Archaeologies of Water in the Roman Near East: 63 BC – AD 636 9781463217754

Table of contents :
Table of Contents
List of Illustrations
List of Tables
Preface
Acknowledgments
Abbreviations
Introduction
1. The Introduction and Uses of New Water Technologies
2. Water and the Economy
3. Attitudes towards Water as a Resource in the Roman Near East
4. Water, Hygiene, Cleanliness and Purity
5. Water and Pagan Religion in the Roman and Late Roman Near East
Conclusions
Bibliography
Index

Citation preview

Archaeologies of Water in the Roman Near East

Gorgias Studies in Classical and Late Antiquity

8

Gorgias Studies in Classical and Late Antiquity contains monographs and edited volumes on the Greco-Roman world and its transition into Late Antiquity, encompassing political and social structures, knowledge and educational ideals, art, architecture and literature.

Archaeologies of Water in the Roman Near East

63 BC – AD 636

Zena Kamash

9

34 2013

Gorgias Press LLC, 954 River Road, Piscataway, NJ, 08854, USA www.gorgiaspress.com Copyright © 2013 by Gorgias Press LLC

All rights reserved under International and Pan-American Copyright Conventions. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning or otherwise without the prior written permission of Gorgias Press LLC. 2013

‫ܘ‬

9

ISBN 978-1-61143-421-7 Second Printing

Library of Congress Cataloging-in-Publication Data A Cataloging-in-Publication Record is Available from the Library of Congress. Printed in the United States of America

TABLE OF CONTENTS Table of Contents.....................................................................................v List of Illustrations .................................................................................vii List of Tables............................................................................................ix Preface.......................................................................................................xi Acknowledgments .................................................................................xiii Abbreviations ..........................................................................................xv Introduction ..............................................................................................1 1 The Introduction and Uses of New Water Technologies.......17 River-fed irrigation ..................................................................31 Aquifer-fed irrigation ..............................................................33 Floodwater farming.................................................................37 Well and cistern-fed garden cultivation ...............................42 Spring-fed irrigation ................................................................43 Teleilat al-anab............................................................................44 2 Water and the Economy ..............................................................75 Watermills .................................................................................75 Tanning, dyeing and fulling ...................................................78 Fishponds (vivaria) ...................................................................80 The rural – urban divide.........................................................84 The rural economy in the late Roman period.....................91 Urban productivity and industry in the late Roman period................................................................................94 3 Attitudes towards Water as a Resource in the Roman Near East ..................................................................................................99 Dams .......................................................................................100 Urban water storage ..............................................................101 Aqueduct-fed storage............................................................103 Nymphaea...............................................................................112 Bathhouses .............................................................................117 v

vi

ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST

The use of water in domestic contexts: status and display.............................................................................120 4 Water, Hygiene, Cleanliness and Purity ...................................129 5 Water and Pagan Religion in the Roman and Late Roman Near East ......................................................................................157 Gods with a watery element to their character.................158 Cosmological associations of watery locations: sacred springs, rivers, lakes, pools and wells............159 Watery themes in religious iconography............................167 Cult of Atargatis.....................................................................170 Maioumas and other festivals ..............................................173 Votive deposition, oracles and divination .........................174 Water, healing and purification ...........................................175 Conclusions ...........................................................................................177 Bibliography ..........................................................................................187 Ancient Authors ..........................................................................187 Modern Authors ..........................................................................189 Index.......................................................................................................227

LIST OF ILLUSTRATIONS The illustrations were prepared and drawn by Alison Wilkins. The photographs were taken by the author. Figure 1: Map of the Near East showing major sites referred to in the text and the boundaries of the Empire at its greatest extent................................2 Figure 2: Schematic drawing of a wheel with a compartmented body......19 Figure 3: Schematic drawing of a wheel with a compartmened rim (noria). .............................................................................................................................19 Figure 4: A 5th-century mosaic from Apamea depicting a noria. ...............20 Figure 5: View of the Roman dam at Homs, showing an area of waving steps on its air face. ..........................................................................................28 Figure 6: Map of the Near East showing the locations of irrigation channels (grey triangles)...................................................................................32 Figure 7: Schematic section and plan drawing of a typical qanat...............33 Figure 8: The southern side of the WF4 fieldsystems at Wadi Faynan, also showing the location and layout of the mill complex (adapted from Barker et al. 1998, fig. 4). ..............................................................................................40 Figure 9: Map of the Near East showing the distribution of irrigation channels (triangles) and qanats (stars)............................................................45 Figure 10: View of the northern gate area at Apamea, showing where the aqueduct enters and feeds into the castellum divisorium. ................................52 Figure 11: Schematic drawing of a vertical-wheeled mill with a rightangled gear.........................................................................................................62 Figure 12: Schematic drawing of a horizontal-wheeled mill with an arubah penstock.............................................................................................................63 Figure 13: View of the limestone incrustation on an aqueduct bridge in Antioch. .............................................................................................................70 Figure 14: View of the kalybe in the forum at Shohba/Philippopolis. ....114 Figure 15: View of one of the latrines at Apamea, showing the large drain beneath the seats, the gutter in front of the seats and notches in the wall for fixing the seats. .........................................................................................139 Figure 16: Map of the Near East showing the location of public latrines (grey triangles). ................................................................................................141

vii

LIST OF TABLES Table 1: Summary of water technologies in the Near East before 63 BC showing their earliest attested date and place. ................................................5 Table 2: Chronological summary of introduction of water technologies into the Near East. .............................................................................................6 Table 3: Contexts of saqiya pot finds. ............................................................23 Table 4: Known dates of aqueducts in the Near East.................................47 Table 5: Presence (X) of public and private water supply features associated with a Roman way of life in selected cities across the East: Antioch, Apamea, Beirut, Caesarea, Dura Europos. ...................................59 Table 6: Presence (X) of public and private water supply features associated with a Roman way of life in selected cities across the East: Jerash, Petra, Umm Qes, Zeugma..................................................................60 Table 7: Bases of late Roman legions in remote areas of the East. ...........93 Table 8: Comparison of urban aaqueduct-fed storage installations by capacity (over 500 m3) in the East and North Africa. Data on North African reservoir capacities is based on Wilson (1997:79-80, table 4).....102 Table 9: Aqueduct-fed urban water storage installations (including in bathhouses) by date........................................................................................103 Table 10: Water sources and storage installations in the East. (A: aqueduct, D: dam, R: reservoir, C: cistern, RC: reservoir-cistern, W: well). ...........................................................................................................................105 Table 11: Nymphaea in the East.....................................................................115 Table 12: Doubtful nymphaea in the East. ....................................................116 Table 13: Sites with mulitple bathhouses in the Near East (not including Herodian palaces). ..........................................................................................120 Table 14: Luxurious or decorative water features in urban houses across all periods.........................................................................................................122 Table 15: Luxurious or decorative water features in urban houses by century (not including unphased houses)....................................................125 Table 16: Orders of drainage. .......................................................................130 Table 17: Latrines in the Near East. ............................................................143 Table 18: Evidence for sacred lakes, pools, rivers, springs and wells across the Roman Near East.....................................................................................165 Table 19: Representations of city tyches and water. ..................................169

ix

PREFACE The inspiration for studying water in the Near East originally came from the discovery of a latrine at Zeugma in 2000. From initial research into the installation it soon became clear that there was a wealth of information from disparate sources about individual elements of water management and supply systems in the area. My aim became to synthesise these data in order to create an understanding of how water was managed across the whole region. In looking at the similarities and differences and rhythms of continuity and discontinuity in technological practices, I began to find avenues into many aspects of the social lives of the people living in the Near East in the Roman and late Roman periods. It is this social story seen through a watery lens that I would like to tell here. Zena Kamash 21 June 2010

xi

ACKNOWLEDGMENTS This work derives from my doctoral thesis, which was completed under the supervision of Andrew Wilson and examined by David Mattingly and Janet DeLaine, all of whom have made significant contributions to my thinking. Thanks must also go to the School of Archaeology, Oxford University whose graduate studentship funded my DPhil research and to the Craven Committee, the Thomas Whitcombe Greene Fund and the Society for the Promotion of Roman Studies who funded several research trips to the Middle East. The audiences at the ‘Fluid Thinking – Water in Society, Mellon Workshop, Brown University 2010’, ‘Hygiene Group – London School of Hygiene and Tropical Medicine’, ‘Oxford-Princeton Conference on Syria in Antiquity, 2006’, ‘Cura Aquarum in Ephesus, 12th International Conference on the History of Water Management and Hydraulic Engineering, 2004’, ‘Water Supply and Use in the Roman World, Oxford University, 2003’, ‘BANEA 2005, Oxford University, 2005’ and ‘Loxbridge Postgraduate Ancient History Conference, Manchester University, 2004’ conferences, as well as those at the Roman Discussion Forum and Graduate Workin-Progress seminars provided much useful discussion and many helpful suggestions. Several people have kindly read parts of this book: Mike Edmunds, Miko Flohr, Hannah Friedman, Chris Gosden, Ted Kaizer and Fergus Millar. I am grateful to all of them for their time and their many useful comments and thoughts, which have helped focus my ideas considerably. Chris Gosden, in particular, has been a constant source of support and inspiration, to whom I owe a great debt of thanks. Finally, my family, both near and far, have always given me unwavering support and love; this book is dedicated to them. xiii

ABBREVIATIONS AAAS ADAJ AE AJA BAAL BASOR BCH BMB ESI HA IAA IEJ IES IGLS JNES JRA JRS PBSR PEQ QDAP ZDPV

Annales Archéologiques Arabes Syriennes Annual of the Department of Antiquities of Jordan L'Année Epigraphique American Journal of Archaeology Bulletin d'Archéologie et d'Architecture Libanaises Bulletin of the American School of Oriental Research Bulletin de Correspondance Héllenique Bulletin du Musée de Beyrouth Excavations and Surveys of Israel Hadashot Arkheologiyot Israel Antiquities Authority Israel Exploration Journal Israel Exploration Society Inscriptiones Grecques et Latines de la Syrie Journal of Near Eastern Studies Journal of Roman Archaeology Journal of Roman Studies Papers of the British School at Rome Palestine Exploration Quarterly Quarterly of the Department of Antiquities, Palestine Zeitschrift des Deutschen Palästina Vereins

xv

INTRODUCTION Water is one of the most benign, and destructive powers, in the lives of all people. In semi-arid and arid zones such as the Near East, the presence or absence of water holds the fate of people in the balance. Water must be controlled, not only so that there is enough to sustain human life, animals and crops, but also to prevent the destruction of land by flooding or salinization. Over the past decade there has been an upsurge of interest in water in Roman archaeology, specifically, and also in archaeology more generally, as well as in other disciplines, such as anthropology and geography. In the case of Roman archaeology, it is becoming increasingly clear how valuable water can be as a tool for interpreting Roman culture (Koloski-Ostrow 2008:556). In other disciplines, the interest frequently derives from concerns about environmental degradation and water shortages, but also from trying to understand how people respond to water in its many guises (eg Strang 2004; Strang 2006). Recently, it has become clear that looking at past approaches to water supply and management may have much to contribute to both themes, as shown, for example, by the University of Reading/Council for British Research in the Levant 'Water, Life and Civilisation' project and Masters courses, such as that at Oxford University’s Centre for the Environment, that incorporate both modern and ancient approaches and findings. Interdisciplinary approaches to water have much to offer as demonstrated by Durham University’s Institute of Advanced Study 20092010 programme devoted to interdisciplinary dialogues about water, which has seen much involvement from archaeologists on themes including water and power, water and identity and water and ritual. It is hoped that the ideas and themes presented here will be of interest and relevance to such interdisciplinary audiences and discussions. This recent interest in water has led to a number of significant projects and publications that have contributed directly 1

2

ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST

to this work (for example Madella, Osborne and Shaw 2009; Syria 2008 (vol. 85); Barker et al. 2007; Amit et al. 2002). The aim of this book is to assess critically in technological and sociological terms the water supply and management of the Near East from 63 BC to AD 636. The dates chosen cover the period of Roman and late Roman occupation in the area, from Pompey’s annexation of the Seleucid Empire in 63 BC to the Battle of Yarmouk in AD 636 when the region was lost. Following Butcher (2003:9) and Ball (2000:6) the term ‘Roman’ here means the period between 63 BC and AD 284 (the accession of Diocletian) and the term ‘late Roman’ refers to the subsequent centuries. The area discussed mainly covers the following modern areas: south-east Turkey, Syria, Lebanon, Palestine Occupied Territories, Israel and Jordan (Figure 1).

Figure 1: Map of the Near East showing major sites referred to in the text and the boundaries of the Empire at its greatest extent.

This work is based on earlier doctoral research that aimed to undertake a region-wide, synthetic study of all the water systems in the Near East from 63 BC – AD 636 in an effort to characterise the methods for the access, transport, storage and use of water across the study area (Kamash 2006a). In the course of this re-

INTRODUCTION

3

search a database, derived from published sources and personal research trips to sites in the region, was created of all the elements of water management systems in the Near East from single drains to complex aqueducts; this database forms the core of the data discussed in this book.1 The original version of this work examined the data in discrete units starting with water supply in the rural landscape (water-lifting devices, dams, irrigation techniques and installations, and aqueducts in the rural landscape) and moving onto predominantly urban elements of the system (water management in cities and towns, water supply in public bathhouses and latrines, water supply in the domestic sphere, and the role of water in industrial processes). Each of these units was accompanied by sets of gazetteers that detailed the evidence brought together in the database, including dating evidence and reliability. This version has taken a rather different, thematic approach, which does not therefore enter into some of the specifics in the original research; for those who wish to access this level of detail, a full and complete copy of the doctoral thesis is available at: http://ads.ahds.ac.uk/ catalogue/library/theses/kamash_2006/ In this book five key themes in the water supply and management of the Near East have been chosen for in-depth analysis: the introduction of new technologies (Chapter 1), economies of water (Chapter 2), attitudes towards water and its conservation (Chapter 3), perceptions of hygiene, cleanliness and purity (Chapter 4) and the role of water in pagan religion (Chapter 5). Chapter 1 considers what new introductions were made to the water technology of the Near East. As tables 1 and 2 demonstrate there was already a wealth of water supply and management in the Near East and an extensive range of pre-Roman water management technologies had been employed in the Near East. Their importance extended from the countryside into the urban environment. Furthermore, the inhabitants of the various kingdoms and regions developed a sensitive approach to water supply, harnessing the water for their benefit and advantage by making use of the resources available to them and by maintaining an intimate knowl1 The database has not been updated since 2006, but is still felt to be representative of the general situation in the Near East.

