latium around Rome, already regarded by Varro's time as exhausted. 44 Containing valuable nutrients from volcanic sources, they were still not quite so fertile as those of Campania, though Pliny implies that some land round Casinum ranked almost as high (he perhaps meant the Rapido valley). 45 As the ancients knew, general topography, which affected drainage, was as important as the actual quality of the soil, which they saw in terms of structure or texture, ability to retain moisture, general richness (or "leanness" ) and "temperature" (which is harder to interpret). 46 Without having a very systematic approach, they were perfectly capable of summing up the potential of the different parts of the valley - the patchy soils of the uplands, the red soils of much of the plain, the browner and often very gravelly soils of the dissected area and the alluvium (also often gravelly) of the · valley flats. What locals knew from experience, the knowledgeable incomer would judge from various observations and tests - the natural vegetation, the colour, the handling qualities, even the smell and taste. The richness of the foothills villas was not simply due to a preference for good views, though as Cato pointed out, a well-built house in a pleasant situation would encourage the owner to spend time there and so benefit the estate. 47 While the soils were often similar to those of the plain, better drainage combined with elevation above the valley mists was certainly an advantage, and the sloping alluvial fans particularly attractive. The main factors were however undoubtedly the accessibility of various land form types and of the Pedemontana. Two or three sites may be singled out by way of illustration. The villa known as "Eucheria" or "Euchelia" 48 (a name which may conceivably go back to the late empire) sits on an artificial terrace surveying the alluvial fan below Castrocielo. It has obvious direct access to the exploitable area of uplands known as la Forma. Though substantial (and perhaps

43

independent) farms might exist in the uplands, and some peasants may have been eking out a livelihood, it is reasonable to suppose that major land-owners controlled much of them and used them for stock-breeding (horses are found to-day), or at least for summer pasture . Varro, who had a strong interest in stock-breeding himself, comments on the large flocks of goats to be seen near Casinum, 49 but does not specifically mention sheep or cattle in the context of the valley. Meanwhile, the villa had at its disposal the well-drained land of the fan, and perhaps part of the plain below. A similar pattern is found at the villa Parito, 50 situated two-thirds of the way up an alluvial fan (Figs. 13B, 38). Here there were some smaller possibly dependent sites quite close above, on the way to the uplands. Below, on the fan (and perhaps near the track which linked the villa to the Pedemontana), was found the clearest example (because freshly ploughed) of a peasant hut which the valley has afforded (Fig . 6B). 51 Some 12m. in length by 3m. in breadth, it had been built mostly of daub over flimsy foundations of coarse gravel, but had nonetheless lasted from the 2nd to the 3rd century A.D. The probable remains of earlier predecessors lay near-by, and the dependencies here may have taken the form of a hamlet or hamlets. At both these sites, in addition to agriculture and pasture, sloping land suitable for olives was available, though any remains of ancient terracing have been masked by more recent work. At both S. Pietro in Campea and S. Angelo in Fortunula, 52 on the other hand, the foundations of the terrace-walling in the immediate vicinity are probably contemporary with the Roman building-platform. S. Angelo, unlike the others, had no easily available agricultural land, but could have controlled an area of the plain below. The ancient custom of growing crops, or grazing sheep, between the olive trees would further intensify the productiveness of such estates; the oil of Casinum attracted favourable mention, 53 and was marketable. While there can be no direct archaeological proof of the control of extensive valley land by these villa-owners, the consistent patterns conjoined with the dicta of the agricultural writers make it highly likely, and indeed the pattern has been preserved until modem times by families in Roccasecca and Caprile. Olives and livestock, and perhaps to a more limited degree vineyard s, would be the main source of the profits which encouraged their continued expansion. While some of the smaller republican foothills villas died out, others very notably flourished, Eucheria and Parito being sites which may even have seen continuity into medieval times. Below the slopes, the red soils of the plain were fertile, and would have been considered good for wheat and emmer, as well as some vegetables, root-crops and flax.54 Two problems rendered them less attractive and harder to work. Drainage was indifferent, and especially poor over the travertine areas. Especially when compounded with this, their high clay content made ploughing tricky. When wet, they become what Cato terms cariosa 55 (a word still retained in dialects), and ploughing when in this condition

could be highly deleterious to the soil structure necessary for good growth - Cato warns that up to three years' crops could be lost. This area would then be recognized as hard to work, and susceptible to the vagaries of the weather, though productive if good fortune was allied to care. Small and medium landowners might or might not prosper, and as a sole basis of wealth it was precarious. For a large landowner, it was ideal territory to let out to tenant farmers. Nonetheless, it contained a number of substantial farmsteads, some persisting in a manner that suggests the continued existence of the farm-unit, whether or not it retained independent ownership. South of the plain, the soils were a little lighter and better drained, if marginally less fertile, and the slopes less eroded than to-day. Some areas were however extremely gravelly. According to ancient prescriptions, barley and perhaps millet should have been preferred to wheat, while the steep hill-sides were suitable for vines, if they faced south. The discovery of probable wine-tanks on a site near Interamna 56 (Fig. 41B) obviously suggests vineyards, but there is no sign that the ancient wine of the region was, any more than to-day, of superior quality. The sandier and less gravelly soils of the area sloping towards the Gari - also well watered through the presence of springs - can only have made that area the most attractive part of the dissected area. Even in the area of the inscription of Poppaeus, 57 however, the sites are not of outstanding wealth but simply the substantial and minor farmsteads found throughout the dissected area. Areas of woodland suitable for the pasturing of pigs (e.g. in the area without sites north of Pignataro) should probably be added to the list of resources, and the small valley bottoms were available as meadow-land. What the streams of the dissected area gave on a tiny scale the major river valleys probably afforded on one that was larger and more profitable - meadow-land (its place to-day, subsequent to drainage programmes, largely taken by irrigated cash-crops such as maize). While Cato gave much weight to vines and olives, Varro the stock-breeder put meadowland first in his personal list of preferences, 58 and it may be surmised that this partly accounted for his choice of a river-side property just outside Casinum. For Casinum to have · the reputation it already had in Cato's day for ropes, 59 which were made from ox-hide (not hemp, though that too could well have been grown), this must have been long-established. River-bottoms of course posed the problem of flooding and the subsequent presence of marshy terrains or stagnant waters; it is significant that Varro in his description implies some form of containment. ro The valley near Casinum was also suitable to another type of enterprise - the irrigated garden - highly rated by Cato on the scale of profitability. 61 The lowest-lying land would not normally be used for habitations - though equally, had they existed, their remains would be covered by more recent alluvium - but near Interamna some minor sites are found just above flood-level 62 (curiously this does not happen in the area near the confluence of Liri and Melfa rivers where the

44

land, because higher, would have been drier). It would make sense if a number of these represented tenant- or bailiff-run farms specializing in cattle, yet just south of the river we have evidence for habitations and people of higher standing, witness the Cerrinius of Acquasanta (no freedman) 63 or the site Tromba delle Rose. 64 In fact, some of the valley sites could enjoy the advantages of access to different types of land noted for the foothills: this may be particularly true of those south of the Liri, which have direct access to woodland and hills as well as valley, albeit to north-facing slopes which can scarcely have been as productive as those on the opposite side. The increase in comparatively low-lying sites near Giuntura should imply some change in land-use in early imperial times, logically (if the foregoing is correct) a greater interest in stock-breeding. Some river-valley sites could also have practised more exotic forms of stock-raising - it is tempting to imagine fish-ponds beside the very low-lying villa of a senatorial family near S. Oliva (Site 575). While Varro's aviary near Casinum 65 was for pleasure (and has no necessary connection with a riverside), he also regarded aviculture as one of the most profitable involvements. 66 Less information is available on the southern flank of the valley. For the reason mentioned above, it was clearly less attractive than the other, and the generally steeper slopes and relative inaccessibility can only have compounded this. Yet clearly some families of local importance flourished here. To such sites as the limited survey undertaken here recovered, and to the epigraphical evidence, should be added the name Albianum, characteristic for a Roman estate-name (though no major family Albius is known), which is recorded as the earliest medieval form for S. Apollinare (survey failed to locate the expected villa, which may well be under the village). Perhaps large herds of pigs fed on the forests: at any rate the different terrains available must have been sufficient to support an established aristocracy, perhaps troubled less than elsewhere by competition from outsiders. While certain crops or activities might .be considered especially appropriate to specific conditions, there can never have been anything approaching monoculture. Even on a rich man's estate, given over as much as possible to profit-making pursuits, a variety of staple crops had to be grown in order to support the working staff. Meanwhile the peasant, of even the medium landowner, always had to aim at self-sufficiency in basic needs. Overall, then, while limited specialization could occur on properly situated large estates, the overall impression could only be (as it is to-day) one of mixed farming - arable land where grain crops alternated with legumes or fallow and a variable amount of vineyards and pasture for livestock. Fields would also vary according to the size of farm and its terrain, with small patches (again as today) being frequent. While archaeology is at best a very poor second to documentary evidence on sizes of estates, the settlement pattern revealed by survey can be pressed a little further towards providing reasonable hypotheses. The two areas

best suited to analysis are some 8 km 2 in the plain north of Interamna and some 16 km. 2 in the area between Pignataro and the Gari, both of which have a reasonable number of sites and are sufficiently far from the towns to avoid the bias which these might introduce. To the sites as presented on the maps should be added the information gained from conducting a re-survey of certain areas. This suggested that major sites were normally re-found, and that their number barely increased with re-survey. There can however be considerable variation in the reporting of so-called minor sites, which might show up well, poorly or not at all depending on conditions of ploughing and crops. Some "minor sites" reappeared only as tiny scatters of pottery which would not normally have been thought to represent a site at all. When all such observed formless "scatters" are plotted on a map, the number of sites roughly doubles. (While not all need be Roman, most probably are, and if we may assume that the number of non-Roman scatters roughly equals the number of Roman ones not observed at all, the one cancels the other.) To help translate these into possible landholdings, it is necessary to form an idea of what the "sites" represent in terms of buildings. The "major sites" are in fact almost certainly substantial farms, and sometimes have upstanding remains or traces of solid building materials to confirm this: sites in the upper range of this category are close to the villas, lacking only evidence for ostentatious wealth in the form of mosaics or luxury imported stones. The "minor sites" are more variable. They can include the "capanna" or peasant hut near Parito (Fig. 6B) which produced a fair amount of pottery 67 and also the remains of a larger homestead, a minimum of 15 x 12m. (as shown by the appearance of flimsy foundation walls in the trenches for a new house), where the evidence had been thinned out and more widely scattered by ploughing. Others may have been well-houses or rural shrines, but for the purposes of analysis we have to assume that, as a rule, they represent the homes of small farmers who worked some land on their own account, whether or not they owned any.

If these not wholly unreasonable assumptions are made, we find that on average there is one substantial farmstead for every 2km 2 • or 200 ha. The number of minor sites varies - 8-10 per 100 ha. on the plain, 4-5 in the dissected area. If we suppose that the substantial farms on the plain worked between 25ha. and 50ha., the corresponding averages for the others lies between I Oba. and Sha. In the other area, with a lower density, substantial farms of 50-100 ha. would allow averages of between l 7ha. and llha. for the minor sites. The differences may partly reflect recording techniques, partly ancient realities - the steep slopes of the dissected area were not all workable, and there may have been more woodland generally. While these figures only suggest a range of possibilities, farms of between 25ha. and IO0ha. (100 -400 iugera) lie within the range regarded as proper by Cato. Yet they did not give rise to luxury establishments - either they failed to prosper sufficiently, or for the most part they belonged to larger land-owners who did not make them their principal

45

residence, but installed a bailiff to run them. Smaller farms must have lain between the necessary minimum of a few ha. (perhaps split up into parcels) and respectable farms of some 20ha. Archaeological techniques cannot of themselves recover information on tenurial relationships. For that we need documents, and from Roman writers we derive the impression that the small peasant farmer always risked being swallowed up by richer owners, even if he was never rendered extinct. We know too that Cato's slave-run estates were only one option, enjoying a limited and temporary popularity because of particular circumstances, and that the letting out to tenants of land not conveniently workable from the owner's main estate-centre was probably the norm in the early empire. For the later empire, we are presented (even when due allowance for bias in the sources is made) with a great gulf fixed between a few mighty landowning families and a depressed, dependent peasantry. The archaeological picture, even if not illumining the subject directly, can be more or less consistent or inconsistent with the view derived from literary sources. On the whole it is consistent. The peasant settlers of Interamna were not found: even if this should be due to an inherent limitation in archaeological techniques, it implies that they did not prosper. The sites found show a variation in importance which can be credibly linked with sizes of farms. The importance of multiple ownership of estates for the generation of real wealth may also be implied, for it is reasonably clear (if the calculations above are close to the mark) that a single substantial farm of the size that Cato had in mind as a good purchase for an already rich man could not of itself generate obvious wealth. Put in another way, the holdings of the big families of Interamna should have comprised more than one such unit, unless we are to see them as much poorer than their equivalent in Casinum (on either count we are in fact left wondering how many wealthy families the territory of Interamna could support at any one time). The urge to buy additional land whenever possible is certainly comprehensible, and a trend to fewer, larger estates would follow. It is to the archaeological record that we owe the discovery that the 3 rd century was a crucial period of change in the countryside of the Liri Valley. The drop in the number of rural habitations is so steep that we are almost forced to ask if it is real or merely apparent. And indeed, one clear alternative is to hand - that some (even many) sites continued in occupation but without the types of fine pottery which would render them observable. This nevertheless in its own way implies a big economic change, either a major shift in the supply-routes of that pottery (now mostly from Africa), or else a drop in the purchasing power of the small farmer relative to earlier periods. Since the former problem would surely have rectified itself with time, the latter view seems likelier, and again is wholly consistent with the picture of poverty and dependence obtained from the literary and legal sources. A real degree of depopulation may well also be implied,

but is (and will probably remain) difficult to estimate with any prec1S1on. The interdependence of towns and countryside has already received comment. The flourishing state of the late republican towns, as typified by Interamna, went along with the type of attitude found in Cato, where, despite some attempts at self-sufficiency, the estate-owner normally went to the town for such items as ropes, baskets, agricultural implements and building materials (e.g. tiles). The evidence from Interamna suggests that pottery vessels, fine and coarse, were also distributed from an urban point of manufacture. 68 It is significant that production did not continue at Interamna. The only later kiln-site known (producing amphorae and a range of coarse vessels) was in the countryside, not even near a road. Though not connectable to any specific villa or farm, it must testify to estate production. Its range was very restricted, and other evidence for local variations in coarse wares suggests that rural production became the norm, and one economic function of the towns was thereby replaced. As the trend to larger estates proceeded, more and more goods could be produced on them, and the later Roman writers put considerable emphasis on self-sufficiency of estates, regarding it as undesirable that the staff or tenants should need to go to the town to make their purchases. This could only have an adverse effect on town life, an effect seen in extreme form at Interamna, and to a lesser degree at Aquinum. Unless the towns had some additional raison d'etre, they shrank, and the loss of their population must have been as drastic as anything in the cou~tryside. It is unfortunate that a clearer picture cannot be obtained of later Roman Casinum, for comparison of it with the other two towns would be of obvious interest. It is hard to avoid the conclusion that this was the richest of the three, endowed with the wealthiest and most influential local aristocracy - the Barronii of Interamna and Casinum hardly compare with the Ummidii, though they played a similar role. The manufacture of purple dye at Aquinum 69 may have brought wealth as well as fame to the city for a time. But it hardly seems an accident that most literary references are to Casinum, and that it is the oil, ropes and baskets of that town, 70 and not the others, which are mentioned. Clearly, it was regarded as the major stopping-point on the inland road to and from Campania, perhaps partly because the next stage, either to Venafro or Teano, was a long one. It also lay at a cross-roads of some importance, since a road came down from Atina, while the Gari and Garigliano valleys provided water-routes as well as roads to the sea. As a market-town, it thus had distinct advantages over the others, and probably more general contact with the outside world.

As it is, we have some evidence of the converse - that much of the area, and especially Interamna and its vicinity, was a comparative backwater, largely self-sufficient in rural produce and with remarkably restricted imports and exports. In general, both the city and the rural sites show an unusually low percentage of fine wares compared with

46

the norms for other areas. 71 For the republic, the absence of distinctive types of black-slip commonly found in Campania or northern Lazio is noteworthy, their place being partly taken by local products of lower quality (Fig. 56). The early imperial terra sigillata is however imported, from Campania rather than the north of Italy. North African sigillata of the first two centuries is found, but apart from lamp imports, as noted, become distinctly rare from the early 3 rd century. The story from coarse pottery is no less interesting. Here there is an abnormal absence of amphorae, 72 suggesting that there was not much transport of the products (mostly wine and oil) usually carried in these vessels (though the site of Interamna did produce some African oil amphorae of late Republican or early imperial date, Fig. 60). 73 The amphorae produced in the local rural kiln did not enjoy wide distribution. On the other hand dolia were common enough, indicating storage of local produce. Here then is a reminder that in spite of its relatively fertile soils and good communications, the valley retained a certain measure of isolation, being neither in the immediate orbit of Rome nor of the Campanian cities. It is a characteristic that has persisted until this day.

Notes 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.

1 l. 12. 13. 14. 15.

16.

17. 18.

19.

Cato, De agri cultura 135.3. C/L X.5282. Varro, R.R. III.5.9. Ibid. II.3.10. De agri cultura 1.7 (for 100 iugera); cf. 3.5, 10.1, 11.1. C/L X.5405. C/L X.5411. C/L X.5338-39; cf. also 5186(?), G I.1 (p. 61, pl. IV.2), P & G no. 14. C/L X.5347. C/L X.5399 (Avidii); 5400-1 and 5449 (see also G I, figs. 3-4), also P & G no. 22 (Barronii); 5461 (Caprii), 5378 and 5505(?) (Pacuvii). C/L X.5399. C/L X.5382; 5392-94. C/L X.5193-94; cf. also 5237-40. C/L X.5197-98; cf. also 5196. CIL X.5200; cf. also 5270, and G IV.7 (pl. II.5) Eph. Epigr. VIII.597. Same family at Castrocielo: G IV. 25. C/L X.5182-83, 5198; cf. also 5299-5304. For mention of them at Aquinum, see G IV .16 (pl. IV.4). Epist. VI.11; VII.24. Example: Dessau, /LS (I) 1159 = C/L X.5398 (Aquinum). For a consular patron of Casinum in the Caesarian period (no reason given), see G III. I (pl. I fig. 1)--Q. Pedius M. f. Note also G I, Aquinum 1 (mi-lestone of C. Calvisius Sabinus, consular repairer of the Via Latina in 38 B.C.). Cf. also /LS 1138 = CIL XI.2106 (patron of Clusium); /LS 1216 = CIL X.3732 (of Atella). C/L X.5382.

20. 21. 22. 23. 24. 25. 26. 27. 28. 29.

30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 30. 41. 42. 43. 44. 45.

46.

47. 49. 49. 50. 51. 52. 53.

54. 55.

56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68.