4

ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST

edge of the potential of their landscape and its resources. From relatively simple beginnings with the development of irrigation systems to complex water-lifting devices, the Near East showed a consistently innovative and sophisticated attitude to water provision and capture. Indeed, the extent of the knowledge and innovation was such that by 63 BC most areas of the Near East were already equipped with water-lifting devices, dams, tunnelling technology, water storage facilities and well-developed irrigation systems as well as a legal infrastructure for the protection of the water supply.2 The introduction of new elements into this well-functioning system, then, raises interesting questions about how and why new technological choices are made. The discussion and analysis in Chapter 1 sets up some of the main threads that run through this book, in thinking about what motivates people to make particular technological choices and who governs these decisions. With these questions in mind, this chapter concludes with a review of the evidence for which agents or groups may have been responsible for funding and introducing new water technology into the Near East; these include: Herod, the Emperor, the army and the Church.

There is an extensive literature on pre-Roman water technology in the East. Some major publications include: Bagg 2002; Baumgarten 2002; Bruun 2000a; Braemer 1984; Braemer et al. 2009; Calvet 1990; Dalley 2001-2; Dalley and Oleson 2003; Durand 1990; Glueck 1959; Hammond 1973; Hodge 2000c; Lewis 2000; Lewis 1997; Miller 1980; Oleson 2000; Schomberg 2008; Wikander 2000; Wilkinson 1995; Wilkinson 1998a and b; Wilkinson 2000a and b; Wilkinson and Barbanes 2000; Wilkinson and Tucker 1995; Wilson 2000b. 2

INTRODUCTION

Element Shaduf Pulley Čerd Wheel Water screw Dams Irrigation – channels, dykes Irrigation – qanats Terracotta pipelines Tunnels Inverted siphon Aqueduct Urban - waterholes Urban - wells Urban storage Urban drainage Domestic storage Private baths Watermill

Date Bronze Age 9th century BC 7th – 4th century BC? 3rd century BC 3rd century BC Bronze Age 7th mill. BC 5th century BC 3500 – 2900 BC 1st mill. BC 2nd century BC Early 1st mill. BC Ceramic Neolithic Uruk Bronze Age Bronze Age Neolithic 1st mill. BC 3rd century BC

5

Place Mesopotamia Megiddo? Susa? Egypt Egypt Ugarit Zagros Mountains Ayn Manawir, Egypt Tell Habuba Kabira Nimrud; Levantine coast Asia Minor Pergamon (then Nabataea); Erbil; Nimrud; Samos; Athens Gar Sur, North Jazira Tell Hamoukar Hauran; Bab adh-Dhra; Hazor; Ta’annek; Arad; Ai; Byblos; Jawa Ur Syria and Palestine Neo-Hittite cities Alexandria/ Byzantium

Table 1: Summary of water technologies in the Near East before 63 BC showing their earliest attested date and place.

6

ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST

Period Neolithic (9th to 4th millennia BC)

Bronze Age (3rd to 2nd millennia BC)

End of 2nd millennium and 1st millennium BC 9th century BC 3rd century BC 2nd century BC

Technology Irrigation Waterholes Wells Ceramic storage vessels Long distance pipeline Cisterns and reservoirs Dams Legal infrastructure for irrigation Drainage into streets Shaduf Large irrigation channels Tunnels Private baths Čerd Aqueduct (?) Pulley Water screw (?) Watermill Wheel with compartmented body or rim Inverted siphon (?)

Table 2: Chronological summary of introduction of water technologies into the Near East.

Following on from this consideration of the funding sources for the construction and maintenance of the major water supply and management infrastructure, Chapter 2 addresses some of the major topics in current Roman economy studies and sets the water supply and management of the Near East into the wider framework of this research. Three main areas are looked at in this chapter. The first section looks at the supposed rural-urban divide and the consumer/producer city debate, using evidence from aqueduct branchlines to suggest a more symbiotic relationship between cities and the countryside. The other two sections consider different aspects of the debates that surround late Antiquity. Firstly, attention is given to the apparent intensification of irrigation and agriculture

INTRODUCTION

7

in the late Roman period across the region and some suggestions are made as to how this may relate to wider economic and military changes in this period. Secondly, the evidence for increasing industrialisation of and encroachment on public space is analysed, specifically thinking about what this evidence can contribute to debates about the nature of urbanism in this period. Chapter 3 tackles an issue of much relevance to modern people: how to deal with scarce water resources. The geography of the Near East varies from well-watered coastal plains to arid and semiarid plateaux. The problems related to harvesting water are not so much related to the amount of water available as to the distribution of that water over time and space. Reconstructing the climate of the Near East from complex palaeoclimatic data has produced several different interpretations. The current evidence appears to point to higher rainfall in the Roman and late Roman periods than at present. A direct relationship between settlement density and climate change, however, may be overly simplistic and the contribution of increasingly effective water management techniques means that the aridity of the period should not be under-estimated (Barker et al. 2007, 334, fig. 10.36; Frumkin 1997; Frumkin et al. 1998; Heim et al. 1997; Hirschfeld 2004; Issar 2003; Klein 1986). In spite of these uncertainties, there are, nevertheless, several climatic factors, in particular rainfall patterns, that are governed by topography and so we can assume that they were similar to the present situation (see Anderson 2000; Fisher 1971; Wigley and Framer 1982 on the climate and geography of the Near East). The highly varied relief of the Near East, in particular the Lebanon and Anti-Lebanon mountain ranges and Jebel Ansarieh, have a strong effect on the spatial distribution of rainfall, which results in striking regional contrasts. Orographic effects, the uplift of cold air masses as they encounter mountain ranges, produce heavy rainfall on the windward side of mountain ranges, but reduced rainfall on the leeward side; most precipitation, therefore, tends to fall along the coastal regions. So, where Damascus typifies the open, arid interior, Beirut, behind which a range of mountains rises, typifies the damper, cloudier coastlands. The action of the mountains is brought into sharp focus by the Homs-Tripoli gap: the region west of Homs is noticeably better watered than areas to the north and south that are in the lee of the Lebanon and AntiLebanon mountain ranges.

8

ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST

The Near Eastern climate follows a characteristically Mediterranean rhythm of winter rain followed by summer drought. In general rainfall begins in early autumn with short showers in September and heavier more prolonged falls in October. A fine period then ensues until December when the rainy season starts. By the middle of June rain has ceased over the majority of the Middle East. The problem for water gathering is that much of this rain can fall in a very short space of time. The Levantine coastlands, for example, can receive higher annual falls than Britain, but this precipitation is concentrated in 6 months of which only 14 to 18 days are rainy. This basic pattern is subject to regional variation, again for topographic reasons. The rainfall maximum occurs in January in the westernmost, coastal areas of Syria, for example, but in February elsewhere. Also, in Israel and the Palestinian Territories, the northern parts of the country receive between five and ten times more rainfall per annum than the south. This has an effect on where dryfarming is possible and where irrigation is necessary. Semiarid zones with less than 200 mm are classed as being outside the limit of rain-fed farming. This should be treated as a guideline as other factors can have an effect, for example soils along wadis may remain moist for greater parts of the year and therefore ‘lobes’ of rain-fed farming may extend into semiarid areas (Wilkinson 2000b:4). As well as annual regional variations, interannual variations are also common. In Baghdad, for example, annual rainfall from 1931-1960 varied from 336 mm to 72.3 mm, and similarly in Jerusalem it varied from 957.7 mm to 273.1 mm. The length and dates of the dry period and the unreliable rainfall regime are important considerations for understanding water storage solutions, for example cisterns. The impact of such interannual variations was certainly felt in the late Roman period in the Near East. In Antioch in AD 362, complaints were made to the Emperor Julian about the high price of food. The arrival of the court and army had put pressure on food production that was already strained from a drought, which

INTRODUCTION

9

continued into the following year (Liebeschuetz 1972:130).3 This event shows not only how unpredictable rainfall could affect food production, but also how that could be aggravated by additional factors, such as population pressure. Cyril of Scythopolis (Life of Sabas, chap. 54) also describes in vivid detail the effect of five years of low rainfall on the water resources of Jerusalem in the 6th century AD: ‘…at the beginning of the 5th year of famine [September AD 520], so great was the lack of water that the poor of the Holy City were begging for water and dying of thirst. In fact, because of the long drought and lack of rain the water had disappeared from the Siloam Pool and the Lucillian Pool; moreover the springs of Colonia and Nephtho were much diminished.’

The underlying geology also has a significant effect on whether water can penetrate into the ground. This has an impact on runoff patterns, and therefore on irrigation techniques because where water can pass into the ground (e.g. in limestone areas) there will be less runoff than in areas where it cannot (e.g. basalt areas). All the mountain ranges of the western Levant have a core of Jurassic or Cretaceous limestone and in many places are capped by later calcareous beds; in many areas the permeability of the limestone has given rise to karstic areas (Weulersse 1940a, 31). On the Mediterranean side of the Jebel Ansarieh range are marine deposits that are less porous than the limestone substratum. In limestone massif regions such as the Lebanon range water cannot pass through the limestone itself, but because the formations are highly fissured, water can pass along the bedding planes and down the joints, which means that there will be more groundwater storage and less surface drainage water to contend with. The Terra Rossa soils, which occur on such hard limestones, have a high moisture-holding capacity, which makes them potentially useful for agriculture, but they are prone to gleying, which can prevent water penetration (Fisher 1971:81). 3 Effect of the army: Socrates HE 17.2-4. Poor harvest in winter AD 361-2: Libanius Ep. 699. Shortage of corn and drought in spring AD 363: Julian Misop. 369; Libanius Ep. 1351.

10 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST The Shefala, or western foothills, comprise low, rounded hills. To the east in the Wilderness of Judea, or eastern desert, chalk and marl replace limestone. Water can easily penetrate certain marls either downwards or upwards; if upwards, a hardpan or crust can form. The marls in the southern part of the Lower Jordan Valley, however, are virtually impermeable which causes a high percentage of surface runoff that contributes to the formation of badlands along almost the entire length of this section of the river (Schattner 1962:17 and 19). During the Pliocene and Pleistocene there was considerable volcanic activity in the region that gave rise to areas of basalt and basaltic tufa. This is particularly widespread on the borders of Syria and Jordan in the Hauran, Jebel Druze and the Upper Jordan Valley. The whole region south of Damascus and east of Hermon forms an irregular plateau dominated by extensive lava flows. The volcanic activity originating in the Jordan valley or in the area bordering it has had a noticeable influence on the flow of the River Jordan and the lava flows have impounded the Sea of Galilee and (the former) Lake Huleh. Lava flows also feature in central Syria around Homs, as well as in the regions of Hama and Aleppo (Philip et al. 2002; Weulersse 1940a:3). This has a significant effect on water management strategies as runoff is high on the basalt, meaning that well- and qanat-digging is often not an option, though some infiltration of water is possible in some areas of the Hauran (Dumond-Maridat 2008:74). To the far south, the Negev desert is mostly covered by thin layers of water-deposited sedimentaries and Aeolian strata. The northern Negev is split into the coastal region of dunes and alluvial deposits and the north-western plains and foothills, mostly comprising Eocene, Cenomanian and Senonian limestone as well as chalk. The central Negev is a mountainous sub-region whose exposed rock strata consist of Eocene limestone in the south-east as well as Mesozoic limestone and sandstone. The southern Negev is different to the rest of the area as it comprises older igneous and metamorphic rocks such as granite and porphyry (Evenari et al. 1982:49, 59). Although there are some deep rock aquifers similar to those found in North Africa and the Arabian Peninsula, aquifers in this region tend to be of the shallow alluvial type. This is particularly the case in Syria where the aquifers are situated in the areas with

INTRODUCTION

11

permeable rock. These shallow alluvial aquifers are recharged along broad gravel-floored valleys in the highlands or on the upper part of alluvial fans, where rivers leave the highland zone and divide into a number of smaller channels. This occurs mostly during spring and summer at the time of maximum river discharge and percolation can be extremely rapid, with all river waters disappearing across the alluvial fan zone (Beaumont et al. 1998:84). The most relevant aquifers, for our purposes, in Jordan are the shallow aquifers. Firstly, there is a system that stretches southward from the Jebel Druze in Syria to Azraq and Wadi Dhuleil. Recharge occurs on the slopes of the Jebel Druze; the water flows radially away from the massif and is concentrated in the Upper Yarmouk basin, the Wadi Zerqa basin and the Azraq basin. The second shallow aquifer comprises alluvial and sedimentary deposits. There are widespread surface occurrences, but they are small in scale and recharge mostly takes place directly from precipitation. In Israel and the Palestinian Territories there are four major aquifers. Firstly, there is the coastal aquifer. The lithologies vary, but in general terms the aquifer is formed of numerous layers of sands or gravels deposited under alluvial conditions and functions well. The northern aquifer is located in the highlands of eastern Israel in Mesozoic sediments with limestones forming the major water-bearing units. The other aquifers are located in the West Bank Palestinian Territories: the western and eastern aquifers that flow west and east respectively. The geological structure of the region also plays an important role in the formation of springs because permeable strata retain large quantities of water that can emerge as springs, for example in areas where limestone gives way to basalt. The extensive limestone outcrops to the south of the Anti-Taurus Mountains, for example, give rise to springs from which the Balikh, Khabour and Jagh-Jagh rivers originate (Fisher 1971:35). The distribution of springs can be highly variable, for example Abel (1933:137) has calculated that northern Palestine has 9 water sources per 100 km2 in comparison to Samaria, which has 7 or 8 per 100 km2 and Judaea, which has only 2 or 3 in the same area. Central Judaea does not possess large areas of permeable limestone and therefore water cannot infiltrate and emerge later as springs, hence the low density of water sources in the area.