III.320. X.5372, 5375-76. X.5347. X.5378. X.5366. Comelii: C/L X.5354. Valerii: C/L X.5361. CIL X.5362; G I.3 (p. 62, pl. IV. l). CIL X.5405, 5341. CIL X.5385. G IV.10 (pl. III.2). Giannetti here compares a fragmentary entry in the fas ti, C/L 5405. Cf. n.34 below. CIL X.5277. CIL X.5370. C/L X.5338-39. CIL X.5186. G IV.10 (see n.29 above). C/L X.5459. C/L X.5240. G IV.7 (pl. II.5). CIL X.5221. C/L X.5452. C/L X.5541. C/L X.5540, 5543. C/L X.5516. On the evidence of CIL X.5348-49. Varro, R.R. I.9.5-6. Cf. N.H. XVII.25, 36, XVIIl.110; also XVIl.25, 36 - terra pulla; also Cato, De agri cultura 34; White, Roman Farming, Table on p. 82. Cato 6.lff.; Pliny, N.H. XVIl.25-34passim. See White, Roman Farming, tables on pp. 97-102 for a full list of references. De agri cultura 4. Site 135. R.R. II.3.10 (see n. 4 above). Site 180. Site 186. Sites 98, 193. Macrobius, Sat III.16.12 (citing Varro, Rer. Human. XI). But contrast the list in Varro, R.R. 1.2.6, where Venafrum is substituted. See Pliny, N.H. IX.165, after Varro, R.R. 1.23. Translation: White, Roman Farming 106. De agri cultura 5.6. Site 354. C/L X.5277 (Collenuovo). R.R. 1.7.10. See note 1 above. R.R. III.5.9. De agri cultura 1.7 (quoted by Varro, 1.7.9). Sites 419, 441-42. GIV.10. Site 476. Varro, R.R. III.4.2, and 5.8-17. ibid. III.4.1-2, and 5.1-7. See Appendix A, Fig. 63, Site 186. Hayes and Wightman, Arch. Laz. VI 143, with n.7.

Sat. CIL CIL CIL CIL

47

69. 70. 71. 72. 73.

Horace, Epist. I.10.27. See Cato, De agri cultura 135.3. For oil, see ref. n. 53 above. City: see Hayes and Wightman, Arch. Laz. VI 145-47, tab. III-IV. Arch. Laz. VI 145. Ibid. 145 and tab. II on p. 144.

Chapter 10

THE END OF TRANQUILLITY: THE TRANSITION FROM ROMAN TO MEDIEVAL CONDITIONS E.M. Wightman

For five centuries of the pax Romana the Liri Valley was an integral part of Italy which, as heartland of the empire, still at first enjoyed the exemption from taxes on land that dated from the Republic. Although the development of the area may not have been identical to that of other parts of Italy, there were certainly broad similarities, and while differences existed, both perceived and actual, these were at a_ IlllDlmum. When Augustus divided Italy into 14 regions (largely for census purposes?), Latium and Campania together formed Regio I, and were similarly linked to form one of the juridical districts which existed from the time of Hadrian on. 1 At first, the lower Liri Valley was perceived as forming southern Latium (and, since it also lay just within the 100th milestone from Rome, was seen as part of the central district around the capital). 2 Already before the 4lh century, the name Campania was applied to both areas, and when the division of Italy into provinces was instituted (or rationalized) by Diocletian, the relevant province was so named. 3 Only when the province of Samnium was created, later in the 4lh century, was there a hint that old divisions were not entirely forgotten since Venafrum and Isernia were now detached from Campania. 4 The regular correct ores or consulares of the province had already been foreshadowed by 3rd century officials who had virtually acted as regional governors. The Diocletianic reorganization changed the picture in another way, for Campania was now one of the suburbicaria provinces, still free from certain taxes and from the provision of recruits for the army, but regularly furnishing lime, timber, pigs, perhaps also grain and oil (or the equivalent in cash) for the city of Rome. 5 It is difficult to estimate the import of this change in status. Theoretically the need to provide a regular surplus could have stimulated agricultural production, rendering it more intensive and efficient - provided that the necessary manpower was available. Arguments for supposing that it may not have been had already been discussed. In addition, long before the imposition of taxes, the finances of the Liri Valley communities were (like many others) in a dubious state. Curators had already been appointed to oversee city finances (we know of one from Interamna who was native to Casinum) in the 2 nd and 3 rd centuries. 5 Inscriptions tell us that public buildings (e.g. the baths at Interamna) fell into disrepair, 6 though such problems could be an indication that wealth was diverted in other (private?) directions, not a measure of real poverty. It is significant however that the member of the high-ranking Sentii who restored the baths of Interamna in the early 5lh century also contrived to obtain a remission of one important tax. 7 This unpopular imposition, the 48

chrysargyron or collatio lustralis, 8 was instituted by Constantine on all merchants, traders and sellers of goods and services. Although the historian . Zosimus was certainly motivated by pagan prejudices against Constantine when he complained that the tax had ruined the cities of the empire, there may be an element of truth in this rather sophisticated analysis. Such a tax could only further encourage the decline of urban-based artisans, and lead to their seeking refuge on estates, where they could be better protected from the tax-collector. 9 The rural-based nature society of later Roman society, well-documented in a variety of literary and legal sources, was thereby strengthened by administrators who did not particularly wish it, and subsequently had to pass legislation in an attempt to force both landowners and artisans back to the towns. Until the 5lh century, the valley was troubled only by such general socio-economic ills or imbalances, and by the evils - occasional banditry, the oppression of poor by rich - that were endemic to the Roman ·empire, always underadministered and prone both to minor violence and to corruption at all levels. The first obvious wave of external forces, the Goths who, under Alaric, sacked Rome in 410, may have affected the area very little, unless (as is possible) they marched through it on their way south from Rome to Calabria. The later sack of Rome by the Vandals from Africa in 455 was part of a series of coastal raids on Italy, and need not have affected the internal parts, although the Vandals were caught by the emperor Majorian while returning with their booty to their fleet near the mouths of the Garigliano and Voltumo rivers. 10 Towards the end of the century, the de facto reign of Theodoric the Ostrogoth over Italy probably had few repercussions so far south of his headquarters at Ravenna. If a number of Goths were settled on the land, this merely continued the earlier practice of Odoacer. 11 Theodoric recognized the suzerainty of the emperor at Constantinople and maintained Roman administrative structures. Orderly rule thus continued, and one token of normality in the valley is the appearance of a bishop of Aquinum at a council in 487. 12 A generation later, S. Benedict founded his abbey at Montecassino, just within the old citadel walls: the fact that Casinum does not seem to have had a bishop may indicate that city life at the foot of the mountain had declined further than in Aquinum. A later chronicler 13 commented on the happy state of Italy under Theodoric, including the flourishing of trade: by contrast with what was to come, the evaluation was justified. But it is to be noted that pestilence as well as food-shortages is already noted for Rome in 467 /8 and 471/2, and that by 535 the problems of securing the requisite number of pigs from the

south of Italy were such that widespread immunities from the tax were granted. This latter may however have been due as much to the lack of cooperation of large landowners as to any absolute shortage of beasts, and, indeed, even earlier, the uncertainties of the imperial succession, which might lead to armed conflict (as when Ricimer besieged Anthemius in Rome in 471-2) must have rendered efficient securing of supplies very difficult. Cassiodorus, who describes the problems, and records food-shortages in Campania in 522-3, 14 also depicts a relatively flourishing countryside in some areas at least, although one in which the institution of rustic fairs on saints' days to replace regular periodic markets was already foreshadowing one typical aspect of the middle ages. The reconquest launched by the emperor Justinian and his general Belisarius in 536, a decade after the death of Theodoric, was certainly far more disruptive to society and economy than anything heretofore. 15 Strongly resisted by the Goths, it led to a whole generation of warfare, with atrocities committed on both sides as the Goths fought more desperately and Justinian's objectives hardened to include their extirpation. Campania saw its share of fighting and of passing armies . Belisarius' first speedy push to Rome took him up the west coast of Italy, leading to the siege and capture of Naples: thereafter he took the Via Latina rather than the Via Appia, 16 probably because it was the less obvious route. The Liri Valley also served as a corridor for part of a relief army which landed in Naples in the winter of 537-8. 17 Rome, one main enclave of Byz.antine power, was three times in Iittle over a decade (537-550) besieged by Gothic armies of Vitigis and Totila; the latter also occupied Campania along with the rest of the south, razed the walls of Beneventum and laid siege to Naples in 542, 18 collecting the taxes needed for his forces as best he could. Such irregular demands, whether for cash or kind, were bound to be worse than the steady provisioning of Rome. Moreover, the landowning class in general is known to have suffered, since some were taken and killed as hostages, others ruined. 19 Only after various peoples, including the Lombards, were invited to help against the Goths in 552 was a dearly bought Byzantine victory finally achieved. There would be much need for the recorded attempt to correct administrative abuses and, among other things, to purchase provisions for the army at regular market prices. Peace was short lived, for in 568 the Lombards came back under their King Alboin, this time as invaders bent on settling in Italy. Although the Byzantine empire's representatives retained Ravenna and . Rome, with a corridor between the two, the Lombards broke through to create the Duchies of Spoleto and Benevento within a year or two of their arrival?> Details are scanty, but it is clear that the lower Liri Valley now reverted to its old role as a frontier zone, as the exarch of Rome and the Lombard Dukes each tried to control it and, perhaps even more important, the more central line of communication between north and south that passed by Sora. Having had much less previous contact with Rome than the other Germanic

49

peoples, the Lombards were correspondingly more primitive in their institutions and attitudes. On the death of their king in 574 they were content to remain for a decade under the leadership of 35 duces or dukes, each effectively the leader of a war-band in search of booty. Gregory the Great, then bishop of Rome, describes (not without some rhetorical flourishes) the ferocity of the lowlands and the consequent desolation in both countryside and towns. Even after a king was re-instituted, the dukes retained a measure of autonomy. Depredations certainly continued. The abbey at Montecassino was sacked in 589 (the monks fled to Rome), and Aquinum within a few years of it. 21 Drought, famine and pestilence 22 completed the work of the sword, and bishop Jovinus had no successor, from want of potential flock as well as of shepherd. 23 By 595 Venafrum, which had earlier been under Roman control, was held by the Lombards and the merging of other bishoprics (Formiae with Mintumae, Cumae with Misenum) allowed by Gregory speaks of the poor state of other towns. 24 Nonetheless, Gregory could ask Arichis, Duke of Benevento, to help provide transport down to the sea for big timbers coming from Bruttium for the building of a church at Rome. 25 The lack of a clear administrative structure among the Lombards was much to the detriment of the cities which came under their rule, since they no longer .had to function as local centres in a coherently structured govemment. 26 The cities remaining under Byzantine rule, which included Naples and others on the Campanian or southern coasts, were in this respect better off, and accordingly remained in at least relatively flourishing conditions. On the land, Lombard measures were no less disruptive, for they included the killing or ousting of many Roman landowners, while others were forced to pay one third of their produce to support Lombards, or to hand over much of their resources . 27 A few place-names in the Liri Valley may reflect Lombard settlers, most notably locus Radupranndi, in the Aquino plain probably near Piumarola (other candidates are arcus Giz,z,ulus, perhaps also la Starza). Although historical sources dry up, so that the 7th century is a dark period in all ways, the chronicles (e.g. of Paul the Deacon) 28 of pestilence and of natural disasters continue. Because of a "great solitude" at the abbey of Montecassino, Franks from le Mans were able to make off with the relics of S. Benedict and his sister S. Scolastica, under the pretext of venerating them. 29 Extensive tracts of deserted land in neighbouring Samnium are recorded as having been bestowed on incomers by Duke Romuald. 30 The generally accepted picture is one of stark contrast between a few favoured ports (including Amalfi, which was probably created through the flight of Romans into coastal areas before the Lombards) and inland areas. Certainly, the lower Liri Valley was hardly fortunate in its situation. In the late ~ and the 8th century it remained a bone of contention between Rome and the Dukes of Benevento. Gisulf wrested Arce, Arpino and (more important) Sora from the Romans in 702, 31 and in 755 Liutprand marched

north against Rome itself, taking Ceccano (and doubtless other places) en route. Later Charlemagne, after his defeat of the northern Lombards (technically, of the kingdom to which some Dukes adhered only loosely) in 774 attempted to secure the return of Sora and Aquinum to Rome - or rather to the Pope, whose temporal power was greatly in the ascendancy in this period of weakening secular administration. It appears however that Aquinum continued to be controlled by the Dukes of Beneventum. The way in which the Via Latina was used during the Gothic Wars makes it reasonably certain that it was then still in fair order, although Procopius' encomium on the Via Appia, 32 oddly inserted at the point where he has just said that Belisarius did not take it, suggests that the latter was much more impressive (or simply better known). When Gisulf and Liutprand were marching north, was it still in such good shape? Cayro quotes an unconfirmed source 33 on the army of Ludovicus II in 866--per Roman veniunt ad Pomemcurvum, inde Capuam. If this is right, it implies that the Melfa could no longer be bridged, but that Pontecorvo already was. Editors (and Cayro) do not like this and have amended to ad pomem Cepranum--or at least they used to. But in fact I see that Pontecorvo was created in the 860's, and the bridge may have been new. If this is right, we ha\'e a nice pattern, with the Melfa problems deflecting traffic by Pontecorvo until the bridge at S. Vito (11 th or 12th c.) re-established something like the Roman pattern. Obviously, one can cite the continued use of the Via Latina near Aquino. But the S. Vito bridge caused a slight deflection to the west\l,ard stretch. There are problems with the Latina and the silicare campanina, which should represent one of the old centuriation lines which had a road. Probably there were two or three ways to proceed from Aquinum to Casinum in Roman times - the Via Latina, the centuriation road which became the silicate campanina, and a more southerly one by Piumarola. The worst of the Lombard disorganization was however by now over. Laws had been codified, and doubtless, helped by residual Roman influences, an administrative system gradually developed which gained the admiration of Charlemagne. In the S!.hcentury, signs of returning life can be discerned, the return of organized life can be discerned. By 718, the Abbey at Montecassino was once again properly constituted with an abbot at its head, and starting again on the path which was to make it one of the great landowners of the valley, a power temporal as well as spiritual. Also in the early 8lh century, it was capable of founding the nunnery of S. Maria at Piumarola, which likewise became an important centre. By the early 9th century, if not earlier, Aquinum had raised its head again, probably now on the ea.st side of the stream, occupying a spur between two of the lakes. With both a bishop and a Lombard gastald from c.850 it was the main rival to Montecassino within the valley. At the local level, squabbles over territory were now principally between these two, though when the abbey showed weakness (for

50

instance, after its sack by Saracens in 883), the Counts of Teano were also not slow to pounce. From this period, the records of the Abbey begin to throw light on valley settlement as well as on events, even if fitfully at first. Before following them into a new chapter and dealing with the questions which they raise, we may reasonably pause and consider the archaeological record of the 5th to 8th centuries. The archaeological evidence, at first sight, appears to depict a desolation as great as any described by Gregory. Only at Aquinum and in the rich villas Parito and Eucheria 34 do we find any positive evidence for occupation in the 6th and perhaps the 7th centuries. The real desolation was probably not as great as the apparent, but is extremely difficult to measure. The blank impression is created by the absence of datable forms of pottery, especially those coming from Africa: local coarse wares of this date may well be present, but cannot at present be recognized, given the absence of dated stratigraphic contexts. The absence of fine wares still has to be explained. In fact, the Vandal occupation of N Africa in the mid-5 th century had a slightly disruptive effect on production and on the patterns of distribution, at least for a time. While these later resumed something like their earlier form, the supply to all areas of the Italian peninsula may not have been assured. In some areas, notably Lucania, local workshops produced imitations of African fabrics to meet demand, but this does not appear to have happened in southern Lazio. The absence of these local imitations, more than that of the originals, lends credence to the idea of real decline in the area. Yet the picture is not wholly simple. At Interamna, where pottery certainly later than the 5 th century was not found and that of the 5th century only in a tiny amount, a number of Vandalic coins 35 suggests that relations with Africa had not entirely ceased. The coin record then continues into the 6 th century, with both Gothic and Byzantine issues, 36 so that occupation beyond the latest positively attested date for the pottery is thereby clearly proven. Any reconstruction of the period is thus badly hampered by the failure of the lower Liri Valley to produce recognizable types of local pottery. Yet some positive value can be extracted from this failure if we compare the valley with other regions which did so. Those with distinctive pottery traditions in this period fall into two categories, not mutually exclusive. The south, in addition to its local varieties of the originally N African tablewares, had by the late 4th century adopted the practice of applying red paint in broad bands on common buff wares. This tradition was followed by a number of local workshops and spread as far north as Capua and Naples, without reaching the Liri Valley. Painted ware then continued to be produced, though with some modifications in shapes and styles, into the 6th and 7 th centuries. Meanwhile , Naples itself also benefited from its continued contact with the Byzantine world, and imported pottery is found there later than in the inland regions. Here then we have a case of continuity of

earlier traditions. The Liri Valley by contrast apparently lacked both the local traditions which might have continued and the contact with the outside world which might either allow imports or stimulate later local inventions. However much this fact leaves us in the dark on settlement patterns, it does nonetheless suggest an impoverished or stagnant society and economy, with production and circulation of even ordinary goods at a minimum, and imports rare (there may of course have still been a trickle of perishable luxury goods, for those who could afford them). Even the ruralbased industries, it would appear, had virtually ceased, though wood-cutting presumably continued. At present then, only hypotheses can be offered. Nevertheless, in view of the problems that the area experienced, and its increasing lack of contact with the outside world, it can be reasonably suggested that there was indeed a fair degree of depopulation. Effectively, this would represent a continuation of the trends already inferred for the 3 rd century. While the agri deserti reported for Campania in the 51hcentury 37 are technically lands for which no owner responsible for taxes could be found, and not necessarily wholly unworked and unpeopled, it seems likely that both phenomena were often already present together by this time. Subsequent events would tend to introduce a further measure of discontinuity. Hard times were inevitably inimical to life on the plains, and would tend then as in every troubled period to lead people to take to the hills. Pestilence (which in the 61hand 71hcenturies included bubonic plague) and famine could have further reduced the population, even if the former spreads most rapidly in urban conditions and other areas of dense population. In addition to their propensity to dispossess Roman landowners, the Lombards may have also encouraged a further move away from agriculture towards pastoralism: certainly, their laws valued swineherds above peasant farmers. Possible direct evidence from the valley takes the form of only a few pieces of coarse pottery which could be datable to between the 6 th and 10th centuries (and which raise the possibility that sherds of the period may not be readily distinguishable from those of the Iron Age). Some of these in fact come from the upland area Pozzo dei Monaci. 38 Certainly, parallels from southern Etruria and from the Molise make it likely that some hilltop or upland communities would be established at this time. These would not necessarily be on the same sites as later ones, and indeed to mention one potential example, survey makes it unlikely that the ridge which later bore Roccasecca Castello and the forerunners of Caprile and Castrocielo was already settled so early. On the other hand, it is not unlikely that Arce had already moved up the hill towards the Rocca which formed the medieval, as it had the Iron Age, citadel. One site may well have seen a measure of continuity through the entire period, though evidence falls short of proof. This is the villa Parito, on the alluvial fan near Piedimonte (Figs. 13B, 38). In Roman times, as seen,

51

there appears to have been some form of straggly satellite hamlet or village just below it. 39 Pottery from the villa itself goes on into the 61hcentury, and a few coarse wares may be early medieval, representing continued occupation which was the precursor of the straggly medieval settlement (Site 181) which later covered the area southeast of the castellum of Piedimonte itself. The surrounding hills are readily accessible for refuge in times of need, and such temporary moves need not have entailed the abandonment of the site. Villa Eucheria 40 offers itself as another candidate, with similar advantages, and also with 61hcentury pottery, though lacking evidence for any form of nucleation. Unless it is merely an illusion created by the scanty literary sources, the 71hcentury probably was the darkest period. The demise of Interamna had very likely preceded the virtual abandonment of Aquinum. References to Casinum, as opposed to the abbey above, are non-existent, suggest that it too had sunk below the level of organized urban life of any kind. The destruction of the abbey meant that even it was not always available to organize rural life on the estates which were probably from an early date given to it. 41 Inhabitants of the area must have been tom between the need to stay on the land and work it for a reasonably predictable subsistence and the disruptive factors which encouraged flight. If a settlement such as Parito survived, it was probably at a rather miserable level. In view of its position, it may be doubted if Piumarola saw anything which may be described as continuity: certainly, organized forms of exploitation probably started afresh with the foundation of the nunnery in the 81hcentury, and the need for it to be supported from estates. It is not clear whether the archaeological record of Piumarola goes back to this period, although of all the medieval sites of the valley it has, along with Parito, produced the most distinctive and probably the earliest pottery.