12 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST One fundamental difference between regions in the Middle East is the location of rivers with perennial flow and those with seasonal or intermittent flow. The intermittent flow zone is extensive and covers large expanses of the region. It is characterised by wadis: stream courses that are normally dry, but are sometimes subjected to large flows of water and sediment. Their characteristic process is the flash flood whose hydrograph (the expression of flow rate changes over time) has a steep, rapidly-rising limb, a sharp peak and an equally steep falling limb (Anderson 2000:73). Only two rivers in the study area are considered major by world standards: the Euphrates and the Tigris (Beaumont et al. 1998:83; Anderson 2000:74). The Euphrates receives two important tributaries while crossing the Syrian steppe: the Balikh and the Khabour. The Khabour is the last tributary to enter the Euphrates and the river suffers from a large amount of evaporation as it crosses the hot Iraqi plains. Both the Tigris and Euphrates have a distinctly deep thalweg and are at their lowest in September and October. However, due to differences in their catchment areas, the Tigris is in maximum flood in April, whereas the Euphrates does not have its high waters until May. The effect of winter rainfall is less visible in the Euphrates because much of the precipitation percolates into the porous subsoil. In addition to these major rivers, there are other smaller, but still significant rivers in the study area: the Orontes, the Litani, the Zerqa (or Yabbok), the Yarmouk, the Jordan and the Barada. Nearly all flow north to south. The Orontes, however, runs south to north from the Beqaa Valley, near Baalbek to the coast southwest of Antioch (Weulersse 1940a:15, 20, 22, 31, 34 and 38). The largest group of sources for the river is Ain Zerqa at its source in the Beqaa. The regime of the river is quite exceptional in the Levant since it does not suffer from the caprices of the Mediterranean climate. The regime is exclusively one of sources: in winter, the river valley acts as natural drainage from the permeable rocks, basalt and limestone and then in summer, the river is fed by the mountain reservoirs and the slow percolating waters from snowmelt. In general, high waters occur from February to May and low waters from November to January. The Orontes runs in a deep channel, which means that water needs to be raised artificially if it is to be used for irrigation. The Litani River also has its source in the Beqaa Valley near Baalbek, but unlike the Orontes it flows

INTRODUCTION

13

south and discharges into the sea near Tyre (Abel 1933:158). The river is perennial and has its maximum flow in the spring. Both of these rivers break through to the sea. Another major feature of the Middle East are the areas of endoreic or inland drainage, which are produced not only by low rainfall, but also by the land structure i.e. interior basins shut off from the sea by mountain ranges, lava flows across valleys and tectonic basins along fault lines (Anderson 2000, 76). At the lowest level of such areas there is usually an expanse of water, marsh or salt desert; in the case of the River Jordan this is the Dead Sea. The Barada River that runs through Damascus and the River Jordan collect in such endoreic areas. The River Jordan has three principal sources: the Dan River, the Hasbani River and the Banias Spring/River (Schattner 1962:24-26, 36-37). Lake Huleh and Lake Tiberias (or the Sea of Galilee) divide the valley into three major parts. The gradient of the River Jordan is disproportionately steep in its northern- and southernmost parts. The drainage basin of the river is markedly asymmetrical with the catchment area on the east being three or four times that of the west. The Yarmouk (or Seri’at alMenadire) and Nahr az-Zerqa are the two major affluents on the east bank of the River Jordan. The Yarmouk is the most important affluent: its drainage basin is larger than the whole catchment area of the western side of the River Jordan and it supplies almost as much water as Lake Tiberias. It is the junction of three rivers that come together at al-Maqarin near Tell al-Gamid: Wadi al-Aweiret, Wadi as-Sallale and Wadi al-Ehreir (Abel 1933:171-2). In addition, the high discharge of the Yarmouk, which is raised by cloudbursts, is the primary cause of the flooding of the Jordan in January and February. The Nahr az-Zerqa, the second most important affluent, has a source in Amman, but its permanent source is at Wadi adDleil; all the affluents of this river are on its right bank (Abel 1933:174-5). The Zerqa shows effects of direct runoff from rain superimposed on regular flow from springs as well as seasonal snow melt (Fisher 1971:35). How these variations in landscape might be countered may be largely dependent, for example, on what alternative water sources are available. In hard limestone and basalt areas with high levels of runoff, the effects might be countered by directing water using a system of walls and terraces to fields or to reservoirs and cisterns for storage. In aquifer zones, rainwater storage need not be so

14 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST heavily relied upon, if access can be gained, in the form of qanats or wells. Irrigation from rivers can be achieved using water-lifting devices or derivation channels. Chapter 3, in particular, considers to what extent the climate and geography of the Near East was a contributing factor in the design and implementation of water supply and management systems. The chapter begins by looking at the constant-offtake principle, which is a frequently-cited theory in which Roman water supply systems worked on the basis that water flowed constantly through the system and was not stored in significant quantities at any point in the management system (Hodge 1992:3, 79, 89, 279, 280, 296, 303, 322. Also see Forbes 1964:172; Leveau and Paillet 1976; Shaw 1984; Cotterell and Kamminga 1990:51). The principle suggests either technological underdevelopment or a wasteful and ostentatious approach to limited water resources and strengthens the view of urban, consumptive, aqueducts as non-functional luxuries. By looking at dams and urban water storage facilities, this chapter challenges the idea that water systems in the Near East functioned according to the constant-offtake principle. In the second part of the chapter, more attention is given to the cultural impact of concerns over conserving water. Here the low numbers of nymphaea and fountains in the region in both public and private contexts are discussed. It is suggested that the reason behind this pattern lies in deeply-embedded attitudes towards water as a resource that might make these display elements seem overly lavish and wasteful. Chapter 4 builds on this discussion by considering other cultural factors that may have effected technological choices, looking in particular at perceptions of hygiene, cleanliness and purity. The chapter begins by looking at the variety of methods used to dispose of waste water, in which it becomes clear that while modern conceptions of hygiene may not have been at play, there was still a concern to dispose of waste water effectively. The second part of the chapter focuses on private miqva’ot (Jewish immersion pools), public bathhouses and public latrines. By looking at various aspects of these installations, including their spread, distribution and decoration, it is suggested that simple conceptions of the purity of water cannot be upheld in the Near East. Furthermore, Roman ways of living and behaving, such as using latrines, presented such severe

INTRODUCTION

15

challenges to the social mores of some of the inhabitants of the Near East that it was not possible for them to be widely accepted. Finally, chapter 5 adds a new element that was not part of the original research, which is the role of water in pagan religion (the relationships between Christianity and Judaism and water management are encountered in some of the other themes and chapters). In this chapter it can be seen that some of the attitudes towards water conservation seem to reflect long-standing concerns in the Near Eastern psyche as demonstrated by the number of preRoman storm and rain gods. The chapter assesses the evidence for the role of water in pagan religious life. Much of this evidence is literary or epigraphic, but some of the accounts of, for example votive deposition in water, are strikingly similar to those known archaeologically from northern Europe, which suggests that there is more archaeological potential for this topic. Water is a precondition for life and health, if properly managed and controlled. This book hopes to demonstrate how the inhabitants of one region negotiated their ways of life in response to this resource. In addition, it is hoped that the study of one of the most vital resources for life in the region will have elucidated the dynamics behind the acceptance of or the refusal to accept an existing culture and tradition.

1 THE INTRODUCTION AND USES OF NEW WATER TECHNOLOGIES This chapter will assess what new technologies were introduced to the Near East during the Roman and late Roman periods. The main introductions will be discussed starting with predominantly rural technologies such as water-lifting devices, dams, irrigation strategies and aqueducts, before moving onto the predominantly urban technologies such as what happens to aqueduct water once it arrives in cities, how it is distributed, stored and used. Each of these sections will present the archaeological, and where appropriate literary and epigraphic evidence, for the uses of these technologies in the Near East, bringing attention to any specific technological details that are germane to the book or to a wider understanding of water technology in the Near East. Finally, there will be discussion of who may have been the agents of any introductions to the Near East, from individuals such as Herod to wider-ranging imperial agents, such as the army and the Emperor himself. It can be seen from the overview in the Introduction that there was an extensive range of pre-Roman water management technologies employed in the Near East, whose impact extended from the countryside into the urban environment. Furthermore, the inhabitants of the various kingdoms and regions developed a sensitive approach to water supply, harnessing its power for their benefit and advantage by making use of the resources available to them and by maintaining an intimate knowledge of the potential of their landscape and its resources. From relatively simple beginnings with the development of irrigation systems to complex water-lifting devices, the Near East showed a consistently innovative and sophisticated attitude to water provision and capture. Indeed, the extent of the knowledge and innovation is such that by 63 BC most areas of the Near East were already equipped with water-lifting devices, dams, tunnelling technology, water storage facilities and 17

18 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST well-developed irrigation systems as well as a legal infrastructure for the protection of the water supply. What follows will review what major new introductions were made to the Near East and what their significance may have been.

WATER-LIFTING DEVICES Two seminal synthetic works exist on this class of water installation, but it is striking, particularly in Oleson’s work, that there is a paucity of evidence from the Near Eastern provinces (Schiøler 1973; Oleson 1984). To some extent this reflects a survival bias in the archaeological record, but there is also a geographical bias in research. Evidence for some technologies such as simple waterlifting devices over wells is particularly limited, though they must have been in frequent use. Rabbinical literature seems to provide evidence for the use of water-lifting devices, such as water screws and compartmented wheels, in the Roman period of the Near East, but chronological problems make these data very difficult to use convincingly (Oleson 1984:7-9). Overall, it would appear that the techniques available from the pre-Roman period continued in use throughout the Roman and late Roman periods. The two additions to water-lifting devices were water-lifting wheels (norias) and pot garlands (saqiya).4 These wheels come in two main types: wheels with a compartmented body and wheels with a compartmented rim (Figures 2 and 3).

4 Following Oleson (1984:11) noria refers to a hydraulic (water-driven) wheel and saqiya to an animal-driven wheel.

CHAPTER ONE

Figure 2: Schematic drawing of a wheel with a compartmented body.

Figure 3: Schematic drawing of a wheel with a compartmened rim (noria).

19

20 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST Wheels with a compartmented body could only be turned by treading or by a right-angled (saqiya) gear (see below) because the torque was too great for paddles (Figure 2) (Oleson 2000:229-232). Although this type of wheel could lift large volumes of water, it could only do so over a short distance that was less than its own radius. The wheel with a compartmented rim was an improvement on the previous design: it could be water- or animal-driven and used for greater lifts (Figure 3). This second type is well known in the Near East and is commonly called noria. The Septuagint (Ecclesiastes 12:6) probably refers to a wheel with a compartmented rim turned by treading or by a saqiya gear that ran backwards at a well (Oleson 1984:100). The most famous piece of evidence, however, for the use of norias in antiquity in the Near East is the 5th-century mosaic from Apamea (Figure 4). This mosaic seems to depict a wheel with a compartmented rim in a rural or garden scene with strong local elements (eg a camel caravan) (Mayence 1933:4-6; Schiøler 1973:157; Dulière 1974:26, 36-7, pls 25, 26, 62, 63; Oleson 1984:185-6). Dulière (1974:8) notes that this is the only known example of a noria depicted on a mosaic and suggests its subject matter reflected the local technological familiarity.

Figure 4: A 5th-century mosaic from Apamea depicting a noria.

CHAPTER ONE

21

The most convincing piece of literary evidence for the use of norias elsewhere in the Roman world comes from Lucretius (De Rerum Natura V.515) where he compares the eternal rotation of the heavens to ‘wheels and pots…turned by the river.’ This passage does suggest that these installations must have been recognisable outside the East. Another literary reference, in SHA Heliogabalus (XXIV.5), tells a story, set in Rome, of punishment by tying people to a waterwheel, which has been interpreted as a noria (Oleson 1984:96-7). It has been suggested, convincingly, that this was used because Elagabalus would have been familiar with them from his upbringing in Emesa (modern Homs) on the Orontes (Baroja 1954:37). While this technology does seem to have been known outside the Near East during the Roman period, it does not seem to have been a Roman introduction, but rather a technology that may have originated and spread out from the East. The water-lifting devices that required a saqiya gear were bucket chains and pot garlands, which both comprised a string of containers hanging on a turning shaft or drive wheel. By fastening a series of containers to long loops over a wide axle, one could combine the efficiency of a rotating series of scoops (as seen in the waterwheels) with the depths possible when using a simple bucket on a rope or a čerd. 5 These were expensive devices usually only used where water was too deep to be lifted by other devices (Oleson 2000:256, 258). Both devices needed to be propelled by a tread wheel or a saqiya gear (Schiøler 1973:267), which is essentially a watermill gear in reverse. While bucket chain types with wooden buckets are known from, for example, London (Blair 2002; Fitzpatrick 2002; Blair and Hall 2003), Pompeii (Oleson 1984:177) and Cosa (Oleson 1984:361), conclusive evidence for bucket chains has not been attested archeologically in the Near East. It is possible that the evidence for the wooden buckets has not survived due to unfavourable preservation conditions. A shortage of timber in large areas of Most water-lifting devices only have a short lift: a wheel can lift water only approximately the height of the radius; the shaduf c. 78% the length of the tip-beam and a water screw less than half the length of the barrel Oleson (2000:251). 5