In view of all the foregoing, it is remarkable that the earliest documentary evidence suggests that the actual pattern of settlement and the system of exploiting the land in the 8th century had not radically changed, but represented a continuity, if at a lower level, from those of Roman times. This evidence of course deals only with church lands, but suggests that the abbey's scattered holdings, with isolated settlements, were organized into a domusculta or estate organized from the new monastery of S. Germano, founded about 800 at the foot of the mountain on the site of Casinum. Although the estate so administered was a new grouping of property, the system is essentially that of the late Roman multiple estate. An element of continuity, however limited, had then survived the disruptions. Certainly, the changes in organization of settlement which were to produce the landscape of the high Middle Ages were still in the future, and were the result of initiatives taken in the face of new problems and needs.

Notes 1.

2. 3. 4.

5. 6. 1. 8.

9.

10. 11.

12. 13. 14.

15. 16. 17. 18. 19. 20. 21. 22.

23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38.

Pliny, N.H. IIl.56ff; Thomsen, The Italic Regions lff, 55ff, 153ff. Thomsen, The Italic Regions 69-71, 155-60 passim, 162-3. Ibid. 198ff. Cod. Theod. 14.4.3; Nov. Val. 35.11. Cf. Cod. Theod. 11.28.14, and 11.16.12; also Thomsen 213ff. CIL X.5197. C/L X.5348, 5349. CIL X.5349. Zosimus II.38; Zonaras XIV.3; Jones, LRE 4312, with n. 53. See Cod. Theod. 13.1.10 (of374), and Jones, loc. cit. for certain exemptions for rustics, if not negotiatores. Sid. Apoll., Carm. 5.385-430 [ed. Loyen I 43-4]-the locality is there not specified. Cassiodorus, Var. 2.16.5; Ennodius, Epist. 9.23; Jones, LRE 251. The matter is examined in detail by Goffart, BR 58-102. Also Procopius, Bell. Goth. I. 1.8, 1.28. Cagiano de Azevedo, Aquinum 18. Zacharias of Mytilene, Hist. Eccles. 16.10. Cass., Var. 8.33 (A.D. 527). Musset, Les invasions 99ff., Jones, LRE 275ff. Procopius, Bell. Goth. 1.14.6. ibid. Il.5.2. ibid. IIl.6.1-7. ibid. IV.34.3-8. Musset, Les invasions 138ff, Jones, LRE 308. Paul. Diac., Hist. Langob. IV .17 (misdated); Greg., Dial. 111.8. See Paul. Diac. IIl.24 for bubonic plague still occurring in 590. For the outbreaks from 540s onward, see Jones, LRE 288; Procopius, Bell. Pers. 11.22-23; Agathias 5.10; Chron. Min. 2.2134, 238. Greg., Dial. IIl.8. Jones, LRE 312. Paul. Diac. IV .19. Musset, Les invasions. Goffart, BR 176-89, re Paul. Diac. II. 32 and IIl.16. Paul. Diac. V.31, VI.5. ibid. VI.2. ibid. V.29. Paul. Diac. VI. 27. Procopius, Bell. Goth. I.14.6-11. Cayro, p. 34, quoting "Ap. Pratill. Ignot. Casin. N VI Tom. 1 p. 196 "--is this a later forgery? Sites 180, 135. Nicosia (1982) 95-97, pl. III, with comments pp. 81-82. ibid., passim. The general phenomenon is surveyed by Jones, LRE [II] 812-23. Sites 145?, 149?, 150? 52

39. 40. 41.

Site 183. Site 135. The later destruction of the abbey's archive denies us the details.

Chapter 11

THE LIRI VALLEY IN THE MIDDLE AGES G.A. Loud

The Early Middle Ages 700-950 The medieval history of the Liri Valley really begins with the resettlement of the monastery of Montecassino by a group of monks under Petronax of Brescia c.718, after the monastery had stood deserted for nearly a century and a half.1 This is not to say that at this stage the monastic community played a central role in the life of the valley, as it was so signally to do later. But, compared with the real "Dark Age" of the seventh century, for which both archaeological and written evidence is conspicuous by its absence, the restoration of the abbey provides a focus of evidence for the subsequent history of the valley. The monastery's chroniclers, beginning with Paul the Deacon in the Carolingian period, and the charters it generated, which become abundant from the middle of the tenth century, provide most of the written evidence for the valley in the Middle Ages. How disruptive the Lombard invasion had been is impossible to assess. Paul the. Deacon's description is rhetorical in the extreme: "churches were despoiled, the priests killed, the cities overthrown, the people who had grown up like crops annihilated "2 But the sacking of Montecassino and the collapse of the bishoprics of Aquino and Fondi in the late sixth century hardly suggests stability. 3 Our evidence is however so fragmentary as to be virtually impossible to interpret. Arce is recorded as being stormed by Duke Gisulf of Benevento in 702, 4 but whether this reference applies to Arce itself, thus suggesting that occupation had continued from the Roman period, or to the Rocca above, which would imply a retreat to the Iron Age promontory site, cannot be determined. To begin with the refoundation of Montecassino can have made very little difference to the valley. For some time the community was very small. When the English missionary St. Willibald visited it c. 729 he found only a handful of monks. 5 However a substantial donation of property followed from Duke Gisulf II of Benevento in the 740's. Quite what was comprised in this gift is a good question, for the only trustworthy account we have of it, that in the Chronica S. Benedicti Casinensis of the lateninth century simply said that Gisulf gave cuncta in circuitu montana et planiora. 6 The monastery chronicles by Leo Marsicanus c.1100 expanded this description to provide detailed boundaries for the gift, but in fact Leo took these from a tenth-century confirmation of the abbey's property, and the alleged text of Gisulfs diploma in the abbey's cartulary of 1130-3 was a forgery written by the cartulary's compiler, the notorious Peter the Deacon. The genuine charter had clearly long since disappeared. 7 But it was undoubtedly from this donation that Montecassino's 53

dominance of the valley of the lower Liri dated. That domination was never though absolute. The western part of the valley was outside the bounds of the monastic patrimony, the Terra Sancti Benedicti; usually subject, up to the twelfth century at any rate, to the rule of the gestalts, later, from 981/4, Counts of Aquino. 8 The contrast between the two halves of the lower Liri Valley is not the least interesting part of its evolution, though inevitably we are better informed about the history of the Terra Sancti Benedicti, since the monks of Montecassino were both more active in generating records and, despite the ravages of warfare, not least in 1944, in their preservation. In retrospect the monastic memory saw the period of nearly a century after Gisulf Il's donation as a golden era of peace and plenty. The twelfth century historian of the abbey of St Vincent on Voltumo, the other great monastic house of Lombard south Italy, commented (in a passage of considerable significance in the light of later developments that "at this time caste/la were rare in these regions, but everywhere was filled with villae and churches; neither was there the threat or fear of wars since all rejoiced in long-established (alta) peace, until the era of the Saracens". 9 But in fact the period of large-.scale Cassinese estate exploitation was relatively brief. Only in 787 did the abbey obtain a grant of immunity from temporal obligations from Charlemagne, and only under Abbot Gisulf c. 800 was a proper estate organization developed. 10

The centre of this administrative structure was the monastery of the Holy Saviour at the foot of Monte Cassino, the curtis maior. Around it was a group of at least nine monastic cells (Fig. 47), which were not merely centres of coloniz.ation but also of estate administration. In addition there was a nunnery subject to Montecassino at Piumarola in the middle of the valley, founded somewhat earlier than most of the cells' 1• The presumption must be that the cells of Montecassino functioned as the foci of a demesnial economy, worked largely by abbatial servi, who, if not quite slaves in the Roman sense, faced very severe restrictions on their liberty and mobility and could be transferred from one owner to another. 12 The Cassinese lands would thus have been organized very similarly to those of the abbey of Farfa in Sabina or the papal domuscultae north of Rome. But it is worth noting that the most recent study of these last has been sceptical as to how far they represented a significant reorganisation of the existing rural economy. It is probable that in the Liri Valley fiscal land was simply transferred to new ownership with its existing servile population, and it may well be the case that freedmen with

relatively light labour services and leaseholders also played a part in the cultivation of this land. The relatively late development of the cells would thus be·explicable first of all through a need to cater for an increased number of monks, and secondly as collection points for renders in kind rather than as centres for the cultivation of blocks of abbatial demesne. 13 Needless to say, given the state of the evidence any interpretation is speculative.

In any case the Cassinese curtis had only little more than a generation of peaceful development before southern Italy was menaced by the attacks of Arab raiders, some of whom had originally been imported as mercenaries by the rivals in the civil war which led to the break-up of the Duchy of Benevento after 839. The short-term effects of these piratical raids on the Liri Valley are difficult to measure. Certainly south Italian monastic sources furnish a woeful picture of destruction, but the hard information they provide is sufficiently sparse to suggest that the early raids were sporadic. Historians are now far more cautious than once they were in accepting highly-coloured monastic estimates of Viking depredations in NW Europe, and perhaps we should be equally careful here. But undoubtedly the raids were a disruptive influence. Arce was sacked, and the west_of the valley plundered in 844, and the Cassinese cells of S. Andrea, S. Giorgio and S. Stephano burnt not long afterwards. According to the Chronica Sancti Benedicti Casinensis the cell of S. Apollinare was only saved from destruction at the same time by a heavy overnight rainfall which left the invaders stranded on the other side of the river. 14 All these cells, it should be noted, were in the southern part of the valley, suggesting that they were the victims of a hit-and-run raid, or raids, from the coast. But by 867, when they sacked the cell of S. Elia, the invaders were penetrating up into the Rapido Valley. 15 These raids also led to an important development with the foundation of the castellum of Pontecorvo on a crossing point of the Liri to defend Aquino, which is described as a villa at this time, a term which implies that it was still without walls - and probably therefore the original undefended Roman town on the west bank of the Forme di Aquino. 16 Similarly Abbot Bertharius of Montecassino (856-883) fortified the monastery itself in modum castelli according to the later abbey chronicle. He also constructed a town (civitas is the word used in the Chronicle, not castellum, though it too may imply fortification) around the Curtis Maior of the Holy Saviour, which he called Eulogimenopolis or the city of Benedict.17 The function of this civitas is difficult to determine at a time of disruption, but one might suggest that it was intended as a refuge for inhabitants of the abbey lands displaced by the raids. If so, it was not a very safe one. In October 883 an Arab force sacked first the monastery of Montecassino itself, and ten days later that of the Saviour, and during the second attack Abbot Bertharius himself was murdered. 18 The destruction of Montecassino was the immediate result of the establishment of a fixed base for the Arab raiders at the mouth of the Garigliano. This menace remained until 915, when a coalition of local rulers under papal auspices, which even included a 54

contingent from By:zantine Langobardia (Apulia), extirpated the colony. 19 In these circumstances the monks of Montecassino remained in exile, first at Teano and then at Capua, and indeed the monastery was not rebuilt until the late 940's. But the valley was hardly left in a vacuum and the local nobility profited from the Cassinese exile. The gestalts of Aquino took over much of the abbey lands, the gestalts of Teano those from M. Trocchio eastwards. 20 And indeed the beginnings of the incastellamento of the valley can be traced to the lords of Aquino as much as to the Cassinese monks. The foundation of P~ntecorvo c. 860 we have already seen. It was probably in this period too that the site of Aquino itself changed, from the old Roman one on the west bank of the Forme di Aquino to a more defensible (and fortified) site on the east bank. Certainly the change had taken place by 954 when a charter referred to the "old city", and the late ninth century is the obvious time for such a move. 21 But these two fortified centres, and perhaps the civitas established by Abbot Bertharius at the foot of Monte Cassino, which the abbey chronicle implies as still in existence in the 940's, 22 are the only fortified sites for which we have evidence before 950. How far did the destruction of Montecassino and the sixtyseven year exile of its monks actually affect the settlement pattern? Ironically, although the fortified monastery on the mountain was sacked by the Arabs, the nunnery of St. Maria at Piumarola in the plain survived, to judge by a charter given to it by the By:zantine strategos (governor) of Langobardia in 892. 23 Similarly a charter of Prince Pandulf I of Capua of 961 refers to the churches of S. Ambrogio (the first mention which can be found of this dependency) and S. Andrea, S. Giorgio, S. Stephano and S. Apollinare, all of which had existed a century and a half before, and the first three of which had been sacked by the raiders. 24 But it is not clear whether these cells had a continuous existence since the 840's, whether they still played a part in Cassinese estate administration, or whether they were foci for settlement as they appear to have been in the early ninth century. The Lfri Valley in the tenth century was certainly a relatively undeveloped area, ripe for further settlement and exploitation. Hence in 918 Landulf I of Capua gave eight inhabitants of Sora a substantial tract of land on the east bank of the Garigliano at Bantra to cultivate and develop, and exempted them from all burdens to the state except that of military service . But two important observations must be made at this point. The document makes no specific mention of the creation of a settlement, let alone the construction of a fortified castellum. And whatever the problems which the monks of St. Benedict faced in retaining their lands during their exile at Capua, this charter did expressly mention property at Bantra belonging to the abbey. 25 Hence we should be cautious in accepting at face value the account of this period in the Montecassino Chronicle of Leo Marsicamus, which, whatever its earlier sources, was written c.1100-1105, a hundred and fifty years after the events in question. The chronicle records that not only the monastery but nearly all the surrounding

area was deserted as a result of the Saracen infestation, and almost nobody could be found there who owed obedience to the servants of God. 26 The scale of the desertion may well be exaggerated here, and the problem have been not that there was no one who ought to have owed obedience to the monastery, as much as that there were few who were willing to do so, or able to even if willing.

The Age of lncastellamento 950-1100 When the monks did eventually return to Montecassino there can be no doubt that Abbot Aligern (950-986) faced a formidable task in recovering the alienated property of the monastery, particularly from the gestalts of Aquino, and in recolonising and properly exploiting the reconstituted patrimony. The support of the Princes of Capua was crucial in forcing the local nobility to disgorge monastic property, both that of Montecassino itself and that of its dependencies. 27 The abbot attracted immigrants, especially from Marsia, to cultivate the valley, usually by a sharecropping arrangement by which they rendered 1/7 of grain and 1/3 of wine to the monastery. 28 It was as part of this process that the Cassinese castella originated. In 966 Aligern came to an agreement with a group of over thirty men to set up a castellum at the apparently already existing church of S. Angelo in Theodice on the west bank of the Gari, giving the settlers land in return for a money rent and 2/5 of the wine harvest. 29 The great contrast between this document and the princely charter granting land at Bantra in 918 (cited above) was that the construction of a fortified village was an integral part of the agreement. The abbot was to pay and feed the magistri fabricatores building the walls. A year later the monastery received a charter from the brothers Pandulf I and Landulf III, Princes of Capua and Benevento. This confirmed to the abbey "the castella and towers which have been built in the patrimony of the said monas~ery, or shall be constructed hereafter, namely the castellum of Rocca Janula, and that of S. Angelo in Theodice, and the tower of S. Giorgio, that the monastery and its abbots and rectors shall hold them firmly and securely in perpetuity. "30 These two charters are of cardinal importance for the history of the Liri Valley, marking as they do the beginnings of the major economic and social change of the Middle Ages, the effect of which was to last up to the twentieth century .

Incastellamento was not of course a phenomenon confined to the Liri Valley, indeed far from it. To a greater or lesser extent it affected much of Italy in the tenth or eleventh centuries, and in the last twenty years has become a major preoccupation of the historiography of the medieval peninsula. 31 The process saw the concentration of the rural population in fortified villages, around which the tenurial structure, agriculture and local administration of the countryside was arranged. Such villages were described in the sources interchangeably as castella or castra (except perhaps in Calabria where the word castellum usually referred to an urban citadel). 32 Put 55

crudely, the pattern of dispersed settlement which is seen as the norm in the early Middle Ages, with an economy based on large estates, curtes, worked by servile or semiservile labour, was replaced by a more concentrated settlement structure. Political instability, but also a desire for the extension of cultivation, on the part of the landlord as much as on the part of the tenant, led to this change in settlement geography. New villages were planted, on easily defensible and walled sites, and became the basis for the economy. Demesne farming, already in decline, virtually collapsed and in its place arose a complicated structure of lease, rent and sharecropping. · Many questions however remain. Within the broad outlines of the economic and social evolution sketched above there were very significant local variations, both of chronology and of extent. Some areas were transformed by incastellamento, others hardly affected. In the Tuscan highlands, for example, it appears to have been of minimal significance. Even within a relatively restricted area there might be very real differences in the nature and completeness of the change. On the lands of S. Vincenzo al Volturno, to the east of the Terra Sancti Benedicti, there was a significant contrast between the central part of the complex, where castella played a crucial part in settlement, clearance and economic organiz.ation, as described above, and the north-eastern area where, on generally poorer land, dispersed settlement and undefended sites remained more usual. 33 Such micro-topographical differences applied also to the chronology of change. On the Farfa lands in Lazio incastellamento commenced in the 920's. But elsewhere in Lazio the process was later: in the Tiburtina commencing from the 970's and not complete until well into the eleventh century, although by 1050 the pattern of settlement was very concentrated. But on the Subiaco lands to the east of the Tiburti.Ba incastellamento was late (after 1050), partial and largely defensive in purpose, not part of a process of concentrating the population. The causes too varied. Toubert's classic study of Lazio stressed the desire of the great landlords to recover, consolidate and exploit more efficiently their patrimonies, but Taubert himself has later admitted that perhaps he has underestimated the military function of the castellum. 34