22 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST the region also may have hampered their use. It could be argued that where timber was more easily available, for example in the Lebanon and Anti-Lebanon mountain ranges, the noria was the water-lifting device of choice, especially on the Orontes River. The pot garland was a late variant of the bucket chain and seems to have come about with the ‘invention’ of saqiya pots that could be lashed to a wheel or looped as a necklace over the turning shaft. With the use of pots, rather than wooden buckets, cheaper, mass production became possible, probably giving rise to their increased use (Oleson 2000:350, 354). So, it is also possible that, by the time of the introduction of this kind of technology to the Near East, pot garlands had largely superseded bucket chains. Archaeological evidence for pot garland installations in the East is more widespread, and comprises the well from which the water was lifted and/or the pots in which the water was raised. It is possible that wells without saqiya pots may have functioned with bucket chains. Saqiya pots often have a characteristic knob at the bottom, are ribbed and have wear marks caused by the rubbing of the rope (Schiøler 1973:101; Oleson 1984:358; Guz-Zilberstein 1995:324; Ayalon et al. 2000:220). Of the 46 sites in the Near East with saqiya pots of the Roman and late Roman periods, 41 are in Israel; this probably reflects a research bias and it is very likely that more sites exist in the rest of the Near East (Ayalon et al 2000:2256; Schiøler 1973:100; Decker 2001:126, 129; Ein Gedy 2002:90*; Peleg et al. 1992:143, 146; Guz-Zilberstein 1995:324; Stern and Berg 1995:104, 233, 264; Haddad and Velednizki 2003:37*-38*; Segal 2003:36*-37*; Getzov 1993:21; Sussman 1976:99; Oleson 1984:217, 355, 366; Johnson 1988:figs 751, 754, 759; Harper 1980:340; Baumgarten 1999:66*; Ginzburg 2000:42*; Birman and Wienberger 2001:40*-42*; Lauffray 1991:125; Orssaud 1991:248). The pots were found in late Roman contexts at 43 of these sites, which seems to match the pattern found in Egypt. Only 13 pots have recorded contexts (Table 3). The two bathhouse examples have parallels in Egypt and Italy. It is unclear in the case of the pot associated with a kiln whether the pottery was being manufactured at the site or whether a saqiya was being used in the production of the pottery, lifting water from the pool recorded on site, or even both (Getzov 1993:21). If a saqiya was being used for industrial purposes, this would add a so far unattested (but not unexpected) use of this device in the Roman Em-

CHAPTER ONE

23

pire. There seems to be, then, a large amount of evidence for the use of pot garlands in the Near East. While Oleson (1984:354,379) may be correct to suggest that this kind of technology was largely restricted to Egypt in the Roman period, it had certainly reached the Near East by the late Roman period, if not before. Context Number of pots 3 Well Bathhouse 1 Well in bathhouse 1 Cistern 2 Drain 1 Pottery kiln 1 Burial 3 Near wall 1 Table 3: Contexts of saqiya pot finds.

So, the Near East exhibited a wide range of water-lifting devices and can be compared to Egypt. Although Egypt undoubtedly possesses more papyrological evidence, devices that Oleson believed to be in common use only there, such as the pot garland, were obviously in widespread use in the Near East too. Indeed, with more awareness of the traits of these installations, for example pottery forms and the shapes and sizes of wells, it is certain that further fieldwork will add many more installations. This use of pot garlands makes the East, and Egypt, stand out from other regions of the Empire. In addition, there may have been more use of norias in the East than elsewhere. There is less evidence for bucket chains and force pumps, particularly wooden ones, than in the western parts of the Empire. While this may in part be due to a preservation bias, it is also possible that it reflects the availability of certain materials, for example wood (though this would also have been needed for norias), or a chronological difference, for example with pot garlands in the East being generally later than bucket-chains, their western counterparts. There also appears to have been a particularly Near Eastern association with norias.

DAMS Given that dams were particularly important in areas such as the Near East with low rainfall, it is not surprising to find that dams

24 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST also played an important role in water supply and management in the Roman and late Roman periods. Dams fulfil several roles in the overall scheme of water management. These include storage of water in the form of reservoirs, provision of water to aqueducts for water supply of settlements (derivation), provision of water for irrigation and water diversion and/or flood alleviation. In most instances dams perform a combination of two or more of the above roles as well as providing water for subsidiary purposes such as milling. Conventionally, there are three basic forms of dam design: the gravity dam, the arched dam and the arch dam (Hodge 1992:80). A gravity dam functions on the principle that it is too massive to be affected by the pressure exerted by water stored behind it. Pressure on a gravity dam is concentrated at its base, hence its design as a wide-based structure. Gravity dams can be constructed out of either masonry (a rubble core with a dressed-stone face) or earth, sometimes with a stone facing. In general, gravity dams are long and low. The arched dam is very similar to the gravity dam since it too resists water pressure due to its weight. It is, however, usually curved in shape. The arch dam, however, functions on a different basis. The arch dam has a convex water face and resists water pressure by transmitting the stress horizontally and hence does not require the weight or thick base of a gravity dam; a modern example is the Hoover Dam on the Colorado River, USA. Almost all dams that survive in the archaeological record in the Roman Empire were either gravity or arched dams and this was no exception in the Near East. It has been proposed, however, that a dam at Dara was not a gravity dam, but an arch dam, which resists pressure due to its shape.6 This proposition has been founded on a passage from Procopius (De Aed. 2.2.13-17; 2.3) that asserts that Justinian built a dam in order to protect the city of Dara from flood damage: Ἐν χώρῳ διέχοντι τοῦ τῆς πόλεως προτειχίσματος ἐς τεσσαράκοντα μάλιστα πόδας, μεταζὺ σκοπέλου ἑκατέρου, ὦν δὴ κατὰ μέσον ὁ ποταμὸς προϊὼν φέρεται, ἀντιτείχισμα ἐτεκτήνατο ὔφους τε καὶ εὔρους ἱκανῶς 6 The others are at Glanum, France and Kasserine, Tunisia (Hodge 1992: 92).

CHAPTER ONE

25

ἔχον. Οὗπερ τὰ πέρατα οὕτω δὴ ὄρει ἑκατέρῳ πανταχόθι ἐνῆφεν, ὡς τῷ ὕδατι τοῦ ποταμοῦ, ἣν καὶ σφοδρότατα ἐπιρρεύσειεν, ἐνταῦθα ἐσιτητὰ μηδαμῆ ἔσεθαι. Τοῦτο δὲ τὸ ἔργον οἱ περὶ ταῦτα σοφοὶ φράκτην ἣ ἀρίδα καλοῦσιν, ἣ ὅ τί ποτε ἄλλο ἐθέλουσιν. Οὐκ ἐπεύθείας δε τὸ ἀντιτείχισμα πεποίηται τοῦτο, ἀλλ'ἐπὶ τὸ μηνοιεδὲς τετραμμένον, ὅπως ἄν τὸ κύρτωμα πρὸς τῇ τοῦ ποταμοῦ ἐπιρροῇ κείμενον ἔτι μᾶλλον ἀντέχειν τῷ ῥείθῳ βιαζομένῳ δυνατὸν εἴη. ‘At a place about forty feet removed from the outer fortifications of the city, between the two cliffs between which the river runs, [Justinian] constructed a barrier of proper thickness and height. The ends of this he so mortised into each of the two cliffs, that the water of the river could not possibly get by at that point, even if it should come down very violently. This structure is called by those skilled in such matters a dam or flood-gate, or whatever else they please. This barrier was not built in a straight line, but was bent into the shape of a crescent, so that the curve, by lying against the current of the river, might be able to offer still more resistance to the force of the stream.’ (Procopius De Aed. 2.3.16-20)

Most importantly, the passage seems to suggest that the dam was designed as an arch, which is important for the history of technology and assessments of the Roman contribution to it. The passage has been thrown into doubt, however, because the remains of such a dam have not yet been found. Three dams have been located at Dara, but are not of the design suggested by this passage (Garbrecht, and Vogel 1991; Sinclair 1989:221, pl. 101). This leads to two possibilities. Firstly, there was never an arch dam at Dara and that Procopius has been misinterpreted and/or was not writing truthfully. Secondly, there was an arch dam at Dara in addition to the ones already found, but it has been destroyed and is no longer visible. The likelihood of this second possibility rests to some extent on the likelihood of the first: how likely is it that Procopius could make up this description of an arch dam at Dara? Although Croke and Crow (1983) have cast doubt on the verity of some of Procopius’ claims, it seems unlikely that Procopius was speaking pure fantasy. Procopius tended to embellish on the truth, rather than lie

26 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST outright. In particular, he claimed that Justinian was responsible for virtually all the changes and new additions to the water works of Dara, but it was more credible that at least some of these should be credited to Anastasius. Therefore, this evidence should not be ignored outright. There is also a second point strongly in favour of the existence of an arch dam. The description Procopius gave was quite precise in its description of the attributes of the dam. Of particular interest is his apparent understanding of fluid dynamics and the effect of pressure and force upon a curved surface. It seems hardly credible that Procopius could have explained only accidentally the action of an arch dam. Therefore, even if there was no arch dam at Dara, it still must stand that the technology was known and available. If the technology of the arch dam was available in our period, one must ask: why was it not used more often? An important issue in the construction style of dams is where they were located, which may offer a first explanation. In general terms, gravity dams are suited to shallow, wide valleys whereas arch dams are suited to narrow gorges. It has been suggested that one reason for the lack of arch dams in the Roman world was that Romans preferred to build dams in broad valleys (Hodge 1992:81). The statistics for the Roman dams recorded in the Near East, however, appear to show a rather different story. Of the 45 dams that definitely belong to the Roman and late Roman periods 18 were narrow and sited in narrow locations such as gorges or small wadis and 18 were broad and sited in broad valleys across larger rivers.7 These figures, therefore, seem to indicate a rather different story, i.e. that the dams of the Roman and late Roman periods do not seem to favour a broad location over a narrow one. Of dams with known lengths from elsewhere in the empire, seven were over 100 m long (Cornalvo, Proserpina, Alcantarilla, Consuegra, Esparragalejo, Böget, Kasserine and up to 900 m at Wadi Caam), three were between 50 m and 100 m long (Subiaco, 7 On the dating of dams, see Kamash (2006b). Also see Kamash (2006b) for details of the Nabataean and Roman dams analysed in this study.

CHAPTER ONE

27

Çavdarhisar and Löştügün) and three were 50 m or under (Faruk, Semalı and Örükaya) (Bildirici 2002; Hodge 1992:82, table 39; Schnitter 1967; Schnitter 1979; Smith 1971). This shows a reasonably similar variety. It may be significant that the majority of the shorter dams were geographically closest in Asia Minor, though as three of these were not included in the empire-wide studies of Hodge, Schnitter and Smith, it is also possible that other short dams may be found by further work on dams in eg Spain and North Africa. This diversity in choice of site is brought into sharp relief by a comparison with Nabataean dams. Eleven of the fifteen well-dated Nabataean dams were located in narrow clefts, gorges or necks of wadis, one was parallel to a wadi and for three geographical data are not available. These data indicate a clear preference for siting dams in narrow locations. It is difficult to see if this pattern extended across the empire, as evidence for pre-Roman dams elsewhere is limited. It is clear though that in the Roman East, at least, dam construction was not constrained by using only one type of location. Therefore, we are still left to explain why full use was not made of the arch design. A second alternative and plausible explanation could be survival bias. It is possible that if breached, an arch dam may leave little or no trace due to its small span and lesser thickness (in comparison to the broad and thick gravity dams), as may have happened at Dara. Thirdly, it is possible that since arch dam technology appears to have developed at a late stage, after the main period of dam building in the East in the 3rd and 4th centuries, fewer dams needed to be constructed and therefore the technology was not used to its full potential. A further important point is that it does not necessarily follow that because a technology was ‘superior’ it would be used. There may, for example, have been only a limited number of engineers able to design and construct such a new and sophisticated technology. This must surely have been an important factor when selecting from a suite of designs. In such cases there would have to be a strong over-riding reason for choosing a more complicated and unfamiliar design (with the potential construction problems that that may entail), over a ‘tried-and-tested’ technique that works effectively. In the case of Dara this reason may have been the need for the greater resistance of an arch dam against violent flood-

28 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST waters, coupled with the fact that the location made its use possible. The presence or absence of stepped courses also presents some interesting evidence for design adaptations in the Roman and late Roman periods. While no dams from the Nabataean period were stepped, at least six of the Roman and late Roman dams catalogued had been stepped on their air and/or water faces. This suggests that this was a technique introduced by the Romans into the Near East; indeed, it was used on other Roman dams in the empire, for example on the Proserpina dam, which fed Mérida, Spain (Hodge 1992:80). In all cases the steps served to broaden the dam at its base. In some cases stepping was only provided for the lowest courses of the dam, for example at Harbaqa. The Homs dam displays an interesting range of techniques, in particular an intriguing waving step on the air face, which as yet has an unexplained function (Figure 5).8

Figure 5: View of the Roman dam at Homs, showing an area of waving steps on its air face.