In short, the whole phenomenon was extremely complex, and only detailed local study to examine chronology, causation and extent can provide meaningful answers as to its nature. In the Liri Valley too our sources are once again inadequate. Despite the extensive general documentation which has survived the vicissitudes of time at Montecassino, and even the destruction of 1944, the charter of S. Angelo in Theodice is the only example from the valley of a charter formally creating a castellum. Admittedly such documents are not very common in other areas either. Toubert cites only ten from ~zio. But there are some seventeen incastellamento leases for the lands of St. Vincent in the tenth and eleventh centuries in the Chronicon Vulternense, 35 and compared with this apparent plethora of valuable material the Liri Valley is poor

indeed. By contrast one has to rely on occasional isolated references in charters and on the terminology of the Chronica .... Casinensis, which althougk very detailed and of the utmost value as general source, based as it was one intensive study of the monastery's archives, is not without the danger of anachronism. Nor was its author interested specifically in the development of the abbey's castella references to them are purely incidental. Nonetheless answers are possible, and they are interesting. The scale of the incastellamento in the valley was considerable (Fig. 48). A papal bull confirming the possessions of Montecassino in 1059 listed no less than twenty castella on the abbey's lands. 36 As one would expect from the nature of the evidence not all these foundations, or in some cases fortifications of existing settlements, can be closely dated. But the process was certainly a relatively slow one; and though by 993 a princely charter could confirm all the lands and castella (in the plural) belonging to the monastery, 37 there was no flood of castral foundations under Abbot Aligern. The 967 charter of Pandulfl and LandulfIII cited above listed only S. Angelo in Theodice, the tower of S. Giorgio (implying that the settlement was not yet fully fortified) and Rocca Janula. And the last was., as its name (rocca) implies, a fortress pure and simple, guarding the road up to the abbey itself, and never a village settlement. The next castellum to be attested on the abbey lands is that of Bantra (just east of the Gari/Liri confluence, sometimes later called Bantra Monastica to distinguish it from Bantra Comitalis or Rocca d'Evandro) in 996. But a few years earlier (c.989) the second redaction of the chronicle described Abbot Manso restoring the church of S. Elia "destroyed by the barbarians" and settling men around it, but made no mention of the creation of a castellum. 38 Otherwise the evidence round about the year 1000 is scanty in the extreme. The civitas at the foot of Monte Cassino may perhaps have had a continuous existence since the ninth century, but under its new name of S. Germano is definitively attested once again, as a civitas, at the turn of the millennium. The chronicle also refers to a church being dedicated at the castrum of S. Pietro in Monasterio at this period. The mention may possibly be anachronistic: there is no other mention of a castrum here until the bull of 1059, though according to that document S. Pietro had of old been called the Castrum Casinum. 39 Quite how one should interpret this cryptic description is a good question. Problems of interpretation are also raised by the Chronicle's brief discussion of a revolt by the inhabitants of Pignetaro in 997. Here Pignetaro is called a municipium (fortress), but the context makes clear that it was a centre of habitation as well, and the passage goes on to describe how a church "next to the municipium" was burned down during the uprising. 40 Are we dealing here with a defended village, or some sort of fortification (like the tower of S. Giorgio in the 967 princely charter) around which inhabitants had congregated? Certainly c.1018 Abbot Atenulf established a garrison of mercenaries at Pignetaro, but the Chronicle confuses matters still further by calling it then an oppidum. Does this imply an open 56

settlement, perhaps with a citadel or tower, or one protected by walls? Similarly Cervaro and S. Pietro in Fine were described in the early 1040's as oppida. What does this mean? The Chronicle described Piedemonte twenty years later as an oppidum when charter evidence makes clear that it was a castellum, and it seems probable, if not entirely certain, that the two words were interchangeable. 41 Cervaro and S. Pietro in Fine were thus likely to have been fortified in the early 1040's. But S. Andrea was at the same time described as a villa, 42 and w~ hence still an open settlement, as were a number of other later castella of the abbey. Furthermore castella were not merely fortified villages but, and this is a point which Taubert and others have been at pains to stress, legal and economic units with their own territorial jurisdictions. In the 1040's Piedemonte was still part of the territory of Aquino (and still under the rule of its counts). And in 1048 Abbot Richer of Montecassino leased some property, including a church, to someone described simply as living ·"in the place called Vallefrigida" (modem Vallemaio), another site not yet a castellum, though it was to be listed as such in a papal bull only nine years later. 43 It therefore seems probable that most of the newly-settled or resettled sites in the late tenth century were as yet unfortified. The abbey lands were protected more by the close relations of the abbots with the Princes of Capua Abbot Manso (986-996) was a kinsman - than by the fortification of all the main centres of habitation. There was little trouble with the lords of Aquino, the other main power in the valley, for a generation after a punitive expedition by Prince Landulfll in the 950's. 44 When fresh dispute flared in the 990's, at a time when the protective power of the Princes of Capua was at least temporarily in eclipse, Abbot Manso built a fortress at Roccasecca on the slopes of M. Cairo, but this was a castle pure and simple and not intended as a centre of settlement. The original site was surely the rocca (now ruined) on the hill overlooking the later, and still-existing, village. Two other castella are known at this period: Castrocielo, another site over the 200 metre line on the slopes of M. Cairo, and Interamna on the plain , both apparently built by the Counts of Aquino in the 990's (the dating is less secure for Interamna), and both very near the boundary between the county and the lands of St. Benedict. Castrocielo may have begun primarily as a fortress, intended as a counterweight to Roccasecca, though the latter was itself soon to fall into the hands of the Counts of Aquino. 45 The extension of settlement in the county of Aquino is much more difficult to follow than in the Terra Sancti Benedicti. Pontecorvo would seem to have developed considerably in the tenth century. By 994 it was the seat of a separate county, ruled by a son of Atenulf Megalu, the first Count of Aquino . By 1004 it was being referred to as a civitas. 46 At this time the area south of the Quesa was described still as "deserted and waste", despite the excellent water supply with the river fed by three sorgenti .

The first settlement in this area was by a small group of refugee Greek monks, who in 998 were given a substantial block of land and permission to build a monastery there by Count Guido of Pontecorvo. This church, S. Pietro di Foresta (the suffix is indicative) was given to Montecassino in 1075. 47 Another monastery, dedicated to St. Paul, was built a little way to the west on a hill site in the early eleventh century. 48 The area along the Liri north-west of Pontecorvo was opened up in the 1030's, or perhaps more correctly by the 1030's. A castellum was attested at Rio Matrice in a charter of 1030, though in the twelfth and thirteenth centuries this site, though inhabited, was not considered to be of castral status. A monastery was founded at Bagnerolo, on a hill overlooking the river, in 1030. The same charter first mentioning Rio Matrice also refers to the castellum of S. Giovanni Incarico, and another document mentions a mill in this area in 1039.49 The monastery of S. Nicola in Pico, near M. Leucio, is attested in 1050 when it was given to Montecassino by the Count of Pontecorvo, and the first reference to the castellum of Pico (some 5 km. to the west of the monastery, and south of the Quesa) comes in 1051. 50 Further to the north, but still in this western part of the valley, there was cultivation near the Via Latina bridge over the Melfa, perhaps encouraged by the foundation of the monastery of S. Vito nearby, though some of this area was still, at this date, wooded. 51 Thus the valley in the early eleventh century saw the development of new settlements (Fig. 48) and the expansion of cultivation, almost certainly reflecting a rising population. But the process of incastellamellto was as yet slow and incomplete, and the majority of the valley's settlements probably still unfortified. What led to a dramatic change in this picture was the arrival of the Normans. Coming to southern Italy first as pilgrims c. 1000, from about 1017 they began to serve as mercenaries for the local rulers, and · indeed for anyone else who would pay them. The soldiers with whom Abbot Atenulf garrisoned Pignetaro c.1018 were Normans, but as yet the mercenaries were few in number, and it was not until the 1030's that their activities began seriously to impinge on the security of the Liri Valley. But from the early years of that decade their plundering became such a menace to the lands of Montecassino that c. l 045 Abbot Richer ordered that all the main settlements on the abbey lands be fortified and that the peasants who lived in outlying villae should reside in them. 52 We have thus an unequivocal statement that up to these dramatic events the settlement pattern was relatively open. But we are not dependent on this one statement alone. Prior to 1040 the only sites within the Terre Sancti Benedicti for which there is definite evidence of castral status are S. Angelo in Theodice, S. Giorgio, Bantra, S. Elia (and that for the first time in 1039), plus the possibly doubtful case of S. Pietro in Monasterio and probably Pignetaro (to which should be added the civitas of S. Germano and the fortress of Rocca d 'Evandro, the latter built before 1022).53 Yet by 1059 57

there were twenty castella on the abbey lands. S. Andrea and S. Vittore were seized and fortified by the Normans themselves, though later recovered. 54 Cervaro was described as an oppidum in the early 1040's, whatever that implies, but as a castellum a few years later. The walls of existing castella were repaired, especially at S. Angelo in Theodice where they were largely ruined. 55 Outside the Cassinese lands Mortola is found as a castrum c.1045, though the site may have been inhabited for some time since by this period there was already a road leading from Mortola to Cervaro, and Piedemonte was named as such in a charter of 1052. 56 The 1040's were thus the crucial period in the incastellamento of the Liri Valley, and it was the destabilisation of the area by the Normans which caused this. In contrast to Lazio, where for the most part (Sabina excepted) the castellum served a primarily economic function, in the Liri Valley defensive considerations were uppermost. Although there was a determined effort to further the economic development of the Cassinese lands after 950, many of the centres of habitation remained for a long time unfortified, and the population was at least partly dispersed. It was only in the 1040's under the stress of Norman aggression that the situation changed. In this decade the Normans were a real threat to the safety ahd stability of the valley, and of the Cassinese lands in particular. Pandulf IV of Capua was also hostile to the monastery; indeed the earliest depredations of the Normans had been by those employed as Pandulfs mercenaries, who had for example garrisoned S. Germano c.1034. 57 Abbot Richer however took effective defensive measures. In addition to initiating the programme of fortification, he developed the milites abbatiae to provide an effective force of troops. Some idea of the size of this can be seen from the fact that in the mid-twelfth century the abbey was expected to provide 60 knights and 200 sergeants for the army of the King of Sicily. 58 But the alliance between Abbot Desiderius (1058-87) and the first Norman Prince of Capua, Richard (1058-1078), was to bring about a major change, and to benefit the abbey greatly. Not only was princely protection restored once again to the abbey, but in the years 1065-6 a series of gifts at the expense of the local Lombard nobility who had opposed Richard extended the bounds of the Terra Sancti Benedicti, giving Montecassino the castella of Mortola and Interamna in the valley, that of Fratte (modem Ausonia) further south, and the monastery of the Holy Saviour on M. Cocuruzzu. Perhaps most significantly, in March 1066 _Prince Richard gave Montecassino land at the mouth of the Garigliano, and hence a corridor to the sea. 59 And in 1067 the abbey acquired Piedemonte from the Counts of Aquino. ro The sites of the Cassinese castella in the Liri Valley give eloquent testimony to their function. A few can be found on the plain in relatively open positions. The central part of the valley north of the Liri was simply too flat for there to be·obvious natural defensive sites. But the presence of castella in the particular places where they are can usually

be attributed to special factors. Piumarola was built around the already existing nunnery there, although at some stage before the thirteenth century this monastic house had fallen into abeyance. 61 The presence of substantial Roman spoils may well have facilitated the building of Interamna. Just south of the river the abundant water supply from the springs of S. Giorgio was probably the spur to development there. But the majority of castellum sites were chosen for their defensive possibilities. The bull of Pope Victor II confirming Cassinese property in 1057, which gives the first full list of the abbatial castella, recorded that many of these were in the localities of former cells destroyed by the Saracens (though as we have seen earlier some of these churches may still have existed in the tenth century). 62 But, though the locality was the same, the actual site had often moved. The castella of S. Andrea and S. Apollinare were both about 1 km. from the sites of the former churches of those names. The cell of S. Ambrogio was closer to the river than was the later castellum. The site of S. Andrea moved with the foundation of the castellum and not before. It was described as a villa in the early 1040's, and the site which local tradition assigns to the old church, which is, as said, about 1km. away from the castellum (and modem village), is known as Villa di S. Andrea. Furthermore the Montecassino chronicle mentions the arx (citadel?) of S. Andrea at this same period. It was probably the site of this arx which was used when the castellum was built in the mid-1040's. 63 A characteristic example of a castellum site is that of the now abandoned Mortola (Site 565). This was set on an oval headland to the east of the Garigliano, roughly 150 metres by 50 metres, with a narrow neck across which the remains of a wall are still visible. Otherwise the site is surrounded by small but steep valleys. Similarly Bantra (Site 564), one of the earliest castella, established before 996 and abandoned by the mid sixteenth century, lay on a high terrace, 180 by 150 metres, overlooking the Garigliano near its junction with the Liri, with a sheer drop to the east and steep valleys to north and south. The remains of a curtain wall and towers stand at the east, reinforcing the de'terrent effects of the bluff. Both sites are ideally suited for defence. S. Stephano (Site 77), first listed as a castellum in 1057, occupied the south part of a saddle-shaped ridge (c. 90 by 50 metres), with a bluff on the north side facing the Liri, and steep slopes to the east and south-west. Among the castella which continued to be inhabited after the medieval period is Piedemonte, on a promontory with a narrow neck some 100 metres above the level of the surrounding plain. The papal bull confirming the abbey property in 1057 lists eighteen castella. That of 1059 adds two more, one of which (S. Andrea) had been in existence for some years. In 1061 Abbot Desiderius built Castronuovo to guard the Terra Sancti Benedicti against the attacks of the inhabitants of Fratte (which was itself given to the monastery four 58

years later). 64 The only new castellum found after that date (and not very long after) was outside the Cassinese lands: Roccaguglielmo, the first reference to which occurs in 1067. Two other settlements developed in the shadow of existing fortresses, Roccasecca and Rocca d'Evandro (Rocca Bantra). Both were inhabited by the early thirteenth century, but in neither case is it known when settlement started. 65 Nevertheless, despite these slight reservations, it is clear that the age of castral foundation was effectively over c.1060. Henceforth the expansion of settlement in the Liri Valley was to be away from the fortified centres.

The Valley in the Central Middle Ages 1100-1300. The Montecassino Chronicle suggests that there were already smaller settlements in the 1040's, when Abbot Richer ordered their at least temporary abandonment as a defensive measure. Similarly, in 1106, when relations between the abbey and the Counts of Aquino were at another low ebb, Abbot Otto once again ordered the peasants of the Terra Sancti Benedicti to abandon their villae and congregate in the castella. 66 Throughout the Middle Ages there was a tension between the needs of cultivation and a growing population, which necessitated or at least encouraged a spread of settlement, and the needs of security, which encouraged the concentration of settlement in defensible sites - the castella. However the development of more effective and stable government allowed the demographic increase and the demand for more cultivation priority. The decline in the authority of the Norman Princes of Capua after the death of Jordan I in 1090 was most marked precisely in the north of the principality where the Liri Valley was situated. Hence the lands of Montecassino came under renewed pressure from the local nobles in the early twelfth century. 67 The unification of southern Italy was consolidated under the rule of the Kings of Sicily after 1130 only after a long and sanguinary civil war, which in particular wreaked havoc in the Liri Valley in 1139, when S. Angelo in Theodice, Mortola, S. Vittore and S. Pietro in Fine were all sacked. 68 The unification marked however a decisive change. Henceforth domestic peace was to be the rule rather than the exception, and if this did not necessarily favour powerful vested interests like the abbey of Montecassino, which found both its control over its roccae and its judicial immunity clipped, 69 it was undoubtedly a boon to the local population. On the Cassinese lands the latter also benefited from the grant of franchises protecting their rights and controlling the activities of abbatial officials. 70 There were however two periods when the valley was seriously affected by the rigours of warfare. The first was in the 1190's when first the succession to the kingdom of Sicily was disputed, and then after 1197 effective government was hamstrung by a long royal minority. The second period was much briefer, but nonetheless damaging in its effects; the invasion of the reg no by .papal forces in 1229. And in these periods even the castella were not necessarily very effective protection. During the 1190's S.

Pietro in Monasterio was plundered in January 1192, Pignetaro burnt in 1196, and in the early months of 1199 the troops of the German adventurer Markward of Anweiler burned S. Pietro in Fine, Cervaro, Trocchio, S. Elia and Piumarola, and plundered S. Vittore and S. Germano. 71 Fortunately for the local inhabitants the valley seems to have escaped relatively lightly thereafter, though southern Italy as a whole remained in a highly disturbed condition for several years more. The 1229 campaign was brief, but here the valley was the principal battleground for the forces of Pope Gregory IX and Frederick II. The invading papal army captured and burned Arce, and then apparently marched through the valley on the south side before crossing the river and heading for S. Germano. On their way they plundered Rio Vitellato and the churches of S. Pietro and S. Paolo di Foresta, and then burned Interamna, while Pignataro and Piumarola were abandoned by their inhabitants. Piedemonte was saved from destruction only by the frantic pleas of the Abbot and monks of Montecassino. During the counter-attack by the Sicilian forces which followed the villa of S. Lucia on the slopes of M. Cairo was destroyed. 72 But for the most part the thirteenth century was peaceful. In 1266 Charles of Anjou and his army passed down the Via Latina towards Benevento, but with the exception of a sharp skirmish at S. Germano there was Iittle fighting in the valley. 73 It was therefore in an environment that was largely peaceful that settlement spread beyond the castella, though as we have seen this could occur (fuelled one suspects by demographic imperatives) even at an earlier and more disturbed period. By 1157 the property of the bishopric of Aquino included two farms and the villa of S. Croce (Site 306, east of the Forme di Aquino and 2 1/2 km. NW of Interamna). The same document also refers to episcopal rights in all the vici between Pontecorvo, Roccaguglielmo and Pico, suggested that by this date dispersed settlement was quite common in this part of the valley. 74 Very few of these smaller sites make an appearance in the documentary sources, or can now be identified by name. However, a papal taxation list of 1325 mentions five casales (that is open villages) in the neighbourhood of Pontecorvo and Roccaguglielmo: Rio Vitellato, S. Pietro in Curulis, S. Paolo di Foresta, Tordoni and Teyrano. 75 These casales, like the castella, were administrative and legal units. They suggest an important contrast between the Montecassino lands and the western part of the valley. That some of the more important settlements in this latter area were unfortified suggests either that they were founded relatively late , or at least that their development as settlements of any size was slow. Of these five casales only Rio Vitellato can be clearly attested as an inhabited centre in the eleventh century. 76 The lack of defences would certainly be explicable if these sites had only emerged as significant centres after 1140 and the unification of southern Italy, when a strong central authority was active in suppressing disorder. It is also highly unlikely that the Kings of Sicily would have permitted the construction of fortified positions since they zealously restricted the regalian right of fortification

59

(indeed in 1224 Frederick II even contemplated the demolition of the walls of S. Germano).n A potential difficulty with this hypothesis is the case of S. Pietro in Curulis, the name of which is carved on the bronze doors of the abbey church of · Montecassino, produced in the early 1120's. The church of S. Pietro certainly existed as early as 1059, but since it was then specifically mentioned as a church one may doubt whether there was at that stage much of a settlement around it. When the first documentary reference to a settlement does come it is as late as 1222, at which time S. Pietro in Curulis was called simply a villa. 78 Only in 1270 was it described as a casale.79 Hence the terminology of the fourteenth -century papal taxation lists may well reflect a reorganisation of the rural administrative structure in this part of the valley during the preceding century. Of the other sites S. Paolo di Foresta (Site 574) developed around the monastery of that name, and the casale was probably in existence by 1215, when it is known that the monastic observance had been abandoned there. 80 All these sites except S. Pietro in Curulis were later abandoned , indeed that of Teyrano cannot be identified, through it was apparently near Pontecorvo, and the location of Rio Vitellato on the map is conjectural. 81 Of the other identifiable sites in this area documentary evidence comes only in the late thirteenth century (Fig. 49). A settlement developed around the monastery of S. Pietro di Foresta as well as around its sister house of S. Paolo. The village of S. Pietro is first attested in 1273. By this time the monastic community had probably disappeared, indeed this may well have occurred before 1225. In 1275 the church of S. Pietro was attached to the infirmary of Montecassino, and in the mid-fourteenth century it was made subject to the Cassinese sacristy. 82 In 1296 there was a legal case between the communities (universitates) of Pontecorvo and Roccaguglielmo over the possession of the three villae of Rio Vitellato (later named as a casale in 1325), S. Oliva and S. Tamparus (the site of which has not been identified). All three were confirmed as belonging to Pontecorvo, but the document also noted that the casale of Montevetro (Site 578) and another villa, Valle Ii Morini (that known later simply as La Valle, Site 577?) ought to belong to Roccaguglielmo. 83 Apart from Rio Vitellato none of these smaller settlements can be traced any earlier. Cultivation was going on around the; Monte Vetro sorgente as early as 1034, but the terms of the charter which describes this suggest that as yet there was no settlement here. 84 Another village in the same area, Tiano (Site 573), south of the Liri/Melfa junction, appears in a charter of 1276. 85 Similar small undefended villages can be found in the north of the valley in the shadow of M. Cairo. Villa S. Lucia, also called Villa Piedemonte, developed before 1140, 86 and despite its sack by the imperial army in 1229, unlike most of the smaller settlements survived the vicissitudes of the later Middle Ages. This was probably due in large part to the strategic position of the settlement,

at 320 metres on the lower slopes of M. Cairo. In 1483 its inhabitants were given permission by the administrator for the commendatory abbot of Montecassino to fortify the site for their own protection. 87 The monastery of S. Gregorio, north of Aquino, rebuilt in the late-tenth century, acted as the focus for a village which had developed around it by the second half of the thirteenth century, or possibly earlier, in a manner analogous to S. Paolo and S. Pietro di Foresta. 88 At the same period there was a Villa Sancti Maurici which had developed around the church of that name near the River Melfa, itself in existence by 1040. 89 To the north-east of this area, on the saddle between Monte Cassino and M. Cairo, lay an important group of monastic sites. The earliest foundation, on the saddle about 1 km. NW of the abbey of Montecassino, was S. Maria dell' Albaneta, which developed from an oratory built by a monk called Liutius c. 1011. 90 S. Nicola di Cicogna (Site 190) was constructed on a terrace below M. Castellone at a height of 450 m. (about 1 km. to the west of S. Maria) before 1057.91 S. Matteo Servorum Dei was built higher up M. Castellone, at about 750 m., behind S. Nicola, in the same period, before the death of Abbot Richer of Montecassino in 1055. In the thirteenth century a road connected the two monasteries. 92 A fourth monastic house in this vicinity, S. Angelo di Fortunula (Site 193), was founded between 1155 and 1183 on a terrace on the 400 m. line on the SE slope of a ridge below M. Castellone now known as the Colle Sant' Angelo. 93 All four of these monasteries were subject to Montecassino. Their foundation to cater for an expanding monastic demand reflects the general demographic growth of the period. The needs of monastic isolation, and perhaps also a desire for a safe position (three of the four houses were founded before 1060), were factors which dictated their relatively inaccessible sites in the mountain rather than the plain. But another factor in three of the four cases, S. Maria, S. Nicola and S. Angelo, was the use of sites formerly occupied in the Roman period and on which the remains of Roman occupation still survived. At both Albaneta and S. Angelo di Fortunula Roman cisterns were re-used as part of the new churches. 94 At the latter site wall paintings in the old cistern still attest this. Furthermore some of the Roman terracing survived and was re-utilised. The presence of these remains helped cultivation on a far from promising site, and the creation of these new monasteries encouraged the spread of cultivation from the plain into the upland area in the twelfth century. The charters of S. Maria dell'Albaneta from the early thirteenth century reveal numerous assarts (cesae) on the slopes of M. Cairo and M. Castellone in the vicinity and to the north of S. Lucia, and a Fortunula charter of 1255 reveals houses in the lands of, but outside, that village?5 Archaeologically this extension of cultivation is reflected in two medieval sites (nos. 176 and 178) at the head of the valley above S. Lucia, well over the 300 m. line, and also in the group of sites around 800 metres at Pozzo dei Monaci (nos. 148, 149 and 151), though it is just possible that the rather gritty coarsewares found at site