8 For more photos of the Homs dam (which, contrary to popular belief, is still extant), see Kamash 2006b, figs 3-5.

CHAPTER ONE

29

This diversity of design elements coupled with the variety of locations noted above, leads to some interesting conclusions. One of the most pertinent aspects of the sites chosen for the dams is how well suited they are to gathering water, especially in areas with low rainfall. It has been said that the location of the Harbaqa dam was one of its failures because it was prone to silting (Calvet and Geyer 1992a:126). Two factors argue against this. Firstly all dams by their very nature are prone to silting because the action of the dam against the water movement means that particles carried by the water settle out. Secondly, and of importance here, the dam is located in a particularly good position. Its reservoir is fed not only by the Wadi al-Barde, but also by other large wadis that flow into the al-Barde a few hundred metres further up. In addition, the ring of mountains that almost entirely encircle the area around the dam must also contribute large quantities of runoff water. The design of the Homs dam at Lake Qattina on the Orontes also seems to have been affected by its location. It was shaped like a long, flat V with the point facing upstream into the impounded lake. The reason for this shape seems to be the presence of a basalt spur that does not extend across the whole lake (Hodge 1992:91). It appears that the dam followed the basalt spur across the lake because it provided excellent foundations. Where the basalt spur runs out, the dam changed direction in order to reach the other side of the lake in the shortest distance. Here it would seem is the real art of dam building: the ability to choose the right location and design a dam that would suit the needs of that location most economically. Strategies concerning physical conditions and the locations of dams have not been looked at explicitly elsewhere in the empire, though there does seem to be a propensity for earth gravity dams across broad and shallow watercourses in North Africa (Hodge 1992:84). It seems clear in the East, however, that dams were not constructed according to a strict template and that a more fluid approach was taken for dam designs that reflected the chosen location. There were, then several highly important contributions to dam technology in the Near East that occurred after the arrival of the Romans: a location-first approach to designing dams and progression in construction style. The location-first approach shows that the dam engineers had an in-depth understanding of the physical geography and hydrology of the region and how best to maxi-

30 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST mise its potential in a harsh and unforgiving climate. This approach also influenced the scale of the dams leading to the use of longer dams as well as shorter dams, which has two interesting implications. Longer dams across broader channels or valleys must create larger reservoirs, with concomitantly larger storage capacities. They must also have been large undertakings requiring a lot of manpower as well as a large amount of investment and, importantly, centralised organisation. This is contrast to the high numbers of small dams at Petra, for example, that point more towards to a fragmented solution to water management by dams, possibly at personal discretion (Al Muheisen and Tarrier 2001-2002:517).9 Stepping dams which seems to feature initially in the Roman period, may be only a small aspect of a dam’s design, but is one that adds strength to a gravity dam while minimising the materials used. Furthermore, the late Roman period at the latest saw the important introduction of the arch dam. This design and the surviving description of it reveal a very high level of understanding of engineering principles. The use of the arch dam provides a solution to two of the three modes of failure in dams. While a gravity dam remains the best design against sliding, an arch dam performs strongly against tipping and over-stressing. It appears then that the Roman dam engineers contributed an innovative and fresh approach to dam design and construction, often in a centrally organised framework, as well as developing the pre-existing knowledge and principles. Dam building seems to have been an example of an indigenous technology that moved into the Roman sphere where it underwent ‘fine-tuning’ and was then used in its more sophisticated state in the Near East, its apparent area of origin, as well as in Asia Minor, Spain and North Africa.

IRRIGATION Irrigation plays a key role in any agriculturally-based society in semi-arid areas, such as the Near East. The techniques and approaches to irrigation will be analysed in relation to their environ9 It can be argued that there is some evidence for centralised organisation in Nabataean dams, but it would seem that this is confined to the flood-control dam at Bab es-Siq, Petra.

CHAPTER ONE

31

mental and hydrological setting, which is an important governing element of their use and distribution. Therefore, the installations will be discussed in the following categories: river-fed irrigation; aquifer-fed irrigation; floodwater farming; well and cistern-fed garden cultivation; spring-fed irrigation, and teleilat al-anab.. The economic impact of these systems is discussed further in Chapter 2 in relation to the changes in the late Roman economy. River-fed irrigation Water for irrigation was taken from rivers to the land to be irrigated via irrigation channels (sometimes referred to as canals). Irrigation channels have been distinguished from aqueducts where the channel was wider than 1 m and/or does not appear to have fed a settlement. In our region and period, 33 irrigation channels tapping rivers at 19 sites have been recorded. All of these sites, with the exception of Damiyah, are in Syria. Although Syria has several rivers that are particularly well suited to irrigation channel digging, for example the Euphrates and its tributaries, it is possible that this distribution reflects a research and publication bias. We might expect to find other irrigation channels in relation to other perennial streams in the region in the future, though the lack of known irrigation channels tapping the River Jordan is still surprising and might not be accounted for by lack of research in the area.

32 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST

Figure 6: Map of the Near East showing the locations of irrigation channels (grey triangles).

The irrigation channels in Syria concentrate in the area around the Euphratesand its tributaries with a second, lower concentration in the Damascus area (Figure 6). Several of these irrigation channels flowed within the territoria of late Roman fortresses such as Barbalissos, Callinicum and Circesium. These fortress cities were in areas below the dry-farming precipitation threshold and therefore irrigation channels were vital for their existence, as they could not support a significant population without the use of this kind of irrigation method (Decker 2001:98). Irrigation channels tapping the river directly would not have been possible from the Orontes River as the river valley is too deep. Here irrigation channels would have had to have been fed by norias or, as at Homs, to have been taken from a dammed reservoir. Such tapping techniques would not have been restricted to the Orontes River. Dams at the offtake point of irrigation channels have been recorded or inferred from placename evidence at seven sites on the Khabour River: Nahr Dawwarin (Tell Seker), Haseke (two), Thannouris and Tel Dibs/Thallaba (three) and one on the Euphrates: Auzara (Welles 1937; Lauffray 1983:54; Decker 2001:103). This suggests that the Khabour River, in particular, was well suited to the construction of

CHAPTER ONE

33

irrigation channels by means of derivation dams at the offtake points. Aquifer-fed irrigation Aquifers were able to provide water for irrigation (as well as for use in urban contexts) by the use of qanat technology. A qanat is a subterranean gallery that taps an aquifer, usually located in a hillside or valley side, and leads it to lower-lying ground using gravity flow (Figure 7) (Wilson 2003a:133). Qanats have interested researchers in ancient water supply techniques for several decades and a substantial bibliography has built up around them, including several studies of qanats in the East (Aisenstein 1947; Goblot 1979; Kobori and Endo 1980; Beaumont 1989; Bonine 1989; Honari 1989; Lambton 1989; Ron 1989; Kobori 1990; Safadi 1990; Lightfoot 1996; Lightfoot 1997; Briant 2001; Wilson 2003a). The qanats in this region have been treated with special attention as they may provide clues concerning the diffusion of this technique: a particularly thorny subject in qanat studies. A further reason for the increased interest in qanats in recent decades is their potential to address modern problems of water supply in the Middle East (eg Safadi 1990). This has lead to several useful studies on qanats in the Middle East, for example the work by Joshka Wessels at Shallalah Saghira.

Figure 7: Schematic section and plan drawing of a typical qanat.

34 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST Qanats are particularly well suited for use in areas where surface water supplies are sporadic as the groundwater provides a resource that can be tapped throughout the year even if there are seasonal fluctuations in the water table (Beaumont 1989:13). There is a strong correlation between the location of qanat sites and rainfall, evapotranspiration, topography and geology (Lightfoot 1996:327-9). In the East alluvial aquifers occurring along major river valleys and beneath alluvial fans at the margins of highland areas and larger wadis coming out of mountains are widespread (Beaumont 1989:13-15; Lightfoot 1996:328). These aquifers provide water at a shallow depth and are therefore ideal for qanat construction. Their locations explain the major distribution patterns of qanat sites across the area: Lower Jordan Valley, the western band between Damascus and Aleppo and the Palmyra area (see Figure 9). In Syria 90% of qanat sites lie within 25 km of upland areas and 75% lie completely within piedmont slopes at elevations of 500 – 1000 m (Lightfoot 1996:328). Similarly, all of Jordan’s qanats were constructed in the piedmont zone because of the greater quantity and better quality of groundwater in that zone (Lightfoot 1997:443-7). The majority of Jordanian qanats (across all periods) tap shallow aquifers of 5 – 20 m depth with higher transmissive flow (i.e. where water can flow more easily through an aquifer). These conditions are ideal, but rare in Jordan, so the pattern seems to show deliberate and knowledgeable exploitation of these zones by the qanat builders. The geology of the steppe and desert regions of Syria (and Jordan) has an added benefit: calcium carbonate and silica form impervious layers beneath permeable marly and calcareous formations nearer the surface, which means that the aquifers can be recharged seasonally (Lightfoot 1996:327-9). In contrast, no qanats are found in the impermeable basalt regions of Syria and northern Jordan. In addition, the majority of qanats in Syria are constructed within areas at or below the 500 mm isohyet and in areas of the country with the largest discrepancy between precipitation and potential evapotranspiration. The date of the introduction of qanats to this region is important in qanat diffusion debates. Of 46 qanat lines over 19 sites, however, only 4 lines can be dated with a high degree of confidence within our period; 13 lines can be dated with a medium de-

CHAPTER ONE

35

gree of confidence and over half of the lines (29) can be dated with only a low degree of confidence. A radiocarbon date has recently been obtained for the southeastern reservoir at Andarin, taken from a sample of charcoal in the reservior floor. Provisionally, this seems to date the reservoir to the 6th-7th centuries AD (Mango 2009:76). Another of the relativelysecurely dated lines at Shellalah Saghira in Syria has been given a late Roman date on the basis of an inscribed cross in the tunnel (Decker 2001:122); it is, however, possible that this cross does not belong to the original construction phase and therefore the qanat could be earlier. The dating of the line at Anasartha/Al Hammam is based on an inscription from a large basalt lintel found 400 m north of the qanat shafts and at the eastern extremity of the ruined site: Συμφυής τῷδῃ τῷ παντὶ ἐξ ἄκρας στο[ρ]εν[νύναι πέδιον (?) καὶ ἀλλάξαι] τῷ ὑγρῷ το ξηρόν θείῳ νεύματι τῇ πη[γῇ πέφυ]κε κρα[νάῃ ἧδ'] οὐσία, πρόσφορον πρὸς ὑγίας φάρ[μακον ἑσ]τωσιν τύ[φειν] (?). [Ῥε]ῖθρον χαριζόμενον εὑρών, Γρ[ήγοριος (?) κομ]ιδὴν ἔρκσ[ε]ν, οἴα τῇ πατρίδι ὁ νοσαρθῶν προσῆκον τα[θῇ ῥόος β]ί(ου) καὶ ἀσυλίας [τὸ πᾶ]ν δὴ ἔργον πρὸς ἥβην ἐγειρίστη ..... ἰνδ(ικτιῶνος) ιβ. ‘It is natural in this universe for things to roll to the plains from the summits and for dry to follow humid: by divine good will, fallen from this source, out of this rock came the power to exhale vapours, offering a salutary remedy suitable for passers-by. Having discovered an abundant source, Gregorios worked on its capture, which extended easily to his homeland the stream of life and prevention against illnesses. All this work was carried out in [the year of] the 12th indiction.’ (Mouterde and Poidebard 1945, 207-8)

It is thought that Gregorios must be Gregorios Abimenos, a rich Arab, who in AD 604 restored the city gates and possibly the citadel at Hanaser (Ibid.). This dates the inscription securely to this period. Other attempts have been made to date qanat lines using written documentation. The water from the qanat systems at Hierapolis, Syria may have been used for the cult and sanctuary of Atargatis, for which water was an essential component. It has been proposed that this theory is supported by a reference to a sacred lake

36 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST in Lucian’s De Dea Syria on the basis that this lake may have been fed by a qanat. If this is the case, this would provide an exceptionally early date for qanats in Syria as Lucian was writing in the early 2nd century AD. Wood also notes Palmyrene inscriptions, but none in any other language on the Umm al-Omi qanat at Palmyra (Wood 1753; Crouch 1975:166). This is used as an argument for an early date for qanats at Palmyra (i.e. before the late Roman period), because otherwise one would expect to have found Greek inscriptions as well. Further work on these (unpublished) inscriptions would be very useful given the potential importance of the date of these qanats. The problems associated with dating these systems centre on problems of longevity and finding (dateable) artefacts from key contexts. In the absence of epigraphic or literary evidence, qanats are virtually impossible to date, except by spatial association with a settlement, which may be problematic, or by fortuitously preserved stratigraphic relationships (Lightfoot 1996:324). As is usual, attempts have been made to date the lines by using settlement evidence with varying degrees of success. Lightfoot notes that there is a high level of correlation between qanats and Roman/late Roman sites (termed Roman-Byzantine by Lightfoot) with 40% being found within or adjacent to Roman-Byzantine towns or fortified villages and 50% in the immediate vicinity of smaller RomanByzantine outpost or guardhouse ruins. This looks like good evidence for Roman-Byzantine dates for qanats in this area and is particularly convincing at Qdeym, which would otherwise have had severe problems with water supply. Lightfoot (1996:324) does, however, supply a caveat over these data as he also notes that this settlement evidence is not conclusive and that some of the qanats equally may have been built for use at Islamic sites in the area. This situation has clear consequences for any conclusive assessment of where the East lies chronologically in the transfer of qanat technology. The evidence seems to point to the fact that qanats were present in the area by the late Roman period. There is currently no positive evidence to suggest that qanats were in the region in the pre-Roman period. If we claim that qanats were introduced during the Roman period, we need to consider the source of this introduction. Within the empire, Egypt would be the most likely contender, but qanats were only used to a limited extent in outlying areas. It seems more likely that the influence would come

CHAPTER ONE

37

from Persia via desert trade where qanats were in common and widespread use. Significantly, this would make qanat technology an example of a technology that was introduced into and disseminated around the Roman Empire from the east. If desert trade was the carrier of this technology, the date of the qanats at Palmyra becomes very important and a thorough survey of these systems and their inscriptions must become a priority for research into qanats. Alternatively, qanat technology may have been transferred from the Arabian peninsular, presumably also via desert trade. Reluctantly, however, we must agree that any such theories can only be supposition for the time being. Finally, the size and storage capacity of the reservoirs associated with these qanats is noteworthy in comparison to those found in urban centres. Whereas the reservoirs in urban centres seem to have comparatively low storage capacities with just 1 reservoir out of 23 with known dimensions having a capacity over 10,000 m3, 4 of the 5 reservoirs associated with qanat systems had capacities over 10,000 m3 (if the depth is presumed to be at least 3 m deep). This shows a clear concern for storing irrigation water; the consequences of this observation will be discussed in more detail in Chapter 2. Floodwater farming Work on field systems in the Near Eastern region has seen an increase in recent years; this must be linked directly to an increase in field survey projects over the past decade that have focussed on understanding the greater landscape of the region, for example the Wadi Faynan project in Jordan and the Homs Regional Survey in Syria (Barker et al. 2007; Newson 2002; Philip et al. 2002). Projects such as these, which are based on a strong methodological framework and make good use of innovative techniques and computer technology, have revealed large amounts of useful and reliable information on field systems in the Near East. Work of this calibre is needed desperately in the East as the advent of more intensive modern farming techniques, in particular the liberal use of bulldozers, is rapidly destroying much of the archaeological evidence for ancient, irrigated landscapes. Extensive work has also been undertaken on several sites in the Negev (Kedar 1957; Glueck 1959; Mayerson 1959; Mayerson 1960a; Mayerson 1960b; Negev 1974;