60

149 are Dark Age products, perhaps derived from a retreat into the hills at the time of the Lombard invasions. However, since this upland area was not part of the main and systematic survey, and work there was very much hampered by the heavy brush cover, such chance discoveries present only a very impressionistic picture. Two further foundations, both catering for women, complete the portrayal of the monastic settlement of the valley. The nunnery of S. Maria Palazzolo was founded at Villa Eucheria, 1 km. east of Castrocielo, before 1134, again on a Roman site and utilising the Roman remains. The only thirteenth-century foundation was the nunnery of S. Petronilla at Piumarola, set up in 1278 to replace the long-defunct house there which had been dedicated to the Virgin. 96 But to produce a realistic analysis of the medieval settlement of the valley as a whole one must also take into account a further eighty minor sites in the plain (Figs. 30, 37, 46) which cannot be specifically identified from the documentary sources. These show a substratum of extended secondary settlement outside the castella of the Terra Sancti Benedicti, and, to a lesser extent, in the County of Aquino. Such sites were thos~ abandoned by Cassinese tenants in times of tension for the safety of the castella, as in the 1040's and in 1106, and were akin to the vici alluded to in the papal bull to the see of Aquino of 1157. Only very occasionally do such rural sites surface before the historian's gaze, and then usually in a thoroughly unsatisfactory manner. One example comes at Pareto (Sites 179-185), on a broad and gentle slope east of Piedemonte and south of S. Lucia, an important Roman villa site occupied as late as the fifth century, where quite extensive medieval pottery remains have been discovered (which are as usual undatable). But the only documentary reference comes in a charter of 1042 which reveals the presence of an enclosed vineyard. 97 The implication here would seem to be that the medieval settlement was another product of the twelfth and thirteenth century demographic upsurge common both to the valley and to Europe as a whole. But what form the settlement took, whether farm, hamlet or church, cannot be discovered, although the extent of the scatter suggests more than one habitation. The presence of such extended settlement, except for certain specialised buildings, is curiously difficult to discover from the written evidence. But· it undoubtedly existed . In 1271 for example Abbot Bernard of Montecassino issued a general confirmation of leases "to all our faithful men of our castrum of Trocchio, living inside or outside [the castrum]' .98 Occasional and fleeting references in boundary clauses of charters mention isolated houses in the countryside, but their overall number is very few. 99 Where there is abundant evidence in the thirteenth and fourteenth century sources however is for olive presses (montana) , apparently quite widely distributed, and fulling mills and water mills even on quite minor stretches of running water as well as on the Liri itself. Such presses and mills could only be constructed on the Terra Sancti

Benedicti with abbatial license, and this jealously-guarded seigneurial prerogative was of considerable fiscal value to the abbey. Hence the high profile which such buildings occupy in the late thirteenth and fourteenth century abbatial registers. 100 It is conceivable that many of the unidentifiable medieval sites discovered by the survey are of this type. The possibility certainly exists that any sites very close to stretches of water were connected with mills, for example Sites 222/223 on the Rio Pioppeto NW of Piumarola, or Site 507 next to the Liri west of S. Apollinare. Site 432 north of S. Giorgio looks to be a bridging point. The remains of a track south of the river stop just short of the bank at a point where the latter seems too steep for a ferry crossing. Some of these sites too may have been rural churches and chapels. For example the papal taxation list of 1325, the most detailed of the period, records eleven churches and the nunnery of S. Maria Palazzolo in the territory of Castrocielo; and not only the archipresbyteral church of S. Maria at Cervaro but also five others within the confines of the castral territory. 101 Not all were within the castellum itself. That of S. Maria Magdalena, for example, was to its SW on the way to Trocchio. Another rural church in the territory of Cervaro, S. Maria Pipemis, was founded 2 km. to the NE of the castellum as late as 1399. 102 However the sections of the fourteenth-century papal taxation lists for the Cassinese lands are sometimes rather cursory, simply recording a single payment by the archpriest and clerics of a particular castello. Fortunately Cassinese sources frequently provide further information. Thus an abbatial fiscal census of 1371 records six churches in the territory of S. Vittore, five in that of Trocchio and no less than eleven in that of S. Angelo in Theodice, all localities for which the papal list of 1325 is unhelpfut. 103 Another document a few years earlier recording the income of the Cassinese sacristy refers to thirteen churches in the territory of S. Vittore and seventeen in that of S. Angelo in Theodice. 104 But how far the extension of baptismal and other sacramental rights beyond the castral pieve contributed to the growth in the number of these subordinate churches is very difficult to answer. One must furthermore ask how far there could be a growth of rural churches without a growth of population outside the walls of the castella. It seems highly unlikely that all the minor sites revealed by the survey were of the above specialist type. Occasional references suggest that open settlement was not confined to the Pontecorvo/ Aquino region, although given the relative lack of castella it was clearly more prominent there. At least two open settlements are recorded in the territory of S. Elia in the later thirteenth century, one, Fargnola, described as a villa, the other, Virdaro, as a casale. 105 One such settlement would seem to be revealed by a clu ter of sites discovered by the survey south of the Liri in a loop of the river about 4 km. west of S. Giorgio (Sites 480, 482, 483). These sites were strung along a road leading from a ferry point across the river which is marked by a further deposit of sherds (Site 484) down to S. Pietro in Curulis

61

and Roccaguglielmo. (North of the river the road probably split, one fork going to Piedemonte, the other to Pignetaro). The topography of Site 480 suggests two platforms for cottages, with Sites 482 and 483 marking a further cottage apiece. At both Sites 480 and 482 glazed as well as coarse pottery was discovered. Another rather scattered hamlet along a road may be represented by Sites 273, 277 and 280, 2 km. north of Pignetaro. These cluster around the junction of the routes from S. Angelo in Theodice to Piumarola, and from S. Germano to Interamna. Only coarse wares were discovered there, although Site 280 was also inhabited in the Roman period. Generally the sites revealed by the survey tend to be near roads or along rivers, the latter especially near the Liri from S. Stephano eastwards, the lower part of the Forme di Aquino or on the ridges near the Gari between S. Angelo in Theodice and Bantra. A group of sites lay on the Campo di Fiori north of Bantra. Other examples of sites along roads are no. 221 on the Piedemonte/Pignetaro road, and nos. 527 and 528 on the probably Roman road running east of S. Angelo in Theodice on the other side of the Gari. Site 528 revealed Roman pottery, both republican and imperial, as well as medieval. The major exception to this picture was the scatter of sites around Piedemonte and on the flat plain running south of it. But one must emphasise that only part of the valley was surveyed, although it is clear from what was surveyed in the west around the River Melfa that medieval settlement was much less thick here than on the lands of St. Benedict. The proliferation of secondary centres of population, assarting, the extension of land under cultivation into the hill areas of the valley boundaries, and . the increasing rarity of references to woodland, 106 all reflect the growth of population in the twelfth and thirteenth centuries. Measurement is more difficult. In a fragmentary and rather ambiguous Angevin taxation document of 1269-70 S. Germano was assessed at 576 hearths, Aquino at 62, S. Elia at 52 and Roccaguglielmo at 9. But the figure for S. Germano - more than that given for Naples in the same document - looks suspiciously high and may well represent the assessment for much of the Terra Sancti Benedicti. And whether these figures represent more than the crudest judgement of wealth and or population is open to doubt. Those given for Naples (448 hearths) and for other localities look suspiciously low. Sora, the seat of a bishopric after all, was assessed at only 12 hearths and Atina at 9. 107 In 1320 S. Germana's tax assessment was only 1/7 of that of Naples. 108 But the development of S. Germano as a genuine and flourishing urban centre cannot be doubted. Contemporary charters reveal palatii . ... , twostorey houses, bath-houses, shops, hospitals, a Jewry, and a borgo beyond the walls. 109 A Franciscan friary was established outside the walls as early as the 1220's. 110 S. Elia too possessed at least one palatium in 1275, and it too had spread beyond the walls. 111 Borghi w~re also attested in the later Middle Ages at S. Vittore, Castronuovo and Cervaro. 112 Probably the best estimates for the size of particular places come from the register of Abbot Thomas

from the late 1280's, but even these are difficult to interpret. Pignetaro was estimated at 150 households (was this just the castellum or did it include its territory as well?), S. Ambrogio at 70, but Trocchio at only 10 (this was surely just the castellum). 113 For non-castral settlements information is even sparser, although it is recorded that at S. Pietro in Curulis in 1271 the oath of fealty to the abbot was taken by 25 men, including 2 clerics, presumably the total adult male population of the

casale. 114 The inquests into the obligations of the inhabitants of the Terra Sancti Benedicti undertaken on the orders of Abbot Bernard in 1269-70 and 1273, and by Abbot Andrea a hundred years later, provide quite full information as to the crops cultivated in the valley. Wheat, barley, buckwheat, bearis, wine grapes and olives were grown everywhere. Usually 1/7 or 1/10 of the crop was owed to the monastery, although in a few caste/la the proportion of the wine crop owed was as high as l /3. Since renders of l /7 of grain and 1/3 of wine had been the norm in the later tenth century, these figures show that the conditions of tenure had tended to ease since that time. Oats were only mentioned at S. Giorgio in 1270. There are occasional references to millet and spelt, the latter far more common in 1371 than earlier. Cultivation of hemp was also fairly general, and a special tax, the aquaticum, was levied on this, usually of 1/30 . Abbot Bernard's regulations for the income of the Cassinese hospital of 1275 also mention flax, which one might have suspected to be generally cultivated, were not often mentioned, although at S. Vittore and S. Pietro in Fine they were part of the salutes, special customary renders at Christmas and other festivals. 115 But fig cultivation cannot have been confined just to the east of the valley. A charter of 1227 mentioned figs in the territory of Villa S. Lucia. 116 Nonetheless this may suggest that the characteristic coltura promiscua of southern Italy was rather less marked than elsewhere.

The Age of Contraction 1300-1600 If in the thirteenth century some of the casrella were expanding outside their wall s and the population generally was growing, in the fourteenth there began an era of contraction, with a real disproportion between some inhabited centres and others exacerbated by the four horsemen which beset that century of crisis: plague, earthquake, economic recession and endemic warfare. When the general economic and demographic reverse began is of course one of the most deba ted questions of medieval historiography. But with regard to southern Italy the decade of the 1340's does seem to mark a real turning point. Insofar as estimates for the population can be made from tax returns, and this is by no means easy, it looks as though the population of the Kingdom of Naples remained more or less stable between the 1290's and the l 340's. 117 But then the onset of plague coincided with the era of domestic warfare which began with the Hungarian invasion of 1347. The Liri Valley admittedly escaped the worst effects of the invasion, since King Louis entered the 62

kingdom via the central Via Flaminia from Aquila to Benevento in December/January 1347 /8. 118 But it is unlikely to have avoided the Black Death so easily. The direct impact of the plague cannot be established - there is no contemporary local chronicler or abbatial register. But the Apulian historian Dominic of Gravina recorded that: "So great was the mortality in this land that scarcely a handful of the population survived here, and this occurred universally in every city and village of the regno and the 9 world" .11 Exaggerated this testimony may be, but it was written soon after the event, and the effect of the plague can hardly have been small. Barely a year later, on 9 th September 1349, an earthquake wreaked havoc in the area. The abbey of Montecassino was reduced to ruins and not rebuilt for over twenty years. All the caste/la in the valley except S. Vittore were badly damaged. The villa of S. Gregorio was annihilated and subsequently abandoned. 120 At the same time, though it is not clear that this was directly due to earthquake damage, the castellum of Mortola was virtually deserted. 121 In 1353 the abbatial revenues from Mortola were combined with those of Bantra and leased out for three y-ears. In 1371 an inquisition found Mortola uninhabited and derived information about the abbey's rights and property from an inhabitant of neighbouring Cocuruzzu, 122 .although other documents suggest that an archpriest remained at the castellum until the end of the century . 123 However ecclesiastics and the ecclesiastical administration may well have taken time to adapt to shifts of population. The decline in the economic prosperity of the area was eloquently shown in 1373 when the government of Queen Joanna reduced the annual tax liability of the Terra Sancti Benedicti from 649 ounces and 28 tari to 312 ounces "since we have heard that these lands and castra, through the ruin of a terrible earthquake, from the frequency of wars and the continued incursions and depredations of the companies, have been brought to a low ebb and almost to annihilation". The new assessments for individual localities varied from Pontecorvo (45 ounces), S. Germano (35), S. Vittore (20), Cervaro (15 1/2), and S. Elia (15), down to Trocchio (3 ounces), Interamna (2), and lunctura and Mortola (1 apiece). 124 It is notable that those caste/la with larger assessments were those which other sources portray as substantial - S. Vittore, Cervaro and S. Elia all had borghi at this period - while those at the bottom of the scale were all later abandoned, indeed Mortola was already virtually deserted, and it seems unlikely · given the tax figure that Iunctura was in much better case. By 1388 Piumarola, the tax assessment of which had been reduced to four ounces only in 1373, was depopulated. The reason cited for this was the disruptive effect of warfare on cultivation. 125 But the signs had been clear for some time before. In 1375 two knights from Piumarola were actually resident in S. Germano, and . when the men of the castellum made formal submission to the Abbot-Bishop of Cassino after a rebellion in 1377 only nineteen were named. 126 By 1401 not only Mortola and Piumarola, but also S. Stephano, Iunctura and Interamna were "destroyed

and scarcely inhabited, on account of their poverty, terrible wars and mortalities and other fiscal burdens pressing daily upon them" . 127 Neither were the surviving castella immune. In 1397 a document from S. Angelo in Theodice lamented "this pernicious time of wars", and similar complaints came from Cervaro in 1400. 128 The valley was the scene of the climactic struggle between Louis of Anjou and King Ladislaus in 1411, and serious damage was done to its agriculture by the plundering of the latter's army as it mustered for the Roman expedition of 1413. The area was once again the main centre of operations in 1421 when papal troops fought those of the condottiere Bracchio Fortebracchio. 129 Attempts were made to preserve and revive some of the castella in trouble. In 1381 Charles of Durazzo exempted the inhabitants of S. Pietro in Monasterio from taxation for twelve years, a concession extended by King Ladislaus for a further twenty in 1388. The inhabitants of Trocchio were given a ten-year immunity from tithe, terraticum and cathedraticum dues in 1423. 130 It is significant that both a papal bull of Boniface IX of 1391 and a charter of Ladislaus of 1401 which gave lists of the abbey's castella omitted these two settlements, although neither had as yet been abandoned. 131 But whereas the later fourteenth century saw considerable changes in the settlement of the valley (Fig. 50), and the abandonment of several caste/la; despite the political upheavals and military operations of the fifteenth century the pace of change apparently slowed. It may be that the real desertions of this period were those of the hamlets and villas outside the main settlements. Those sites discovered by the survey which have revealed only medieval pottery, with few or none of the glazed early modem types, had probably disappeared by 1500, but though one assumes that most of these desertions came after 1350 one cannot be any more speci fie than that. Some of the smaller castella continued to decline in the fifteenth century, but even those recorded as virtually deserted at the beginning of the century were a long time dying. By 1454 S. Stephano was only classed as a casale, but it still possessed its own territorial jurisdiction at this period. 132 It was probably only finally deserted c. 1500. In 1529 the "rural church" of S. Stephano was under the control of the archpriest of Roccaguglielmo, and in 1554 there was an unambiguous reference to the rural church in the destroyed castellum of S. Stephano. 133 A similar chronology occurred at S. Pietro in Monasterio, where the church was in bad condition in 15 I I, and the archpresbytery was suppressed in 1542. 134 What the fifteenth century did see was the virtual extinction of monasticism m the valley outside Montecassino itself (Fig. 50). The nunnery of S. Petronilla at Piumarola probably vanished with the castellum in the late fourteenth century. 135 By that period the revenues of S. Angelo di Fortunula and S. Nicola di Cicogna, both now referred to as churches rather than 63

monasteries, had been combined together and assigned to laymen. S. Angelo's books had been moved to Montecassino by the mid-fifteenth century. 136 There was still a resident abbot of S. Matteo Servorum Dei in 1420, but by 1441 the monastic observance had collapsed and the buildings were in ruins. 137 S. Maria dell' Albaneta was going strong in 1398, and perhaps as late as 1454, but by 1481 only eight monks were left and monastic discipline and observance had collapsed. However in that year it was given a new lease of life for some years as the school for Cassinese oblates. 138 In the early modem period however it became a massaria. S. Vito sulla Melfa was still an active monastery in 1450, but by 1527 when it and its property were taken over by Montecassino monastic life had ceased. 139 The contraction of the settlement pattern continued into the sixteenth century. Bantra was abandoned c. 1550. Pignetaro was reduced to a casale, and placed administratively under the jurisdiction of S. Germano. Trocchio was finally abandoned after 1600, when the remaining inhabitants moved to Cervaro. 140 And as late as 1526 the monks of Montecassino were proposing to resettle Interamna, Iunctura and a casaletto nearby with Greek, Albanian and Slav refugees. 141 This continuation of late medieval change into the sixteenth century applied not just to the castella but to smaller settlements as well. The villa of La Valle south of the River Quesa (Site 577) was still in existence up to the 1530's. 142 To judge by the high proportion of post-medieval glazed pottery discovered there, so too was the nearby villa of Montevetro (Site 578), about 1 km. to the south. 143 (By contrast the pottery scheme would suggest that Tiano (Site 573), just south of the Liri/Quesa junction, had been abandoned earlier.) In a few cases some sort of habitation remained on the site of a former village. Thus, just as at S. Maria dell' Albaneta, the castellum of Piumarola became a Cassinese massaria. 144 But, whereas in the fourteenth and much of the fifteenth century the contraction of settlement was combined with, and largely caused by, a decline in the overall population level, one cannot be so certain that this was continued into the sixteenth century. According to Beloch's figures the number of hearths in the whole Terra di Lavoro (exclud ing Naples) rose from 35,500 in 1505 to over 67,000 in 1595. 145 Unless the Liri Valley provided an exception to this general demographic growth we may be dealing with a grouping of settlement into fewer and larger units rather than an overall decline. But a more recent study than Beloch's has suggested that the northern part of the Terra di Lavoro, including the Liri Valley, did indeed see a fall in population, at least up to 1550, and it was only in the southern area, Campania, that the populace was on the increase. 146 This would seem to be borne out because individual caste/la which survived into the modern period did not necessarily increase in size. S. Giorgio, which had 40 hearths in 1532, had only 24 in 1648. S. Vittore grew very slightly in the same period, from 165 to 175 hearths, but then fell to 106 after the great plague of 1656. 147

In conclusion therefore, one can raise more problems about the settlement of the Liri Valley in the Middle Ages than one can solve. But the Liri Valley survey does suggest at least three lessons for the settlement historian and historical demographer. There existed in the Middle Ages a whole substratum of settlement which largely escaped documentation, discovery of which modifies one's view of the castello, though it certainly does not destroy the significance of the latter. Secondly it reinforces the importance of regional variations, even between areas geographically quite close. The process of incastellamento in the valley differed markedly from that which occurred in eastern and southern Lazio, both in chronology and cause. Thirdly a survey ranging from the eighth through to the seventeenth century shows how important it is for the historian to be conscious that, behind the exigencies of the moment, lie the ineluctable processes of time. For the settlement historian there can only be la longue duree.