38 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST Kloner 1975; Evanari et al. 1982; Rubin 1988; Rosen and Finkelstein 1992; Finkelstein 1995). A comprehensive, in-depth study of such floodwater farming techniques, in particular the Wadi Faynan field systems, has been undertaken by Newson (2002; Barker et al. 2007). Therefore, rather than repeating this research, what follows will try to highlight some striking aspects of the systems. One of Newson’s (2002:242-3) main findings is that local topographical aspect was vital when establishing a field system. The following environmental factors in particular governed their establishment: an adequate supply of floodwater; a point suitable for floodwater capture in large enough volumes for agriculture; and a large enough floodplain to supply crops for the local population and possibly a surplus for trade. Unsurprisingly field systems were established in locations where they could support settlements (Newson 2002:243-4). These settlements generally fell into the following categories: small permanent settlements founded by direct governmental authorisation (‘powerful establishments’ such as Wadi Faynan); large areas of scattered settlements whose inhabitants engage in agriculture, brought about either by colonising policies or by favourable conditions for settlement. Irrigation field systems associated with floodwater farming fall broadly into two categories: terrace farming and wadi farming (see Mayerson 1960a; Frösen et al. 1998:495). Terrace farming retards the flow of rainwater and diminishes soil erosion from runoff on sloping topography. Wadi farming is sometimes combined with slope terracing methods and is found in two different forms: tributary wadi cultivation and main wadi cultivation. Tributary wadi cultivation, which is used in smaller wadis, ravines and gullies, is characterised by stone walls that traverse the wadi in order to create small plots of land. This technique was also used widely in Tripolitania in North Africa and was studied extensively in the UNESCO Libyan Valleys Archaeological Survey (Barker and Jones 1982; Barker et al. 1996). The walls retard the velocity of the water allowing for deposition of fertile silt on the plots and also raise the water level so that water can spill laterally onto land along the sides of the wadi. In main wadi cultivation dams divert water from the wadi into fields along the wadi banks. Five sites in the Near East have been investigated as areas of terrace farming: Wadi Faynan, Khirbet Abu an-Nasur, Sbeiteh

CHAPTER ONE

39

(Shivta), Sumaqa and ‘Site 637’ (Barker et al. 2007; Waheeb 1996:345; Kedar 1957:182-184; Evenari et al. 1982:114-118; Dar 1999:125-134; Glueck 1959:221). Only Khirbet Abu an-Nasur does not seem to have functioned in conjunction with wadi-farming techniques. Fifteen sites in the east made use of wadi farming: Wadi Faynan, Rosh Ha’ayyin, Jabal Harun, Nakhl, Nahal Hevron, Zikhron Ya’aqov, Horbat Kohal, Jerusalem, Qadesh Barne-a, Sbeiteh/Nessana area, Wadi Mshash, Sumaqa, Ruheibeh, Kurnub and Diyateh. Tributary wadi cultivation seems to have been the favoured type of wadi farming with 10 of the sites functioning on this basis: Wadi Faynan, Rosh Ha’ayyin, Jabal Harun, Nakhl, Zikhron Ya’aqov, Horbat Kohal, Wadi Mshash, Ruheibeh, Jerusalem and Qadesh Barne-a (Barker et al. 2007; Berda 2000:99*; Frösen et al. 1998; Mattingly et al. 1998:332-334; Yavor 1998:42; Negev 1999:88*; Kirk 1938:214, 224; Glueck 1959:261; Rapuano 1999:74*-75*; Heimann 1984:90). Three of these sites, Jabal Harun, Sbeiteh and Kurnub, seem to use both types of wadi farming, though in the case of Kurnub, at least, the types were not used contemporaneously (Frösen et al. 1998; Lavento and Huotari 2002; Kedar 1957; Evenari et al. 1982). Only three sites functioned solely as main wadi cultivation areas: Diyateh, Nahal Hevron and Sumaqa (Sadler 1990:429-434; Negev 1996:129; Dar 1999:125-133). Wadi farming techniques showed varying levels of sophistication and a variety of features were employed in the systems. One of the most complex examples was at Wadi Faynan, where terrace, main and tributary wadi techniques were used (Barker et al. 1997: esp. 31-2; Barker et al. 1998: esp. 13-16; Barker et al. 1999: esp. 276278; Barker et al. 2000: esp. 43-44; Barker et al. 2007: 150-164, 305348). In some ways Wadi Faynan was atypical because of its excellent state of preservation, its large scale and complexity and the high numbers of pottery sherds across its area that helped to refine the understanding of its chronological development (Newson 2002:226-7, 256; Barker et al. 2007:327). The work at Wadi Faynan focussed on an area referred to as WF4, which was subsequently subdivided into 20 units, which themselves were divided into individual field units. One of the main reasons behind the success of this floodwater system was the knowledge of how to channel water over long distances over very low angle slopes. In addition, the various methods used were closely tied to the topography and landscape.

40 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST Firstly, the predominantly flat, northern half of WF4 was characterised by parallel channels. Trial trenches through these channels illustrated that they were clay-lined and filled by water-lain sediments, which confirmed their interpretation as water channels. They could be split into two principal types. One type exploited wadi-water through damming and diversion; these were c. 2 m - 2.5 m wide and formed by free-standing walls. The other type redirected overland flow; these were narrower and predominantly formed by a dwarf wall at the foot of a faced terrace. Channels that tapped the main wadi (Wadi Faynan) deviated to the south before continuing west. They appeared to use the velocity of the floodwaters so that flooding was spread into the main area of the field system. The minor wadis from the south were usually tapped at the confluence of wadi channels to take advantage of the higher volumes of water. Channels here were usually built at 45º to the wadi channel in a herringbone pattern and diverted water into fields on either side.

Figure 8: The southern side of the WF4 fieldsystems at Wadi Faynan, also showing the location and layout of the mill complex (adapted from Barker et al. 1998, fig. 4).

CHAPTER ONE

41

On the higher ground on the southern side of WF4 floodwater was captured where three tributary wadis break through the surrounding hills (Figure 8). In WF4.3 water was directed to certain parts of the field system using a system of sluices and baffles (stones arranged below a sluice gap to spread water as it flows past), so that water could be prevented from running into the main wadi for as long as possible. The system diverted water at point E into a long meandering parallel-walled channel that ran north-west. Small sluices on this channel, some with baffles, allowed water into the upper fields. The channel ended at a cairn, where a junction channelled water to fields to the west, north-west and north. In WF4.18 at the western end of the field system a further solution was used. Here cross wadi walls were constructed at the confluence of wadis. These walls stemmed the wadi flow and forced water out onto the surrounding fields. While Wadi Faynan may have been exceptional in its complexity and ingenuity, other sites also illustrate sensitive and sophisticated solutions to floodwater farming. At Nakhl walls were constructed perpendicular to the tributary wadi walls/dams, thus dividing the area into smaller sections that may have been pools or small reservoirs (Mattingly et al. 1998:332, 334). In addition, three cisterns were recorded in this system. The system at Wadi Lavan, Sbeiteh was built on two main levels (Kedar 1957:183; Mayerson 1960a:34; Evenari et al. 1982:114-118). The upper level, which was laid out at the same level as the alluvial fan, obtained water from a runoff gully and was divided into 8 subplots (totalling 30,000 m2). The lower level (totalling 80,000 m2) obtained water from the wadi itself and was fed by three channels, which supplied water to different sectors of the lower level. In addition to the three channels, three types of spillway served as drop structures that carried water from the upper terrace to the lower terrace, which may represent stages in the development of the system. There seems to have been a general trend towards an increase in floodwater farming in the late Roman period. Several changes seemed to occur in systems that were in use over long periods of time, for example at Wadi Faynan and Kurnub. At Wadi Faynan it has been proposed that wadi down-cutting, possibly in the Roman period, made the earlier floodwater farming techniques based on diversion barrages and terracing on upper slopes less effective. In response to this, the emphasis was changed to capture water at

42 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST lower elevations and to spread it across the lower slopes by means of parallel wall channels (Barker et al. 1998:24). The development of the water management systems at Wadi Faynan was also reflected in other parts of the system such as the re-plastered reservoir inlet and the mill, whose upper level consisted of opus signinum and the lower level a coarse lime plaster (Barker et al. 1997:37; Barker et al. 2007:316-8). Similar problems were faced at Kurnub where the wadi walls that had been sufficient when the wadi was a shallow depression were no longer effective as the wadi cut deeper, with the result that the runoff water ended up being 1 m – 2 m below the level of the floodplain. Therefore, a diversion channel (400 m long x 9.5 m wide) was created that led water to a broad series of terraced fields. At a later date again the diversion channel filled with silt and the lower section of the system was converted into a runoff farm where small diversion dams diverted runoff from small wadis into conduits (Evenari et al. 1982:112-4). These examples illustrate that field systems were not static over time, but rather show diachronic developments and adjustments to their layouts and organisation, which, as demonstrated by the survey at Wadi Faynan (Barker et al. 2007:174), were not necessarily linear or constant. Secondly, and most importantly, some of the developments that occured may not be related necessarily to the social or political milieu in which they took place. More simply they may be related to the environment of which they were part, which itself undergoes modifications. In some cases the changes that we see in the field systems may not mark technological progress or imposition of new ideas, but may be a ‘natural’ response to changing topographic circumstances. Similar observations were made during the UNESCO Libyan Valleys Survey, in which it was noted that the creation of the field system was a knowledge-based activity that relied on an intimate knowledge of the topography, geology and hydrology of the area to be irrigated (Gilbertson and Hunt 1996:224). Well and cistern-fed garden cultivation Evidence for irrigation fed from wells in the Roman and late Roman periods was surprisingly lacking, though one would expect that such a simple technique of accessing water for irrigation of small areas and gardens would have been commonplace. This is probably because the evidence for garden irrigation from wells,

CHAPTER ONE

43

rather than drawing water for animals and humans, is frequently ephemeral, such as small mud channels. Recent excavations of saqiya installations over wells used in irrigation projects at Yavne Yam and Tel Ashdod suggest that further fieldwork may redress this balance (Ayalon 1999:72*-73*; Ayalon et al. 2000:225; Baumgarten 1999:66*). The evidence for the use of cisterns in irrigation and garden cultivation has also not been widely published. One exception is the Monastery of St Martyrius (Damati 2002). This monastery, founded in the early 470s AD, featured three garden areas that were fed by runoff water stored in cisterns. Overflow water from the monastery cisterns, directed via rock-cut channels, fed the gardens at Khirbet ad-Deir and Chariton (Hirschfeld 1992: 153, 159). In the case of Khirbet ad-Deir, this supply supplemented that of the dam/reservoir system, which protected the monastery when the wadi was in flood as well as providing water and soil for the garden. The garden of the Monastery of Euthymius was similarly fed by a combination of cistern and reservoir water (Hirschfeld 1992:200; Cyril V. Euth. 15, 24.17-18). At this site two reservoirs and a cistern cultivated an area of c. 2500 m2. Spring-fed irrigation Eight field systems in the East are known to have made use of spring water rather than runoff or wadi water: En Gedi, At Telah, Ein Yalu, Abu Gosh, Nahal Zippori, Wadi al-Nazazat, Emmaus, En Boqeq (Ron 1966:113; Glueck 1959:201-2; Newson 2002:244; Gibson and Edelstein 1985:143; Patrich and Amit 2002:17; Hirschfeld 2002a; Porath 2002a:34; Fischer and Shacham 2002:406-408; Ofer and Porath 1986:28-9; Syon 1994:46-49; Amit 2002a:174-5). This is a surprisingly low number and may represent a publication bias, rather than a genuine low use of such an obvious source of water for irrigation. At En Gedi and At Telah water from springfed reservoirs watered terraced and walled fields in the surrounding area (Glueck 1959:201-2; Ofer and Porath 1986:28-9). The regular layout of the fields at At Telah, which is uncharacteristic of other field systems in the Near Eastern landscape, points to a highly-

44 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST ordered system that probably required a large workforce to create.10 Newson argues that this may point to input from a Roman or late Roman authority (Newson 2002:244). Although unproven, this hypothesis is tempting and does provide some explanation for these uncharacteristic field layouts in the Near Eastern landscape. The systems at Ein Yalu, Abu Gosh and Emmaus made use of spring flow tunnel technology to access spring water (Ron 1966:113; Gibson and Edelstein 1985:143; Ron 1985:168; Hirschfeld 2002a). These systems stored the spring water in reservoirs before channelling it out into the surrounding fields.

Teleilat al-anab A third type of structural remains, teleilat al-anab, lit. ‘grape mounds’, has also been associated with irrigation in the Near East. These are rows of artificially-created mounds of soil mixed with gravel, covering areas of between 2 to 2,500 hectares that are usually found on hammadas (stony desert areas) fairly close to urban sites; examples are to be seen at Auja, Sbeita, Mishrefa and Abda (Mayerson 1959:20). Mayerson (1959:21-2) has distinguished three types of heap: conical, ridge and ‘flowerpot’. Conical mounds range from 1.5 m – 3.5 m in diameter and from 0.25 m – 0.5 m high; they are usually arranged in rows 2 m –5 m apart. Ridge mounds are long strip heaps that do not follow the contour of the slope (and therefore are not terrace walls); they are usually 2.5 m – 3 m wide at the base, 0.15 m – 0.25 m high and spaced 6 m – 10 m apart. These two types can be found on the same slope. ‘Flowerpot’ mounds are only found on very stony hammada with little or no soil cover and are circles or rectangles of stone with gravel and soil on the inside. They are usually 2.5 m – 3 m in diameter, 0.5 m high and are either arranged in rows at 15 m intervals or in an irregular pattern. Debate on the purpose of these curious structural features has been ongoing since the late 1950s (Kedar 1957; Glueck 1959:218; Mayerson 1959; Evenari et al. 1982:127-147). Overall Mayerson’s (1959) theory, that the mounds provide areas of micro-irrigation in 10 Though recent work on the Homs regional survey may suggest a form of centuriation in this area too (P Newson and G Philip pers. comm.).