8.

9. 10.

Notes The following abbr eviations are used throughout. Chron Cas : WP Chronica Monasterii Casinensis, ed. H. Hoffmann, MGH SS xxxiv (1980) (citations are from book and chapter and by the page number of Hoffmann 's edition).

11.

FSI: Fonti per la Storia d'Italia (Rome 1887 to date).

12.

Gattula, Accessiones: E. Gattula, Accessiones Historiam Abbatiae Casinensis (Venice 1734) Gattula, Historia: E. Gattula, Casinensis (Venice 1733)

Historia

ad

13.

Abbatiae

Leccisotti: Abbazia di Montecassino, Regesto dell'Archivio, ed. T. Leccisotti & F. Avagliano (11 vols., Rome 1964- 77).

MGH: Monumema Germaniae Historica (following the usual conventions, SS = Scriptores, SS Rer Lang. = Scriptores Rerum Langobardorum (Hannover 1878) etc.

1. 2. 3.

4.

5. 6.

1.

Chron Cas 1.4, pp. 22-3. Paul, Historia Langobardorum 11.32, ed. L. Bethmann & G. Waitz, MGH SS Rer Lang p. 91. Paul, Historia Langobardorum IV .17, p. 122. Gregory I, Dialogi Libri Quattuor, ed. U. Moricca (FSI, 1924) p. 153. Gregory I, Epistolae lib. IIl.13, ed. P. Ewald & L.M. Hartmann, MGH Epistolarum I (Berlin 1887) pp. 171-3. Paul, Historia Langobardorum VI.27, p. 174. Vita Willibaldi, MGH SS xv.102 . Chro11ica S. Be11edicti Casi11e11sis,ed. G. Waitz, MGH SS Per Lang, c. 21, p. 480. Chro11 Cas I.5, pp. 25-8, taken from Gattula, Historia p. 105 (928). The "Gisulf" diploma is edited ibid. p. 27. For a discussion of the complex relationship between the vanous 64

14. 15. 16. 17. 18.

19.

20. 21. 22.

recensions of the chronicle and this forgery see H. Hoffmann, "Chronik und Urkunden in Montecassino", Quellen und Forschungen aus Italie11ische11Archiven und Bibliotheken 51 (1971) 193-4. For the bounds see L. Fabiani, La Terra di S. Benedetto (Miscellanea Cassinese xxxiiixxxiv, Montecassino 1968) i.17-19, 41-3. On whom F. Scandone, "II Gastaldato di Aquino della meta del secolo IX alla fine del X", Archivio storico per le provincie napoletane 33 (1908) 720735, and 34 (1909) 49-77, especially 61-3 on the date of the comital title. Chronicon Vulternense, ed. V. Federici (FSI, 3 vols., 1924-38) i.231. G. Falco, "Lineamenti di storia cassinese nei secoli VIII e IX", Casinensia. Miscellanea di studi cassi11esi pubblicati in occasione del XIV centenario della fondazione della badia di Montecassino (Montecassino 1929) pp. 486-8 . For Charlemagne's diploma, E. Caspar, "Echte and gefiilschte Karolingersurkunden fiir Monte Cassino", Neues Archiv 33 (1907) 53-73, especially 72-3. Fabiani, Terra i.146-150. For the nunnery, Chron Cas 1.8, p. 54, for the cells of Valleluce, S. Apollinare and S. Gregorio, ibid. o 259 (1271) . E.g. the exchange of servi between Montecassino and St. Vincent on Voltumo recorded by Chron Cas 1.35, p. 95. c. J. Wickham, 'Historical and topographical notes on early medieval south Etruria" , Papers of the British School at Rome 46 (1978) 173- 7. Fabiani, Terra ii.290-4. Cf. P. Toubert, Les Structures du Latium medieval . Le Latium meridionale et la Sabine du !Xe siecle a la fin du XII siecle (2 vols., Rome 1973) i.465-472 , and also C. J. Wickham, "The terra of San Vicenzo al Voltumo in the 8th to the 12th centuries: the historical framework" , in R. Hodges (ed.), San Vincenzo al Volturno. The Arch[a)eology, Art and Territory of an Early Medieval Monastery (Oxford 1985) pp. 223- 4. Chronica S. Benedicti Casinensis cc. 6-7, pp. 472-3. Chron Cas I.7-8, pp. 77 -82. Chro11 Cas 1.35, pp. 97-8. Chronica S. Benedicti Casinensis c. 14, p. 475. Chron Cas 1.33, p. 90. Chro11 Cas I.44, p. 114. Erchempert, Historia Langobardorum Beneventanorum, ed. G. Waitz, MGH SS Rer Lang c. 44, p. 251. Chron Cas I. 52, pp. 133-5. J. Gay, L 'ltalie meridionale et l 'empire byzantin (2 vols., Paris 1904) i. 155-163. Chron Cas 11.1, pp. 166-7. Montecassino, Archivio dell' Abbazia, Aula II Caps. xxxvi no. 4. Chron Cas 11.1, p. 167. The sense is slightly ambiguous. But according to Chron Cas 1.45, p. 121, Abbot Angelarius began the repair of the

23. 24. 25.

26. 27.

28. 29. 30. 31.

32.

33.

34.

35. 36. 37. 38. 39. 40. 41. 42. 43.

church of the Saviour in the late 880's. F. Trinchera, Syllabus Graecarum Membranarum (Naples 1865) pp. 2-3 no. 3. Gattula, Accessiones pp. 58-9. H. Hoffmann, "Die Kirchenstaat im hohen Mittelalter", Quellen und Forschungen aus ltalienischen Archiven und Bibliotheken 51 (1977) 44-5. Chron Cas II. l, p. 166. See especially I Placiti Cassinesi del secolo X con periodi in volgare, ed. M. Inguanez (4th edition, Miscellanea Cassinese xxiv, 1942) pp. 13-15 no. 1 (960), Chron Cas Il.1-2, 6, pp. 167-171, 176. Fabiani, Terra i.47-56. N. Cilento, Italia meridionale longobarda (2nd edition, Naples 1971) pp. 199-201. Chron Cas 11.3, pp. 171-2. L. Tosti, Storia della badia di Montecassino (3 vols. Naples 1842) i. 223-6. Tosti, Storia i.226-8. See P. Taubert, "Les destinees d'un theme historiographique: "castelli" et peuplement dans l'Italie medievale", in his collected essays Histoire du Haut Moyen Age et de l 'Italie medievale (London 1987) essay no. IV. G. Noye, "Feodalite et habitat fortifie en Calabre dans la deuxieme moitie du Xle siecle et le premier tiers du Xlle siecle", Structures feodales dans I 'Occident mediterraneen (Xe-XI/le siecles) (Rome 1980) pp. 620-1. Here the various works of C. J. Wickham are of crucial importance, notably The Mountains and the City. The Tuscan Apennines in the Early Middle Ages (Oxford 1988), especially pp. 36-9, 289-306; "The Terra of San Vincenzo" [above n. 13], especially pp. 235-243; ll Problema dell 'incastellamento nell 'Italia centrale. L 'esempio di San Vincenzo al Volturno (Florence 1985), especially pp. 24-52. Taubert, Latium i.303-368, especially 321-6. Wickham, ll problema dell 'incastellamellto pp. 53-74. P. Taubert, "Pour une histoire de l'environnement economique et social du Mont Cassin (IXe-XIIe siecles)", Comptes Rendus de l 'Academie des Inscriptions et Belles-Lettres (1976) 701, reprinted in his Histoire du Haut Moyen Age [above, n. 31] essay no. IX. Toubert, Latium i.322-3. Wickham, "The Terra of San Vincenzo" p. 235. Patrologia Latina, ed. J.-P. Migne (221 vols., Paris 1844-64) cxliii.1305-8. Gattula, Accessiones p. 90. ChronCasII.13,p.192. Gattula, Accessiones p. 91. Chron Cas II.25, p. 210. Chron Cas II.20, p. 204. Chron Cas II.38, 67-8, 74, Ill.16, pp. 241, 302, 304, 316, 380. Chron Cas II.72, p. 312. Toubert, "Les Destinees d'un theme 65

44. 45.

46.

47.

48.

49. 50.

51.

52.

53. 54. 55. 56.

historiographique" pp. 14-15. Leccisotti vii.214 nos. 29-30 (1042), Montecassino, Aula II Caps cxxviii (fasc. 2) no. 15. Chron Cas II.2, pp. 168-9. F. Scandone, "Roccasecca. Patria di S. Tommaso di Aquino", Archivio storico di Terra di Lavoro 1 (1956) 33-45. For Castrocielo, Gattula, Accessiones pp. 89-90 (994) and Chron Cas II.14, p. 194. For Interamna, Gattula, Accessiones pp. 90-1 (1000-1007), Chron Cas II.15, p. 195. But the mention of lnteramna comes only in the second and third recensions of the Chronicle, not in the original version of Leo Marsicanus, and the charter is known only from the dubious copy of Montecassino, Archivio dell' Abbazia, Registrum Petri Diaconi (Reg. no. 3) fol. 255r no. 615, and Interamna was not, as claimed there, in the possession of Montecassino at this time, nor was it to be until 1066. Interamna was certainly a castellum in 1051, Gattula, Historia pp. 269-270. H. Bloch, Montecassino in the Middle Ages (3 vols., Rome 1986) i.174-5. Gattula, Accessiones p. 89. Scandone, "Roccasecca" 121-2 no. 15. The bounds of the territory of Pontecorvo were recorded by MGH Diplomatum Ottonis Ill (Hannover 1893) p. 761 no. 333 (999). For the foundation, Gattula, Historia p. 293. A. Nicosia, "La Valle della Quesa e ii monastero greco di S. Pietro", Benedictina 24 (1977) 115122 shows that this document refers to S. Pietro, and not to S. Paolo. For its acquisition by Montecassino, Gattula, Historia p. 267. See also Bloch, Momecassino in the Middle Ages i.213-5. It was given to Montecassino in 1093, Gattula, Historia p. 294. On its later history Bloch, Momecassino in the Middle Ages i.215-6. Leccisotti viii.133-4, 136 nos. 1-2, 7. F. Avagliano, "Le Piu antiche carte di S. Nicola di Pico conservate nell'archivio di Montecassino", Scritti in onore di Filippo Caraffa (Anagni 1986) pp. 213- 7 nos. 1-2. No. 1 = Gattula, Historia pp. 213-4. Cf. Bloch, Montecassino in the Middle Ages i.211-3. Montecassino, Aula II Caps. cvii no. 7. R. Bonani, Ricerche per la storia di Aquino (Alatri 1922) pp. 190-1 suggests that the monastery was founded c. 1000, but without citing any evidence. Chron Cas II.73, p. 315. For the background, ibid. Il.57, 68, 71, pp. 277, 304-6, 309-312. Fabiani, Terra i. 73-4, and (better) R. Palmarocchi, L 'Abbazia di Montecassino e la conquista normamia (Rome 1913) pp. 79-86. Chron Cas II.43, p. 251. Bloch, Montecassino in the Middle Ages i.192-3. Chron Cas II.71, pp. 309-310. Chron Cas II.68, 70, 72, 73, pp. 304, 308, 312, 315. Chron Cas II.72, p. 312. Gattula, Historia p. 75.

51. 58.

59.

60.

61.

62. 63.

64. 65.

66. 67.

68. 69.

70.

71. 72.

73.

74.

75.

Chron Cas II.57, p. 277. Catalogus Baronum, ed. E. M. Jamison (FSI 1972) p. 150 art. 823. In 1228 the abbey provided 100 "well-equipped" sergeants for Frederick H's Crusade, Richard of S. Germano, Chronicon, ed. C.A. Garufi (Rerum Italicarum Scriptores, 2nd ed., Bologna 1938) p. 149. Gattula, Accessiones pp. 165-7, Historia pp. 312313. G. A. Loud, "A Calendar of the Diplomas of the Norman Princes of Capua", PBSR 49 (1981) 120-2 nos. 5, 6, 10, 12-13. Chron Cas III.16, p. 380. Scandone, "Roccasecca" 129 no. 51. Bloch, Montecassino in the Middle Ages i.174. Regesto di Tommaso Decano o Cartolario del Convento Cassinese 1178-1280, ed. M. Inguanez (Montecassino 1915) pp. 239-242 no. 112 (1278). Patrologia Latina cxliii.831-3 . A. Pantoni, "Le Chiese di S. Andrea sul Garigliano", "S. Apollinare e le sue chiese", BollettinoDiocesano Montecassino 19 (1964) 104, 142; "S. Ambrogio sul Garigliano", ibid . 20 (1965) 65. Chron Cas II.71 , p. 309. Chron Cas III.11, p. 373. Montecassino, Au/a II Caps. Ii nos. 13-14. (Leccisotti viii.182-3). Richard of S. Germano, Chronicon pp. 29, 32, recorded the fortress and villa of Rocca d'Evandro. Chron Cas IV.29, p. 495. See G.A. Loud, Church and Society in the Norman Principality of Capua 1058-1197(0xford 1985) pp. 91-3, 132-5. Amiales Casinenses ad an.1139, MGH SS xix.309. Chron Cas IV.128, p. 604. Anna/es Casinenses ad. an. 1140, MGH SS xix.309. Fabiani, Terra i.119-122, ii.19-20. P. Toubert, "La terre et les hommes dans l'Italie normande au temps de Roger II: l'exemple campanien", Societa, potere e popolo ne/l'eta di Ruggero II. Atti della Ille giomate 11orma11110 sveve, Bari 1977 (Bari 1979) especially pp. 6671, reprinted in his Histoire du Haut Moyen Age [see note 31] essay no. X. Richard of S. Germano, Chronicon pp. 14, 18, 20-1. Richard of S. Germano, Chronicon pp. 153-4, 162. Bloch, Montecassino in the Middle Ages i. 175 is mistaken in viewing this as the final destruction of Interamna. Saba Malaspina, Rerum Sicularum Historia 111.45, ed. G. Del Re, Cronisti e scrittori sincroni napoletani (2 vols., Naples 1845-68) ii. 247-9. P.F. Kehr, "Papsturkunden in Rom", Nachrichten der K. Gesellschafi der Wissenschaften zu Gottingen. Philo/ogisch-historische Klasse (1901) 260-2 no. 6 . Rationes Decimarum ltaliae nei secoli XIII e XIV. Campania, ed. M . Inguanez, L. Mattei-Cerasoli & P. Sella (Studi e Testi 97, Vatican City 1942) 66

76.

77.

pp. 37-8 nos. 428-433. Montecassino , Aula II Caps. Ii nos. 2 (1034), 4 (1042), 5 (1042), 11 (1065). Leccisotti viii.176 9. Richard of S. Germano, Chronicon pp . 113, 116-

7. 78.

79.

80.

81. 82.

83. 84.

85. 86.

87.

Montecassino, Aula II Caps. cxxvi no . 28. Another Charter, Aula II Caps. cxxvi no. 27 has the date 1050, and is listed under that date in Gattula's MS. inventory. But the use of Frederick H's regnal years shows this to be a scribal error , and the actual date should probably be 1250. It too calls S. Pietro a villa. I Registri de/la cance/laria angioina ricostruit i, ed. R . Filangieri di Candida et al . (28 vols., Naples 1950 to date) vi .251 no . 342. Regesti Bernardi I Abbatis Casinensis Fragmenta, ed. A. M. Caplet (Vatican City 1890) pp. 97-8 no . 228 . Reg. Tom. Dec . p. 7 no. 4 (1227) refers to an earlier concession by Pope Innocent III (d. 1216) of the church to an apostolic notary . The monastic community had surely by then disappeared, and thus Bloch, Montecassino in the Middle Ages i. 216 is in error in asserting that it continued into the fifteenth century. In 1278 the church was apportioned specifically to the support of the Cassinese convent, Reg. Tom. Dec. pp. 287-9 no. 133. Based on the map in Rat. Dec. Italiae Campania. Nicosia, "La Valle della Quesa" 125. Reg. Bernardi I Abbatis p. 198 no. 446. Montecassino, Archivio dell'abbazia, Registrum Angeli et Andreae (Reg. no. 11) fol. 34r. Gattula, Accessiones pp. 394-5. Montecassino, Aula II Caps Ii no 2 (Leccisotti viii. I 76-7), the donation of a piece of land next to the surgile to S. Pietro in Foresta. Reg. Tom. Dec. pp . 271 -2 no. 125. The first reference comes in December 1139, Regesto di S. Matteo di Castello o Servorum Dei, ed. M. Inguanez (Montecassino 1914) pp. 51-4 no. 25. L. Mattei -Cerasoli, "Tre Registri del Cardinale Giovanni d' Aragona, commendatario di Montecassino", Casinensia [see note 10] pp. 598-

9. 88.

89.

For the vill, Reg. Bernardi I Abbatis p. 88 no. 200 (1258). The earliest charter of the monastery, Gattula, Historia p. 75, there dated to 1052, should probably be re-assigned to 977, A. Pantoni, "Una memoria scomparsa. S. Gregorio di Aquino" , Benedi ctina 1 (1947) 255. This article describes the remains of the church which were demolished completely in 1943 to makeway for an airfield . Bloch, Monte cassino in the Middle Ages i.205-7. Reg. Bernardi I Abbatis p . 159 no. 386 (1273) records the villa. For the church, Gattula, Accessiones p . 141, and Chron Cas 11.79, pp. 324-5 . Bloch, Mo ntecassino in the Midd le Ages

90.

91.

92.

93.

94.

95.

96.

97. 98.

99.