CHAPTER ONE

45

which vines can grow and be watered by hand, seems to be the most compelling solution to these soil and stone heaps, though it, like the others, does not yet explain why three different types were used. It seems possible that they had varying functions, for example planting in flowerpot mounds and directing runoff with ridge mounds, which may explain why no single theory can explain them satisfactorily. This analysis of irrigation techniques in the Near East has provided a picture of a densely irrigated rural landscape in places. In addition, it has shown the variety and complexity of the techniques used, each suited to the needs and demands of the local landscape. On a broad scale this can be illustrated schematically (Figure 9), showing that irrigation channels were used in zones distinct from qanats, i.e. irrigation channels in areas with perennial rivers and qanats in the steppe. The different irrigation techniques were restricted by their geographical location, with little or no overlap.

Figure 9: Map of the Near East showing the distribution of irrigation channels (triangles) and qanats (stars).

Another important observation that can be made despite the poor dating of the installations is their broad similarities over time (though individual details may change). One wonders, for example,

46 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST how different the landscape at Auara would have looked if it had been set out a century later. There are only so many ways of irrigating land effectively according to the geologic and topographic circumstances of the land needing to be irrigated; the similarity of these field systems to those recorded in the Libyan Valleys Survey demonstrates this point even further (Gilbertson and Hunt 1996: 217, 224). Rather, the variable is the intensity of the application of this technology. One possible exception to this lack of change may be the introduction of qanats to the region. In spite of the problems concerned with dating the majority of these installations, some interesting hypotheses do present themselves. The high numbers of irrigation channels in use during the late Roman period suggest that there was a higher level of irrigated agriculture at this time. This may also be supported by the qanat evidence, which showed that they were in relatively common use by this period, as well as the probable upward trend in floodwater farming and the evidence of cistern-fed gardens in monastery settings. This is discussed further in Chapter 2.

AQUEDUCTS An aqueduct takes its water from a spring, river or runoff water and transports it to a settlement. ‘Aqueduct’ here refers to the entire length of a water transport channel; channels that did not clearly feed a settlement have been classified as irrigation channels. The channel specus and course includes those parts of an aqueduct where the water is transported in an open or roofed conduit; these can be built or rock-cut. Tunnels pierce ridges, watersheds, saddles between two mountains and projecting spurs. The channel specus is often no more than a small conduit cut into the floor of these tunnel sections. Bridges, often the most visible sections of aqueducts, take the channel over valleys, which it would not be economical to circumvent. Hydraulically, the water in all these parts of the aqueduct system can be considered as flowing in an open channel. When the water runs in pipelines and inverted siphons, it can be considered to be flowing in a closed pipe. An aqueduct may comprise one or more of these components in any combination, i.e. the entire length of an aqueduct may be a pipeline or may have tunnel and bridge sections, but no pipeline. Unsurprisingly, the majority of aqueducts in the region appear to have been constructed in the Roman period reaching the height

CHAPTER ONE

47

of new construction in the 2nd century AD (Table 4), which is also reflected by a peak in nymphaeum and fountain building. There was a clear drop-off in numbers in the late Roman period; this slump in construction probably reflects the fact that the majority of sites already had aqueducts and possibly that the finance for new aqueducts was no longer available (on the cost of aqueducts see below). The early introduction of aqueducts in significant numbers in the Hasmonean and especially Herodian periods seems to be associated with palace construction; the influence of Herod will be discussed further below. Period Hellenistic 2nd century BC 1st century BC Nabataean Hasmonean Herodian Hasmonean or Herodian Total Hasmonean and Herodian 1st century AD 2nd century AD 3rd century AD Roman Total Roman 4th century AD 5th century AD 6th century AD Late Roman Total late Roman

Dated aqueducts 3 1 1 8 7 14 4

High confidence 1 1 0 4 5 10 4

Medium confidence 2 0 1 4 2 4 0

25

19

6

7 12 4 11 34 1 2 2 13 18

4 11 3 7 25 1 2 2 11 16

3 1 1 4 9 0 0 0 2 2

Table 4: Known dates of aqueducts in the Near East.

As one might expect from such complex structures, the introduction, or not, of new elements of aqueduct technology was highly variable. There are few differences, if any, between the Near East and the rest of the Roman Empire in the design and use of

48 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST channels, tunnels and ceramic pipelines. The use and design of bridges/arcades and inverted siphons, however, demonstrate some interesting differences and developments. The majority of the bridges (52) were constructed between the Herodian period (15) and the end of the 2nd century AD (28). The high number in the 2nd century AD probably reflects the peak in aqueduct construction in this period. The bridges varied in length from 9 m to 220 m. The shortest bridges (less than 25 m long) were most common (8 out of 16 with recorded lengths) and belonged to a variety of time periods. Three of the four longest bridges were attributed to the Hasmonean and/or Herodian periods, for example Hyrcania had a 150 m long bridge (Patrich 2002:343) and Alexandrion had a 187 m long bridge (Amit 2002c: 308). The longest recorded bridge was on the Beirut aqueduct and measured 220 m long with an arch spanning 20 m in the centre (Davie et al. 1997:274). This bridge had three tiers of arches and is comparable in design to the Pont du Gard. Five arcades were recorded. Two were dated to the Herodian or procurator period, one to the 1st century AD, one to the 2nd century AD and one was of unspecified date. As arcades were clearly a Roman technology, it is particularly interesting that two of the aqueducts with arcaded sections may be Herodian: Jerusalem High Level and Caesarea High Level channel A. This suggests that Herod may have been an agent of transmission of this technology to the east. Twenty-six pipelines and inverted siphons were recorded; of these, ten have been identified securely as inverted siphons.11 An inverted siphon is the common term used to describe the section of an aqueduct that is led through a pipeline in order to cross a depression (as an alternative to circumventing the depression or building a bridge across it) using the hydraulic principle that water will find its own level. Of the ten inverted siphons, two of which were on rural lines that also supplied irrigation water, three used stone pipes (Jerusalem – High (Mazar 2002a:227-230), Tiberias (Winogradov 2002) 11 On the use of the term siphon see Smith 1976: 51; Blackman and Peleg 2001: 411.

CHAPTER ONE

49

and Alexandrion (Amit 2002c:307-310)) and two ceramic pipes (Caesarea – High (Porath 2002b: 105-117; Siegelmann 2002:134-5; Peleg 1991a) and Emmaus (Hirschfeld 2002a)). This goes against the trend of lead being the most commonly used material in Roman inverted siphons, though stone pipelines are known from elsewhere in the Roman Empire, particularly in south-western Asia Minor, as well as Italy and North Africa (Stenton and Coulton 1986:46, 53 fig. 9; Wilson 2000e:599). Ceramic pipelines in inverted siphons, however, were very rare and were restricted to the Greek world with three other examples from Spain (Hodge 1992:131, 232); in general they appear to represent a continuation of previous, Hellenistic practice (Lewis 1999:157). When one takes into account the preference for ceramic pipelines over lead pipelines in urban and domestic contexts (see below and Chapter 3), it is not surprising to find that for inverted siphons ceramic was preferred over lead too. It is possible that this difference derived from varying lead resources between east and west. The chief sources of lead were in Spain, Gaul, Sardinia and Britain, though small reserves are known in Asia Minor (Healy 1978:61-2; Stenton and Coulton 1986:48). The costs involved in transporting such a heavy material, large amounts of which would be needed in an inverted siphon, may have been a strong reason for choosing local materials. In addition, as for arch dams, it is possible that it was not felt necessary to replace an existing, well-functioning technology, i.e. technologically ‘superior’ options were not necessarily the ones selected from a suite of techniques. Rather, the motivation for innovation may have been more pragmatic. The two inverted siphons on the Apamea line are described as being underground and ‘roofed using large rectangular stones in such a way to form a paved road which is level with the surrounding ground, so any rainwater streams can flow over easily. Also the aqueduct water cannot escape under pressure resulting from water flowing into and out of the siphon’ (Shahada 1957:160).12 This arrangement in the base of the valley illustrates an astute solution to overcoming damage to inverted siphon lines that ran in the base of seasonal wadis. Damage and subsequent repairs 12

Translation kindly provided by Dr Khalid Kamash.

50 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST caused by seasonal runoff have been recorded on sections of the Jerusalem Low Level aqueduct that were located on slopes (Nadelman 2000:161). Furthermore, the inverted siphons at Apamea seem to have been slab-built, rather than being lead, stone or ceramic pipelines; this is of importance as it has been claimed that only three other slab-built inverted siphons are known: one at Alexandrion, one at Lucus Feroniae and the other at Angitia in Italy (Jones 1962:197-200; Amit 2002c:309). We may now add a fourth site to this list.

URBAN WATER SUPPLY In the traditional model of water supply in Roman towns, the aqueduct would terminate at one of several structures: a castellum divisorium, where the water would be divided and channelled to various secondary points in the city such as fountains, bathhouses and private houses; a nymphaeum that flowed continuously; or a bathhouse. In this model, the aqueduct water flows continuously through the system and is not stored in significant quantities along its course: the constant-offtake principle (see Chapter 3). More recently, it has been argued, contrary to this model, that reservoirs (open ‘pools’) and cisterns (covered storage installations) may also have provided terminal points and storage for aqueduct water (Wilson 1997; Wilson 2001). This section will look at the delivery into the town and the use of castella divisoria, then at the points to which the water was delivered (nymphaea, baths, latrines and storage installations) and finally at how it was moved around the city and then removed. The castellum divisorium was a small tank usually located at the edge of the city. It acted as a distribution point where aqueduct water entered the tank as a single unit and exited in separate branches. Three examples of the castellum divisorium are used as the most common paradigm: the installations at Nîmes and Pompeii and the description by Vitruvius (VIII, 6). In the extant examples at Nîmes and Pompeii a water channel entered the tank, the water was then filtered through a grille or screen and then was split into the various branches leaving the tank (Hodge 1992:281-292; Hodge 1996; Ohlig 1996; Ohlig 2001; Ohlig 2002; Wilson 2006a). At Pompeii it is often said that the castellum gave absolute priority to public drinking water. This theory, however, rests completely on Vitruvius’ account, which suggests that some consumers (public fountains) were given priority over others (baths and private users),

CHAPTER ONE

51

and is now refuted; in actual fact the destinations of the pipes leaving the castellum have never been traced and hydraulically there is no priority in the Pompeii castellum (Hodge 1996:18; Ohlig 2001). A third example at Thuburbo Minus in Tunisia had three lead pipes exiting, which are thought, though without secure grounds, to have fed the baths, the theatre and the reservoirs respectively; like the Pompeii example these assumptions are, again, partly drawn from Vitruvius (Hodge 1992: 289). Three other examples are known from North Africa: two from Carthage and one from Simitthus (Wilson 1997:85). Evidence for castella has been asserted at five sites in the Near East: Caesarea, Apamea, Auara, Scythopolis and Petra. The Caesarea castellum was at the end of the High Level aqueduct in front of the inner city wall (Porath and Yankelevitz 1989-1990:131; Porath 1996:112; Porath 2002b:121-2). The device comprised a basin whose external width was the same as that of both channels A and B (c. 4 m), which indicates that it was not part of the original conception when channel A was constructed. The water was diverted into several terracotta pipelines, but the method of this division has not been recorded. No mention is made of a grille or sluice. A castellum in the form of a large cylindrical building (whose dimensions are not specified) has also been recorded at Apamea in the southern part of town 180 m from the colonnaded street (Lacoste 1941:119; Balty 1987:20). The water entered from the base of the structure, issued through a central hole and then departed via 12 holes set in a circular formation around the central hole. Its location in the southern part of the city suggests that this may have been a secondary castellum as recent excavations may have revealed the main castellum at the point where the aqueduct enters the city near the northern gate (Figure 10).13

13 Thanks to Didier Viviers for the information on the recent discoveries at Apamea.

52 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST

Figure 10: View of the northern gate area at Apamea, showing where the aqueduct enters and feeds into the castellum divisorium.