100.

i.217. Chron Cas II.30, pp. 221-2. Cf. Petri Diaconi Ort us et Vita lustorum Cenobii Casinensis, ed. R.H. Rodgers (Berkeley 1972) c.31, p. 64. Bloch, Montecassino in the Middle Ages ii. 714-6. It was named in the papal bull confirming Cassinese property of that year, Patrologia Latina cxliii.831-3. Bloch's suggestion, Montecassino in the Middle Ages i.309, ii.716-7, that it was already in existence under Abbot Atenulf of Montecassino (d. 1022) is possible, but based as it is on one marginal annotation, must remain a little speculative. Peter the Deacon, Ortus et Vitae c.54 p. 75. Reg. S. Matteo p. 1 no. 1. It is probable that the Alfanus, Archbishop of Salerno, named here is a copyist's error for Amatus. Registrum Sancti Angeli de Fortunula, ed. M. lnguanez (Montecassino 1926) pp. vi, 31 appendix no. I. Fabiani, Terra ii.192-3. The cistern at Albaneta was specifically mentioned by Chron Cas. II.30, pp. 221-2, and that at S. Angelo in a list of the monastery's property in the 1270's, Reg. S. Angeli de Fortunula p. 8. The latter was physically examined. Montecassino, Au/a II Caps. cvii nos. 16 (1196), 8 (1211), 17 (1212), 19 (1242), 3 (1242), 23 (1247), and the inventory of the monastery's property done in 1257 by Abbot Richard of Montecassino, ibid. no. 4. The documents of this section of the Montecassino archives are not filed in chronological order. Cf. also Reg. Tom. Dec. pp. 36-8 no. 23 (1207) in the Fortunula area, and Reg. Bernardi I Abbaris p. 109 no 259 (1271). Montecassino, Aula II Caps. lxvi no. 4 (Leccisotti xi.368-9). Reg. Tom. Dec. pp. 239-242 no. 112. The intention to found a nunnery expressed in the villa of Tuscania, wife of Richard de Babuco, in a will of 1214, Reg. Tom. Dec. pp. 40-3 no. 25, seems not to have been fulfilled. Leccisotti viii.214 no. 29. Reg. Bernardi I Abbaris pp. 103-4 no. 241. Cf. ibid. p. 106 no. 253 to the men of S. Angelo in Theodice. E.g. houses in the territory of S. Pietro in Monasterio, Reg. Tom. Dec. pp. 72-4 no. 43 (1239), 108-110 no. 55 (1253); two houses in the territory of Piedemonte, ibid. pp. 168-171 no. 80 (1267), several adjoining houses in the territory of S. Germano, Montecassino Aula II Caps. lxxvi no. 29 (1281). N.B. the prohibition on the construction of olive presses or of any buildings on watercourses in the territories of S. Elia, Cervaro, Trocchio and S. Vittore without license of the abbot and monastery, Montecassino, Archivio dell' Abbazia, Registrum II Thomae Abbatis (Reg. no. 9) fols. 31v-32r. For examples of such licenses, Reg. Bernardi I Abbatis p. 113 no. 267 (1271), half 67

101. 102. 103. 104. 105. 106.

107. 108.

109.

110. 111. 112.

113.

the revenues to Montecassino, ibid. p. 124 no. 293 (1271), by livello for an annual payment of 1 tari, both of these to construct olive presses; ibid. p. 116 no. 276 (also 1271), to construct a mill on the "Rivo de Sancto" near Mortola in return for 2 lbs. of wax annually, ibid. p. 126 no. 298 (1272), a license to build a mill and a fulling mill on the River Pecia between the territories of Bantra and Rocca Bantra, an annual census of 3 lbs. of wax owed for each. Cf. Montecassino, Archivio dell' Abbazia, Registrum Censuum et Confinium Abbatiae Casinensis 1278-1410 (Reg. no. 17) fol. 159v, the abbess and nuns of S. Petronilla at Piumarola owing a census for a mill on the "aqua de Plumbariola". Rat. Dec. ltaliae Campania pp. 34 nos. 404-5, 50 nos. 531-4. A. Pantoni, "Cervaro", Bollettino Diocesano Montecassino 23 (1968) 71. Reg. Censuum et Confinium fols. 134v-l35r, 144v, 155r. Reg. Angeli et Andreae fols. 34r-'36r. Reg. Tom. Dec. pp. 242-5 no. 113, 297-9 no. 137. Never of course completely absent, for example the reference to a wood in the territory of S. Vittore in 1274, Reg. Bernardi I Abbatis p. 184 no. 431. But this document, which involves property over a wide area, has only this one mention of woodland. N.B. the census owed by the universitas of Pignetaro for its woods, Reg. Censuum et Confinium fol. 157r (1371). Reg. Ang. [see note 79] ii.218-220. K. J. Beloch, Bevolkerungsgeschichte It aliens i Grundlagen. Die Bevolkerung Siziliens und des Konigreichs Neapel (Berlin 1937) 246-7. E.g. Reg. Bernardi I Abbatis p. 182-431 (1274) refers to a palatum in the Shoemakers' square with seven shops on the ground floor, another palatum, also with seven shops, near the palatum of the Jews, a bath-house in the suburb, a house next to the new square with two solars on the ground floor etc. Ibid p. 186 no. 432 mentions another bath-house near the bridge-gate, and a hospital. There were at least two hospitals in S. Germano c. 1400, the Hospital of Piedemonte, which was next to the St. John's Gate, and the Hospital of S. Maria, Montecassino, Archivio dell' Abbazia, Registrum Henrici I Abbatis (Reg. no. 13) fols. 5v, llr. For the Jewry, see also Richard of S. Germano, Chronicon p. 176. Reg. Angeli et Andreae fol. 30r (1231). Reg. Bernardi I Abbatis pp. 205-6 nos. 479-480. Reg. Henrici I Abbatis fols. 122v (S. Vittore, 1402?), 165v, 172r, 185r (Cervaro, 1400), 355r (Castronuovo, 1406). Montecassino, Archivio dell' Abbazia, Registrum I Thomae Abbatis (Reg. no. 8) fol. 106r. A. Pantoni, "Pignetaro Interamna", Bollettino Diocesano Montecassino 20 (1965) 137. Pantoni,

"S. Ambrogio" 114. 115.

116. 117 . 118.

[see note 63] 66.

Gattula,

Accessiones p. 391. Reg. Bernardi I Abbatis pp. 97-98 no. 228. The 1269-70 inquests are Reg. Bernardi I Abbatis pp. 26-50 nos. 79-93, those of 1273 Montecassino, Archivio dell' Abbazia, Registrum II Bernardi Abbatis (Reg. no. 6), and those of 1371 are printed by Gattula, Accessiones pp. 429433, from Reg. Angeli et Andreae. The regulations for the support of the Hospital are printed Reg. Bernardi I Abbatis pp. 154-7 no. 378 and Reg. Tom. Dec. pp. 249-255 no. 117. Reg. S. Angeli de Fortunula pp. 19-20 no. 8. A. Filangieri, Territorio e populazione nell 'ltalia meridionale (Milan 1980) pp. 125-133. Matthew Palmer , Vita Nicolai Acciaolo, ed. G. Scaramella (Rerum ltalicarum Scriptores, 2nd ed. Bologna 1934) pp . 9-10. Dominic of Gravina,

Chronicon de Rebus in Apulia Gestis (1333-1350),

119. 120.

121.

122. 123.

124. 125.

126. 127. 128. 129.

130.

131. 132. 133. 134. 135.

ed. A. Sorbelli (Rerum Italicarum Scriptores , 2nd ed., Citta di Castello 1903) p. 33 . Chronicon p. 49 . Cf. Vita Nicolai Acciaolo pp. 11-12. Gattula, Accessiones p. 407. A. Caravita, Il Codice e le arte a Momecassino (2 vols., Montecassino 1869) ii.21. Fabiani, Terra ii.185. A document of King Ladislaus of 1393 records that the desertion was "at the time of the earthquake", Gattula, Accessiones pp. 475-6. Montecassino, Aula II Caps. ciii no. 8. Gattula, Accessiones p. 430. Rat. Dec. ltaliae Campania pp. 53, 56, 59 nos. 555, 600, 637 (1358/9). Here however the archpriest and clerics could pay only one third of the taxation due. Montecassino, Archivio dell' Abbazia, Registrum Petri de Tartaris (Reg. no. 12) fol. 87v (1397) also records an archpriest at Mortola . Gattula, Accessiones pp. 418-419. Gattula, Accessiones p. 466. In 1371 an abbatial inquisition derived information about Iunctura from the archpriest and men of Bantra, ibid. p. 430. Reg. Petri de Tartaris fols. 16r, 227r. Gattula, Accessiones pp. 498-9. Reg. Henrici I Abbatis fols. 26r, l 86r, 238r.

Notabilia Temporum di Angelo de Tummulis da Sant 'Elia, ed. C. Corvisieri (FSI 1890) pp. 1213, 33-4. Gattula, Accessiones p. 744, 528. A. Pantoni, "S. Pietro in Monasterio", Bollettino Diocesano di Montecassino 17 (1962) 30. Gattula, Accessiones pp. 473, 498-9. Leccisotti viii.169 nos. 5-6. Leccisotti viii.170-1 nos. 7, 11. Pantoni, "S. Pietro in Monasterio" 30-1. The last references to the nunnery which I can find are Rat. Dec. Italiae Campania p. 60 no. 648 (1360), Reg. Angeli et Andreae fol. llv (1362). 68

136. 137. 138.

139. 140.

141. 142. 143. 144. 145. 146. 147.

Reg. S. Angeli de Fortunula pp. ix-xi. Reg. S. Matteo pp. xxviii-xxix, 160 appendix no. 8. Leccisotti i.88 no. 41. Reg. Henrici I Abbatis fol. 2v. T. Leccisotti, S. Tommaso d 'Aquino e Montecassit~o (Miscellanea Cassinese xxxii, Montecassino 1965) p. 33. Gattula, Historia p. 584. Montecassino, Aula II Caps. xxxvi fasc. vi. Leccisotti i.181 no. 47. Gattula, Accessiones p. 752. Pantoni, "Pignetaro" 139. Trocchio still had its own territorial jurisdiction in 1624, Leccisotti xi.252 no. 608. Gattula, Accessiones pp. 604-5. Leccisotti viii.239 nos. 26-7. This was probably the villa "del Monte Yitai(?)" referred to by Leccisotti viii.230 no. 3 (1485). Leccisotti viii.262-3 nos. 28, 31-32. Beloch, Bevolkerungsgeschichte i.215. Filangieri, Territorio e populazione nell'Italia meridionale p. 271. A. Pantoni, "S. Giorgio a Liri", Bollettino Diocesano Montecassino 22 (1967) 133, "S. Vittore del Lazio", ibid. 30 (1975) 225.

The research for this chapter would have been impossible without the assistance of the late don Tommaso Leccisotti and don Faustino Avagliano, archivists of the abbey of Montecassino, which is here gratefully acknowledged.

Chapter 12

SYNTHESIS J.W. Hayes and J.P. Martini

In the preceding chapters the environmental background to settlement in the Valley, and the documentary evidence for Roman and more recent times, have been surveyed, and pictures of the state of the Valley in successive periods have been created. The findings of the survey work, in terms of raw material, are presented in Appendices A-E. An archaeological assessment of the new work, drawing together the various threads of our research, is now required. What new information has been gained? At the time of our work, the region in question, the southernmost limit of present-day Lazio, was underrepresented in the official archaeological record, owing to a quite natural concentration of resources and personnel in the areas closer to Rome; hence distribution maps of finds, for almost all periods, showed an uncomfortable vacuum between the closely investigated areas of central Lazio and Campania. Scattered excavations and researches had been conducted on the town sites (chiefly at Cassino), but the nature and extent of rural settlement were poorly known. Thanks to the work of Giannetti and others, the ancient inscriptions of the region were well documented, but in other aspects the evidence was little known outside local circles (where A. Nicosia and his Pontecorvo group were filling the gap left by officialdom). Survey projects had not, in the early stages, been directed to here, partly because of the paucity of attractive upstanding monuments; hence the region presented itself as "new territory" for a new generation of integrated systematic survey projects. It remained to be seen whether the settlement patterns demonstrated in regions such as South Etruria would be matched here. In the event, some similarities and some major differences emerged. As elsewhere, the artefactual evidence is heavily weighted towards the Roman and High Medieval periods, documenting phases of high population combined with vigorous "market" activities. A "Dark Age", in which the material evidence is hard to identify and probably much sparser, intervenes between the two; in this instance documentary evidence, here relatively rich owing to the presence of the Abbey of Montecassino, goes a good way to filling in the picture. In post-Medieval times a significant rural slump (in terms of sites producing artifacts) is evident between the 15th and the 18th century; this may contrast with the situation in regions further north. A similar contrast may be seen in pre-Roman times: the prehistoric sites are difficult to assign to a specific period, and the cultural upswing and proliferation of settlements seen both to the north (Etruria, Rome region) and the south (Campania) from the early Iron Age onwards is here virtually missing, suggesting less intensive agriculture.

69

In assessing the balance between rural and urban from the survey evidence, one should sound a note of caution: the pottery evidence is not uniform. The Roman peak in settlement is characterized by great quantities of cooking wares (seen typically from the capanna, Site 186), along with a (low) modicum of tablewares, which must surely indicate that the sites were in relatively permanent occupation, and presumably the principal dwellings of their inhabitants. In contrast, the commonest Medieval finds are two-handled jars, which seem likely to have functioned mainly as drinking vessels, and may have been carried around more by countrymen engaged in their daily rounds. Hence these, if found in isolation, may not be direct evidence of a habitation. This seems even more likely for the post-Medieval finds (largely Pontecorvo-type waterjars, without a complementary range of wares)--hence the possibility of minimum rural dispersed settlement at this time. Such considerations must always be borne in mind.

Case study: the Melfa River Area The study of the Melfa River area and its behaviour was conducted in relation to possible crossings of the river. The topography was assessed in terms of its effects on the various possible locations of bridges and fords. In 1978 a team of four researchers conducted a longitudinal survey along two near-parallel roads that run ESE-WNW from the site of Aquinum to the river Melfa. This aimed to establish the line of a Roman road, appearing as a straight line on the map, locally referred to as the "strada romana", 1 and that of a more meandering Medieval road just to the north of it. The survey produced very few sites (only one along the Roman road) but many artifacts were noted incorporated in present-day structures. There is no evidence of road paving, except within the town of Aquinum. Our study was consequently enlarged into an intensive area survey in the sector near the Melfa River. Sites were found throughout, but particularly in the hilly area, close to the 100m contour line. The line of the suspected Roman road was established, as it is clearly marked by tomb sites (Appendix E, Sites 39, 46) and other sites (59, 65, 77). Since the pottery and coin finds from these include Black Slip, terra sigillata, and l''-2 nd and 2nd3n1century material, it appears that the road was in use from Republican times to at least the 3n1century A.D. A series of sites (25, 27, 33, 34, 36) were found on a SWNE axis in the southwestern area near the Melfa-Liri confluence on the high terrace. An ancient road linking these sites was reported by the present inhabitants. These sites may indicate another important road that was in use from the Republican period through the third century A.D.

It most likely also serviced a villa (Site 218) on the rim of the high terrace facing the river valley. In the same localities there are Iron Age finds. It appears that this road led to a crossing of the Liri not far from its confluence with the Melfa. Such a crossing (Figs. 22-23) served to link the Iron Age sites on both sides of the river, and also Aquinum with Fabrateria Nova. The continuation of the road on the right bank of the Liri is identified by the alignment of sites. Another such alignment along the high terrace to the north of the Melfa may be associated with an earlier, more northerly route from Fregellae (destroyed in 125 B.C.) towards Aquinum and the south. In the Melfa area, the concentration and distribution of Augustan sites, to judge by finds of terra sigillata and contemporaneous wares and materials, remains similar to that in the Republican period. Expansion of sites occurs from the mid-first century A.D. in the area east of the Liri and west of the Melfa, and for the first time the line of the road leading from Aquinum directly to the Melfa is confirmed throughout, continuing on the right side of the river where some bridges , such as the Ponte del Diavolo, still stand along the route. This road clearly connected the Melfa crossing near S. Vito with the Liri bridge at Fabrateria Nova. At S. Vito there is no evidence of early bridge building (the river can be forded readily during the summer), but a few hundred metres north of the junction of the road alignment and the river there is evidence of a Late Imperial bridge (the S. Vito bridge). Probably the road was diverted slightly to reach it. From the mid-2 nd century into the 3rd century the sites become fewer though the distribution patterns and the roads, in the main, remain the same. Only five sites survive to ttie 5th century, and they are located along the northern Roman route, indicating its continued existence. Numerous Medieval sites were found close to this northern route leading to the medieval bridges built just north of S. Vito, where the present bridge of the Autostrada del Sole lies.

In this part of the valley there is a direct correlation between topography and settlement. Along the edges of the High Terrace, on the spurs, lie many sites. Each takes advantage of the drainage, summer breeze and views which such positions afford. On the flat land of the terrace itself, sites are much more sparse. It is possible that a series of bridges was built across the Melfa in Roman and Medieval times. The braided, ephemeral, torrential regime of the river, its substratum made of soft silt and clay overlaid with gravel, could not have provided a strong foundation for bridges built in the lower part of the river, particularly as they were subjected to violent floods in the winter and spring . Similar difficulties are experienced with the modem bridges.

On only one site (I 19), at the foot of M. Cairo just below the present-day Roccasecca Alta, was a villa built at a crossing of the Melfa.

Settlement patterns When the site evidence first appears in quantity, in the Late Bronze/Early Iron Age, nucleated settlement, either on defensible positions or close to good water sources, seems to be the norm . We have suggested that this situation persisted for some considerable time after the first Roman settlement of the region, in all probability , as before, for reasons of security . During the 2nd and 111 centuries B.C. the picture changed drastically, as an evidently large population moved out into the countryside in all directions, to cultivate , it would seem, every available patch of land. The presence of significant sites with artifacts of this period in upland areas hints at population pressure. The larger rural sites, whether to be classified as "farms" or "villas", generally begin in this period. These tend to lie on gentle south-facing slopes (mirroring the advice given by the ancient agricultural writers). By the standards of south and coastal Etruria these are not large, and derive their supplies from local sources, having little connection with long-distance trade networks; hence they seem to conform more to the precepts of Cato the Elder . This dispersed pattern of settlement, with relatively weak external connections, continues until the 2nd century A.D. A major change then occurs, affecting both the rural sites and the towns (which in effect serviced them): both, if we may judge by the evidence of the fine pottery, go into a sharp decline. Here the change seems to occur around the late 2nd century-significantly earlier than in South Etruria 2 (though one might just draw a parallel with the Settefinestre evidence). Most of the smaller rural sites fail to produce clear evidence of occupation after A.D. 200 (whether this implies abandonment or merely impoverishment). The evidence for later activity comes mainly from the larger rural sites, suggesting concentration of land-owning in relatively few hands (how wealthy these late proprietors were cannot be proved short of excavation of the sites in question, since the inscriptional evidence now becomes meagre). The relatively early date at which the major change occurred is of note. However, no change in the balance· between town and country is apparent until Medieval times (with the phenomenon of incastellamento). Even the latter may be questioned in terms of its long-term effects: the survey work shows an apparent revival of widespread rural or clustered (village or hamlet) settlement in the High Medieval period (around 11th -13th century), along with increased fortification of the larger settlements. The two might be seen together, incastellamento being initially a reaction to various temporary moments of insecurity, while in the long term the fortified centres served as places of refuge for a substantial rural population in addition to their permanent residents. The major longterm concentration of population in large villages and towns could, on this basis , be largely a Late/Post-Medieval phenomenon --which might account for the paucity of good rural sites of this period-- , reversed only in relatively modem times. Generally speaking, the Medieval and later sites (both

70

urban and rural) show a shift in location from the Roman ones, underlining the artificial nature of the Roman "colonial" pattern in this region--Interamna and Fabrateria Nova, after all, failed, while the more natural centres of Cassino and Aquino survive today (the latter, admittedly, after a shift of location). Continuity from Iron Age into Roman is normal in one area only, and from Roman to Medieval is only present in a few spots. Some corresponden;~ can be shown between soils and ancient settlement. The easily worked, fertile soils in the area west of Roccasecca harboured Iron Age sites, though in the other patch of similar soil, east of the Gari, Roman and Medieval sites wholly predominated, with only some possible Iron Age material. Again, there were few Roman sites, and these small and poor, in one area (near Castelluccio) where the soil has a particularly high clay content. But if soils affected settlement patterns, they did not entirely determine occupation. Even in the area· near Roccasecca, Iron Age sites are not confined to one type of soil, but are found also outside it, notably in and near a valley leading from the Melfa-Liri area to the pre-Roman centres of Rocca d' Arce and Arpino. This suggests that even then communication, as well as security and soil types, had an effect. For the Roman period, in an area east of Aquino underlain by travertine and illuviated clay-thus poorly drained--sites of respectable size were nevertheless found. The big terraced villas in the nearby alluvial and colluvial fans of the flanks of M. Cairo do not immediately command soils that are better than those of the plain, although they are better drained because of the slope, and the soil-texture is more granular. It looks as if, to their occupants, proximity to main roads and a preference for being above the plain were important factors. The last can best be described as a cultural preference, as the establishments are mostly not high enough to enjoy any climatic advantages such as lower summer temperatures; proximity to hill grazing may also have been a factor.

Case study: the Interamna LirenasS. Giorgio area Roman Republic and Early Empire A scatter of Republican sites were identified within 1 km of the town of Interamna, followed by a blank area between 1 and 5 km out. This distribution is taken to indicate the development of suburban rural settlement by the 111 century B.C.; previously the tendency was to live in town and go out to till the fields. A second concentration of sites producing Black Slip or coarse pottery of Republican date was found in the Gari River area, but very few sites were noted between there and the concentration near Interamna. Such a gap cannot be explained either by the soil types or by the possible recent obliteration of sites. Furthermore, all the Black Slip ware found is of 2nd-1st century date rather than earlier, and the few conspicuously earlier pieces come from sites in the Gari valley, not from 71

around Interamna. This suggests that the settlement of the Gari valley during the course of the 3n1century represents exploitation of the lighter soils to be found in that area. The land around Interamna itself is considerably less attractive, and it may well be that the colony, founded for strategic reasons, failed to grow; this might partly explain its refusal to send extra troops at Rome's command in 209. The total number of sites producing terra sigillata and early Imperial coarse wares shows very little change from that of the preceding period. However, there is a decrease in sites in the eastern area (from 35 to 21) and a slight increase in the Interamna area (15 to 21). A difference occurs also in the distribution of African Red Slip and related coarse ware. Sites of Imperial date are more numerous than those of the preceding phases, and there is absolutely no sign at this stage of shrinkage or nucleation. Moreover it is in this period that we find a string of sites on the lower terrace, strongly suggesting the presence of a road running below the drop from the upper terrace. The increase in sites around Interamna may represent fresh colonization under the Second Triumvirate (30's B.C.), despite the suspect nature of the entry for Interamna in the Liber Coloniarum. In the Gari valley, on the other hand, it may be that some smaller farms were swallowed up by larger ones, or gave way to new sites. The pattern is certainly abnormal for this period, when all other surveyed areas show an increase in the number of sites.

Late Roman and Early Medieval

If the 2nd century A.D. may reasonably be described as a flourishing continuation of earlier conditions, the ensuing period seems to have brought radical changes. In the 2nd century the number of sites reaches its peak, again with some regional shifts. Around Interamna the site-count drops from 21 to 17, while it increases from 21 to 39 in the Gari valley, and new sites fill the previously persistent void in that region. Some of these intermediate sites seem specifically designed to exploit the valley flatland, which could attest change in land use or type of farming. Only from the 3rd century onwards does the whole area register a sharp decline in the number of identifiable sites, which shrink to a mere 7 (2 in the neighbourhood of Interamna). However, 5 of these can be classified as major, indicating an even sharper decline among the small outlying sites. Massive depopulation of the entire area is clearly a possible explanation, with the loss of Italy's tax-free status in the late 3rd century one possible reason (uneconomic small farms may have been consolidated into larger holdings), although because the dating of coarseware finds is difficult, the same evidence may mean that the population remained, in humbler living conditions. Inscriptions and possible villa sites attest some presence of local aristocracy in the region. A bridge between late Roman and Medieval occupation is not yet archaeologically detectable in the Interamna area, in part because of the impossibility of dating accurately the

coarseware. It is also likely that the renewal of dangerous times saw occupation shrink to those restricted areas safer from invaders, such as the side valley south of Pontecorvo, which has in fact produced slightly later monastic and habitation sites.

Fann sizes (from notes by E.M. Wightman) In the very broken country east of Interamna, some hypotheses about the sizes of the Roman farms offer themselves from the distribution of sites. It is a reasonable assumption that each site would farm, at the very least, the headland at the base or end of which it was situated. The headlands tend to provide between 10 and 40 ha of arable land, to which can be added valley pasture. This might be regarded as a minimum, but on the other hand, it is not certain that the land was ever totally cleared. It comes nearest to this condition in the Middle Empire, and some further calculations may be attempted. If a square of 2.5 km side is drawn in west of the Gari, in one of the most densely settled areas, it gives some 13 sites within 6.25 sq. km, and an average of just under 50 ha per site, on the assumption that all sites have been located. Further south, nearer Giuntura, the average space is considerably less, for ten sites are found within 2.25 sq. km. Farm sizes of between 20 and 50 ha (80-200 iugera) may thus have been the norm, with some areas of woodland or valley pasture in addition. The largest of these farms could still have been run by single families with the help of a few slaves or labourers, or may be seen as medium-sized estates to be run for profit on behalf of absentee landowners: at any rate, they hardly qualify as true latifundia. Evidence of centuriation The question of formal land division as a result of Roman settlement and resettlement policy (such as centuriation) was posed, but here the survey achieved little in the way of firm results. Remains of centuriation were already known around Aquinum; in the Liri and Gari floodplains any possible systems will have been mostly obscured by later alluviation. 3

Communication patterns Another matter of concern to us was the reconstruction of the road networks over the ages. Here, again, the work of the South Etruria Survey served as something of a model, though in the Liri Valley the paucity of recognizable paving slabs except on the town sites (where what paving survives is mostly of undistinctive limestone) made for a different approach, one centred on the question of river crossings. Inevitably, the conclusions offered (which build on the work of others) must remain rather tentative. One specific case study is presented above.

Roman roads in the Interamna survey area Ancient roads linked Interamna with Casinum, Ausonia, and the area east of the town. The Interamna-Casinum road poses no problems: it is clearly distinguishable as an alignment in air photographs, banks down to creek beds, and some sites. The same road, or a parallel track in places, was used also in post-Roman times; this linked Casinum with Esperia and crossed the Liri by ferry slightly upstream from Interamna (Fig. 46, Site 484). A Medieval village (Sites 480-483) was found to mark its subsequent course. The Interamna-Ausonia road may be identified with Cicero's Via Herculanea leading from Fonnia via Aquinum to Arpinum (Giannetti and Berardi, 1970). The point at which it crossed the Liri has caused considerable speculation. A road descends from Interamna to the Liri River, and remains of a bath house stand close to the likely crossing point. Sightings of bridge remains here at times of exceptional low water are reported by local inhabitants; perhaps these remains are now buried, as we could not observe them. The continuation on the south bank was unearthed and destroyed within living memory. Burials flanked its track, which headed south for some 500m, and then ran SW toward the Ausonia gap in the southern mountains. East of lnteramna there are two roads. One, following the ridges between Piumarola and Panaccioni, is marked by sites (though none link up with Interamna). This road may be more ancient than lnteramna itself. The sites (337, 395-6, 401,404, etc.) flank the road closely, underlining perhaps their comp'aratively work-a-day character; luxury villas are absent. The road continues toward the Gari River, and seems to cross it in the vicinity of Colle Canne. A branch of the early road may have forked off just south of the Rio Poppeto, and followed a ridge route down toward the travertine bluff of S. Angelo in Theodice; this is a natural point at which to cross the Gari River, and was certainly utilized in the middle ages. A second route is suggested by a line of sites (442-4, 450, 461) just below the higher terrace. Such a road avoids the recent alluvial zone and the obviously floodable parts of the valley in the Facciano and Giuntura areas. It approximates to the line for the Via Latina Vetus suggested by Giannetti (1974c). However, the pottery indications are that this road was developed in Imperial rather than the suggested Republican times. It leads directly to the junction of the Gari and Liri rivers where both a Roman and a Medieval site developed,

Epilogue Despite these difficulties, the revelation that a dense rural settlement pattern (admittedly somewhat less dense than in favoured South Etruria) existed in Roman times, and the plotting of it · in relation to local geology and soils,

72

represent significant advances. Equally, the discovery of the minor medieval sites fill in a somewhat one-sided picture created by the surviving documents. The local prehistoric evidence remains rather disappointing, the result of significant movements of soil which have eroded or covered up potential sites, or, in the upland areas, recent reversion to dense scrub which offers few exposures of ancient remains. Here chance finds will continue to do as much as survey to elucidate the picture; a good deal of our survey evidence in fact comes from such temporary exposures. In the lowland areas, however, we arrived just in time: modem redevelopment, including at times largescale earth-moving, is progressively confusing the picture on the ground. Our duty to future generations of researchers has therefore been to present the evidence available in the early 1980s, as part of a constantly changing picture.

Notes 1.

2.

3.

Positive identification of this road as the Via Latina is here deliberately avoided, since the present aim is to describe our survey results. For discussion of its identity, see Wightman, this volume, pp. 31-32. The crucial feature here is the near-absence of African Red Slip Ware of classes A2 and, particularly C (in the Italian TS Africana tenninology); the Al category is the latest to appear regularly on the Liri Valley sites. The later wares are somewhat more in evidence on the town sites. In South Etruria these categories are still quite common. French researchers have suggested a divided network for Interamna; our researches can neither confirm or deny it-readers must pass their own judgement.

73

Figure 1. A. Tractor driven plough (at Site 467). (EMW) B. Results of deep ploughing of land (to depth 60-70cm). (EMW)

74

'

.\ ; :':'_~ .">< -•.-, ~

' ....~

·~...~..,,

' ;.....

0

t

ROM

!-·---

.-.

0

Figure 2. Physiographic map of Italy with location of Liri R. Valley.

75

100km

,

I .'

•

"'~.J{~ ,:' !

~·'

~ -

'>:,

·... ~l ,-,!

.··} )

·r;

• .

.

l

•

/

1'\,,

.

. .

.

l J ..

'

(

j

1

.;

. •'.. -1 ~ ,,. ._ I

>

76

• •f

l=i == ="---.,.10==== =1•5--,...20 milu 1 o

1s

Jokm

Height in metres rculoe 1

Figure 4. Map of Liri River Valley and surrounding area with locations of ancient towns (after Salmon , 1967, map 2). (By permission of the Cambridge University Press)

77

M.

Cairo

Figure 5. A. Cross-valley view of Liri River Valley looking east with M. Cairo in background (1PM); B. Bank of the Liri River (1PM).

78

Figure 6. A. Survey team at work in the dissected area. (EMW) B. Freshly ploughed site (186) near the Parito Villa, showing dense pottery and tile scatter. (EMW)

79

N

2 ◄ 0

I

'' '

'O

~

..

'O C

0

5

;;; ()

~fIHDIDI D ,ldft/Ntf

:ll

tt no

Figure 7. Quaternary geological map of the Lower Liri River Valley (after Devoto, 1965).

80

Figure 8. A. Lacustrine marls (light horizons) alternating with waterlaid volcanic tuffs (dark layers) (1PM); B. Air photograph of terrace (barbed and dotted lines) and low-lying floodplain at the confluence of the Gari and Liri Rivers (Giuntura).

81

Figure 9. Location map of Lower Liri River Valley with indication of topographic cross-sections (A-F) shown in figure 12.

82

200

m

0

C:

0

cu > Q)

w

''

'

100

MELFA-

URI

GARI-URI 0

24

0

48 km

Distance

Figure 10. Longitudinal cross-sections (profile of equilibrium) of Melfa, Gari and Liri rivers, terraces (Tl-T4) of the Liri River and terraces of smaller tributaries (large circles).

83

Figure 11. A. Scarps separating first Liri River terrace from the higher bottom valley main lacustrine plain upstream from Pontecorvo (1PM). B. Gentler scarp separating lowest terrace from modern narrow floodplain of the Liri River, downstream from Pontecorvo (1PM).

84

m 110

(a.s.1.\

URI A.

90 _

T1 _________

SCOPETELE

MELFI

l ------1

A I I I

n,

R. FRATTA I

100~

I

t

MELFA

T3

50

60

_

STERPARELLE

L~-----

T1

I

B

I I

T3

LIRI \ \

m

L

C. CARLOMUSTO

\

PIANA DI RAVANO \

80~

C

\

4J

LIRI / /

80

~

M. TISEO

INTERAMNA

/

URI

4J

M. FELICI

A. SPALL

D T3 \

C. SILVESTRI L

\.

C. VALENTE

C. NARDONE

\

E

\

URI T3

I

I C. OLIVIERI

m

A. FAIO

I

I

L

I

URI

T1

T3

F

I 0

1 km

Figure 12. Topographic sections across valley (see figure 9 for location). L= lacustrine plain, Tl-T3 Liri River terraces.

85

Figure 13. A. Southernflank of M. Cairo. In the foreground:part of the valley floor is shown grading iuto small fans of the mountainrise and the steep slopes of the lower mountainslopes. 1 = High mountainslopes, 2 Uplands, 3 Lower mountainslopes, 4 = Mountainrise, 5 Valley floor (1PM); B. Alluvial fan,

=

=

=

site of the Roman Parito Villa (Site 180), lookingwestward toward Piedimonte (in the backgroundon hill at right). (1PM)

86

100 1 GRAIN SIZE

.,,

sot

J

----

100 , GRAVEL

...

O '---

t Bk - I

Lo - Lv

J c -Bk

Lo -Bk

AQUINO PLAIN

DISSECTED AREA

TERRACES ':Ill

Lo - Hh - Rc

ln l Ve - HI

:~:l;-:~:I: I

I I

200

I I

I I I I; I

MTN Mt;!R~ RISE SLOPE I

LF - Lv

Lo - Hh

I I I

I I

I

I

I I

I I I I I I I I

i

z

0 ~ C

~ 00

-

10km

64

~

Roots

I-=:. I Clay-marl

5:d:]

Fracture

D

[IHJ

Blocky structure

~*frfk] Lacustrine

l~I

Peat

~

Bedrock

~

Pebbles

~

Dolostone

23

25

83

profile No .

sand

v-

zoo

Figure 14. Section across the valley showing major morphological features, Pleistocene sediments and surficial soils. Soil nomenclature follows the PAO-UNESCO (1974) classification: Lf = Ferric Luvisols; Lv = Vertie Luvisols; Re = Calcaric Regosols; Jc = Calcaric Fluvisols; Hh = Haplic Phaeozems; Ve = Chromo Vertisols; HI = Phaeozems; Lo = Luvisols; Bk = Calcic Cambisols; I = Lithosols (surficial soil distribution from Sevink et al., 1984).

87

Figure 15. A. Modem bridge and partially destroyed retention walls in the Melfa River, just upstream from the river crossing at S. Vito (IPM); B. Remains of bridge at the Roman crossing of S. Vito on the Melfa River (1PM).

88

- ------

-----

------

- --- 7 I

I I I I

a:

w

a-.

•

I I I

I

- - - - - - --- - - - - -- -- --- ---

'-

Q)

I

a.

-

.....

• • •

Q)

0

C

0

..... 0 Q)

I

I I

•

u, u,

--

- - - - --

--

--

- - __ _J

~" '-

0

~ 3:

0 '-

Q) ..... «s

3:

• •

•

0

• • • en I =1 ~I C)

C C

0

a:

-w

•

• • • •

(l_

0

• •

•- ---•

E

LO

·ca "--

t-

z◄

• •

0

Figure 16. Diagrams of remains of left pier of old bridge at the Roman crossing of S. Vito over the River Melfa (full dots = wooden piles; open dots: location of missing wooden piles; long dashes on map view = wooden tables; from Martini et al., 1987).

89

Figure 17. Lithics from ·Santa Scolastica area (Site 207).

1.

Random flake.

2.

Irregular chunk.

3.

Utilized flake, white with brown bands.

4.

Blade fragment, translucent chalcedony. Much wear on both edges.

5.

Levallois point, tan, with facetted butt.

6.

Side-scraper or racloir, mottled grey chert with many imperfections or inclusions.

7.

Flake scraper.

8.

Flake fragment, grainy white chert. Scraper retouch along one edge.

9.

Fragment, possibly of scraper. White/grey.

10.

Possible scraper fragment, brown.

11.

Convergent scraper, unbroken.

12.

Irregular chunk lacking signs of use.

13.

Side scraper.

14.

Side scraper.

15.

Fragment of convergent scraper (or utilized piece).

16.

Core fragment.

17.

Core fragment.

90

_; : u-a~ I

I

_Q}f I

I

-

I

-

4

-

3

I

s

tj

-

I

I

6

I

8

~

7

.. (1)

4-,

0

~

~




211 . ,, ,.

l 1618-4, ,~ 211

•,n

•

~

196

1; ,

· 2'90

332 • •

11,.

11•

n,

\

e l34

333

_..,

i

. s21 . s 2s

•

•

'

-;;;,

327

5.32

.122

-~9 1i\ )ha : •'Jo:'36'1• ~''""' 9t...\...... "3 70 ·" ;0. ~ J... ~ , \\

-V.1 1-

~

"lb.S

_/_

Sf'

595

_,,,_;:r ~

378

....J

•

•~'/ ;JI'

= ~~ , •.H1r• '-

l\ \

329 ----....

. 128

.no n1

'

•

~: •$33

~j~

54 1

/"\.,4' ,,... __ .

:i:_.C---,

381.

f"\ • -

'--,

C

.) .-·,

•

'•

/ -~· .,,,-·

'\.

-

\

Liri Valley Survey

/

Ind ex Map

•• -•

\

• Sunr-,td Arta

j

.

\.

..

ii1 ••

, ..6,..

• -~ »f

·, . 46l e

~·

~

}

•'°'

, .,,

'-·

}

'¥'

503

.. .. .-

'°J • ••r

508

1?-..., _ "'"'?-----\.. ~ ./

- , ;j:7 ·e/1r

'~ \,. A

r

\V(·l~ GLJ

j

,,.,

J

i~(...,_ y•~

.

(_'

\

I

11...,_./

..,u,. ·-.-·J

,::>

514 ._t

/:_,

'

os

'"-)

Ui

iSo

•

/•

·':;J'Y

-:;---·

j ..~~-~.,..;'· ~ • ,,,,, ______ _,, ~'¥'

e \ \ { _

onesS • 6 • 7

I\'

~\

/

_?l \~\,..',,-. ;I\ \

~

-------.......-

,,,,...,,...1 . ss1s :

F Lin·

/.

.4- __ ........

/ ;' ~ ~J::'- ----

.-................. . .

420

/

\

•

53!

/ 'f'_;.1

-........

52

•

·t·.: . ) .. /' ....

~-....

8

,..

..... ,.

4 19

.. ,

0

'\ ·

~

19]

•

, 318

(\.___

• ' -..j :;:;0 -, ..,,;,,· (_ .....-·--...-·-·•-< '\.✓ ,.,__;;; -~~i. __ . ·"' _ ,,.} _,,. _»~\

. 210

. 3, :

'-,

..;;3

•

1~:. • ,,.

]'23.

~~

~,~. it,. t

~

/•

•

-~ ~.

•

·n o . 21 1"

• • 311 312

JObJ07°

•

/ 0

•

•

315•• 110 • 3~7

>

/

I

\1267

•

J05 •

JO I

,.J

•. ➔ 260

211

•

•

)

--is 1 258 ,

;(. • 250

/

•264 1

. 26~

01 ,.

• .,,,,,-·

•

I

. 26 1

t;°• , 1>1

/

.2•9

•

.-•j

I .......

~

°'

2ol ...

.,

•

Figure 41. lnt eramn a: A. Site 354, with ancient tank broken in two; B. Detai ls of pro cess ing tank (for wine?) visible on pr evious photograph; C. Roman bridge on the Spalla Bassa creek near Inte ramna.

116

Figure 42. A. Inscription of M. Clodius Anoptes from Interamna (Send of site, near Mass. Morra)= Giannetti (1969) . 62, no. 2, pl. IV.3. (1980 photograph) 8. Cross section of water channel exposed in modem foudnation trench (Site 228).

117

~~

·,.

+~ C

__...r--......

.-•,

,..J

I / \ ~--., l\ ,, ) (

.l

\,~302,

.

{■'

\

a: :

',

I

/

I

. :-•5; -I·· ,,

'