The third suggested device, at Auara, is a monolithic sandstone basin that has traces of a channel cut through its walls (Eadie and Oleson 1986:63). As it was found in several pieces it is unclear that it had any dividing function, which is integral to a castellum divisorium, making its interpretation difficult to confirmm. A probable castellum has also been found at Scythopolis west of the hippodrome (Fahlbusch 2002:62). Excavations are planned to elucidate its nature, but at present all that is known is that there were five distributive systems in the vicinity of the installation. The so-called castellum at az-Zantur in Petra appears to be a smaller-scale junction box (Schmid 2008: 28-9, fig. 20). As well as these installations that have been explicitly called castella, there are also three other sites that have features associated with dividing water where the aqueduct enters the city: Dor, Sbeiteh and Banias. A termination pool fed by a terracotta pipe from the northern aqueduct has also been recorded at Dor (Peleg 2002b:151). There is no evidence for any dividing function in this installation, so it is unlikely that it was a castellum. Its size would

CHAPTER ONE

53

suggest that it might have been a settling pool at the end of the aqueduct. As the aqueduct entered Sbeiteh there was a sluice gate with three openings, two of which directed water to the settlement and one to the North Church (Tsuk 2002c:79). The sluice gate was made of stone blocks with vertical grooves. There does not seem to have been a tank at this point, so while the water was divided, this was not a true castellum as currently defined. A system of 17 small pools was built alongside the aqueduct to Banias at the edge of the city (Hartal 1993:1-2; Tsaferis and Israeli 1995:5; Hartal 1996:6; Hartal 2002:94-101). The pools were not distributed evenly along the course of the aqueduct. A group of eight were positioned where the aqueduct first approached the city. There were no pools along the final 95 m of the aqueduct. The pools were connected to the aqueduct either by an opening in the wall or by a terracotta pipe; three pools featured calix-like devices of lead pipes in conical holes at this point. Each pool had a hole in its southern side (the side exiting the pool towards the city) for a terracotta pipe and two had a pair of pipes; these holes were covered with grilles to prevent debris entering the pipes. The pipes were at or near the bottom of the pools. The suggestion that these pools served as a substitute for a main castellum is not convincing because they were not at the termination point of the aqueduct. Indeed, a termination pool at the end of the aqueduct did exist from which two lines left. These pools seemed to function rather as a series of secondary castella or offtake points/draw basins. Furthermore, they would appear not to have been built as part of the original scheme for the aqueduct, but rather to have been private, ad hoc ventures as none of the pools was identical, but rather constructed out of an assortment of building materials. If the pools were private, it also explains why the calix-like devices were at the entrance to the pool, rather than its exit, thus controlling the amount of water entering the pool from the aqueduct. There are, therefore, only two realistic contenders for the title of castellum divisorium as it is currently understood: Caesarea and Apamea. This indicates that this technique did not make a large impact on Near Eastern aqueduct water distribution, except in two of the most Romanised cities in the region. This is not a surprise as such devices were rare across the empire. Instead there was a wide variety of practices used to divide water as it entered the city. This

54 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST raises the question of whether we should rethink our criteria for identifying castella (also see De Kleijn 2001:34-36 on Frontinus' flexible use of the term). In addition, installations such as those at Banias point to the drawing of water directly from the aqueduct. Nymphaea are found in the Near East, but their distribution is quite restricted with only 17 structures across 13 sites being convincingly identified. Likewise, bathhouses and latrines, which were Roman introductions to the region, demonstrate some chronological and geographical patterning. In both cases, their introduction clearly lagged behind the rest of the Roman Empire, particularly in the case of latrines, which do not become widespread until the late Roman period. The latrines also show a distinct geographical restriction being all but excluded from Judaea. The reasons behind these patterns are discussed in detail in Chapters 3 (for nymphaea) and 4 (for baths and latrines). In brief, the lack of acceptance of these typically-Roman water structures seems to be related to distinct socio-cultural needs, mores and ideals, which are unrelated to perceptions of technical superiority. The water storage installations in the towns and cities comprised reservoirs (open ‘pools’), cisterns (covered storage with a volume less than 1,000 m3) and reservoir-cisterns (covered storage with a volume greater than 1,000 m3). In terms of capacity, over 72% (19 installations) of the urban water storage installations were in the ranges below 2,500 m3 and there was a marked drop-off in numbers in the larger capacity ranges. In general, with the exception of the massive open reservoir at Capitolias (Tsuk 2002a:293; Harding 1967:56; Schumacher and Le Strange 1890:162-166), larger volumes (2,500 m3 and over) were, unsurprisingly, stored in covered installations rather than in open ones, which is probably due to lower evaporation and less risk of water contamination in covered installations. Five reservoir-cisterns and one cistern can be classed as similar in construction style to the chambered cisterns of North Africa. The reservoir-cistern underneath Nea Church in Jerusalem, with a capacity of 5,329 m3, comprised six barrel-vaulted chambers with connecting transverse chambers (Avigad 1983:233-245). The late Roman, Big Reservoir-Cistern at Resafe had two parallel barrelvaulted chambers with connecting transverse chambers (capacity 14,600 m3); it also had settling tanks (Brinker 1991:126). The reservoir-cistern at Dara had ten parallel barrel-vaulted chambers

CHAPTER ONE

55

(Preusser 1911:45; Mango 1974:39; Furlan 1995). Each of these chambers was 4 m wide and 25 m long; assuming a depth of c. 5 m, this reservoir cistern would have had an estimated capacity of 5,000 m3. The two other reservoir-cisterns at Resafe (the Small and the Dome Reservoir-Cisterns) were cross-vaulted chambers with their roofs supported on piers; hydraulically this type of reservoir is single-chambered (Brinker 1991:126, 130-2). These reservoir-cisterns had capacities of 2,050 m3 and 3,400 m3 respectively. The cistern at Dara was similar to these, though with a much smaller capacity (c. 390 m3). In addition to these, tunnel reservoir-cisterns were found at Capitolias (Beit Ras) (Tsuk 2002a:293; Harding 1967:56; Schumacher and Le Strange 1890:162-166) and Sepphoris (Tsuk 2002a:287-293). The example from Capitolias was cut into the limestone and extended underneath the southern part of the city wall; it measured 275 m long, 2.5 – 4.5 m wide and 7.2 m high and had a capacity of 4,900 m3. It is described as being subdivided into small and large compartments by hollow walls; presumably these were piers supporting the roof and therefore the tunnel can be compared hydraulically to the cross-vaulted reservoir-cisterns above. It delivered water via a channel inside the city walls to a reservoir with a 15,500 m3 capacity, the largest in the Near East. Due to the large size of this reservoir, in comparison with the capacity of the tunnel, we can surmise that the purpose of the tunnel was to regulate water supply to the reservoir and to act as supplementary storage if the reservoir was filled to capacity. The Sepphoris tunnel reservoir-cistern, located 1 km east of the city, was very similar in form: it exploited a geological fault in the limestone and chalk, measured 250 m long, 3 m (average) wide and 10 m high and had a capacity of c. 4,300 m3. Its regulating function is more clearly seen than in the Capitolias example. At the end of the tunnel a wall was built, into which a lead pipe was inserted. On the other side of the wall the lead pipe was broken, probably at the point where a stopcock had been. A bronze stopcock has also been found installed in later phase reworking of the outlet point of the Nabataean reservoir at Humayma/Auara (Oleson 1988:123-4; Tsuk 2002a:290-1). These are significant as according to the constant-offtake principle the water supply could not be turned off; the presence of these stopcocks suggests that this was not the case in the Near East.

56 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST On the subject of the constant-offtake principle, however, the largest reservoir (Capitolias: 15,500 m3) and reservoir-cistern (Resafe: 14,600 m3) in the East were significantly smaller than the largest North African reservoir-cisterns from Carthage, Cirta and Zama (Wilson 2001). The significance of this is discussed in detail in Chapter 3 where it is suggested that the Near East did not conform to the principle, but nor did it adopt the same solution as North Africa, rather use was made of complementary sources and storage, which enabled a finely-balanced and tailored response to the water supply needs of different towns and cities. There is little published evidence for non-aqueduct supplies in the form of rainwater cisterns and wells, though it can be assumed that these must have formed a significant water source for the towns and cities of the Near East. As with the wells, pipelines and channels, forming the network by which water from aqueduct and non-aqueduct sources could be moved from one part of the city, and from one building, to another, must have been ubiquitous, yet information on pipelines and channels comes from very few published sites. Lead pipes are particularly under-represented, with only four lead pipelines recorded. Three were from Caesarea, one of which fed the ‘governor’s campus’ (Porath 1996:114; Porath 2000:36*-37*). At Beirut a lead and bronze pipe was found south of the amphitheatre (Lauffray 1944-1945:64). Despite the general lack of recording of pipelines, this low number seems to represent a potentially interesting underlying pattern. This pattern of preference for terracotta over lead is similar to a pattern noted by Jansen (2000b:119-121) in towns first occupied by Greeks; in this case it seems to point to lead pipes being a particularly Roman introduction into the Near East that did not spread widely. As suggested for inverted siphons, this may be due to restricted access to lead supplies and/or a lack of willingness to change technology and construction techniques.

WATER SUPPLY IN DOMESTIC CONTEXTS Evidence for the connection of houses to the urban supply network is limited and there seems to have been a high reliance on rainwater stored in cisterns. While this contrasts with Pompeii, it fits with literary evidence from Frontinus (Aq. 99.3, 103.2, 105.1) and archaeological evidence from other areas of the empire, for example North Africa, which suggests that it was only the wealthy

CHAPTER ONE

57

that had this privilege (Lohmann 1979:178-9; Wilson 1995a). Firm evidence for connection to a piped supply does not come until the 2nd century AD. The supply networks were quite simple, accessing only a few water points. This is a very different picture from that at Pompeii. Jansen (2001:29) has demonstrated that the distribution systems inside those houses comprised three key elements: lead pipes, lead distribution boxes with taps and fountains capping the end of the pipes. The reconstruction of these systems shows that the pipe entered the house and fed a number of fountains around the impluvium, a branch then continued through the corridor to the next courtyard (generally a peristyle) where fountains were again fed by a distribution box and taps. So far, 22 lead distribution boxes have been recorded at Pompeii, whereas none has been found in the East. Distribution boxes have also been found at Vaison-la-Romaine (France) (Jansen 2000b:122). In addition, 112 taps, often near distribution boxes, have been found at Pompeii, in comparison to 3 from the East. While this discrepancy may be due partly to a survival bias as frequently lead and bronze items are robbed, it may also suggest that there was a lesser availability of piped water in the East than at Pompeii. The majority of the houses recorded were supplied with water from cisterns (108 in 92 houses); cisterns from Sbeiteh (Tsuk 2002c:68-73) and Dura Europos (Brown 1944) accounted for 40 and 25 of the cisterns respectively. In areas where water was not easily available, such as the Negev and Dura Europos, the houses were reliant on cistern supply. The proportional bias towards these houses in the data set and then in the cistern numbers may signal that the reliance on cistern water has been overemphasised, in particular against well water. Wells were only recorded at three domestic sites in the region: 1st-century BC Jerusalem (Oren 1972:172), Roman Antioch (House of Ge and the Seasons DH 24-P: Levi 1947:55, fig.17) and late Roman Apamea (House of Consoles: Balty 1982:22). In some cases, cistern water seems to have been used in conjunction with a piped supply. The link between water supply and water for display will be discussed in detail in Chapter 3. The majority of the pipelines in the internal distribution network whose material is known were ceramic: 17 out of 21. Three lead pipes were recorded at Antioch (Seleucia-Upper City: Stillwell 1941:5 and Yakto Complex: Lassus 1938:122) and Beirut (Bey 006,

58 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST Area 2: Perring et al. 1996:195; Perring 2006:28-30). The example from Seleucia – Upper City was fitted with a bronze valve. A bronze pipe was also found at the House of Iphigenia 14-S in Antioch exiting a pool; this may have connected with the lion head spouts that were found in room 1. The use of lead in an area where the use of such Roman building materials is very limited suggests that this was a conscious choice to build in a Roman manner. It is no surprise, therefore, to find that Antioch and Beirut were the cities where this choice was made. A similar point can be made on the use of brick, which was used for three pools in Antioch (House of Menander DH 26 M/N House 1: Stillwell 1941:25-6; House of the Buffet Supper DH 26/27 O: Stillwell 1941:29 and House of Iphigenia: Stillwell 1941:10) and a basin in the palace of the Dux Ripae at Dura Europos (Detweiler 1952:8). Most of the houses in the East had relatively simple water management systems that focussed on supply from cisterns and probably wells. Only the wealthier houses in cities such as Antioch and Zeugma appear to have been connected to an external piped supply. In these houses, additional features, in particular fountains, made use of the piped water to display the wealth and status of the inhabitants. The desire to use water for display was particularly strong in Antioch. Also in Antioch, where there was no connection to a piped supply, pools were used for ostentation and as status markers. To some extent the ability to display wealth in Roman terms was also affected by the nature of the water supply, so houses such as the palace of the Dux Ripae, which one would expect to have several decorative features, actually made restrained use of water because it must have been supplied by cisterns. An interesting comparison can be drawn here with the acceptance of urban and public water installations. When private installations that are associated with a Roman way of life (lead pipes, bathhouses, latrines and fountains) are added to the picture for urban systems, further subtleties of the relationship between identity and material culture can be drawn out. These features were restricted to centres with the strongest associations with Rome: Antioch, Beirut, Zeugma and to a lesser extent Apamea (Tables 5 and 6). This must have reflected which groups of people wished to display their wealth in Roman terms, and by so doing display their desire to seem Roman and their willingness to embrace different ways of living. It seems that the crux is a willingness to make a

CHAPTER ONE

59

change in behaviour that will impact on expressions of identity. It is striking, for example, that Caesarea, which displayed a strong Roman identity in the public world, did not do so in the private realm. This would seem to indicate that people were less eager to change their private identities than their public ones. Antioch Lead pipes (Public) Lead pipes (Private) Nymphaeum Fountain (Private) Bathhouse (Public) Bathhouse (Private) Latrine (Public) Latrine (Private)

Apamea

Beirut

X Bath

X

X

X

Caesarea Dura Eur. X

X X

X X

X X

X

X

X

X

X

X X

X X

X

X X

X

X

Table 5: Presence (X) of public and private water supply features associated with a Roman way of life in selected cities across the East: Antioch, Apamea, Beirut, Caesarea, Dura Europos.

X X

60 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST

Lead pipes (Public) Lead pipes (Private) Nymphaeum Fountain (Private) Bathhouse (Public) Bathhouse (Private) Latrine (Public) Latrine (Private)

Jerash

Petra

Umm Qes X Bath

X

X

X

Zeugma

X X

X

X

X

X X

X X

X

Table 6: Presence (X) of public and private water supply features associated with a Roman way of life in selected cities across the East: Jerash, Petra, Umm Qes, Zeugma.

WATERMILLS Studies of watermills and water-powered installations have suffered in the past from the traditional view of technological stagnation in the ancient world. This view stemmed largely from Finley’s declaration that there was little technical innovation or economic progress in the ancient world (Finley 1965).14 A retarded use of water-power was felt to be a consequence of slave labour and lack of interest in investment in new technologies. Therefore, the rise of water-power was deemed to be a progression of the early medieval period. The idea that mills were not in widespread use during the Roman period, as well as the absence of references to mills in the 14 See Wilson 2002b for a detailed review of the history of ancient technology studies.

CHAPTER ONE

61

literary sources, meant that few scholars dedicated time to their study. This situation began to be rectified by Wikander (1979; 1981; 1984; 1985; 1990) whose on-going work on ancient watermills has opened up the field to study. As well as indicating the archaeological evidence for watermills, he also pointed out that the appearance of documentary genres, such as monastic charters and hagiographies, accounts for the apparent increase in written evidence in the early medieval period. Lewis (1997) has also illustrated the potential of Roman technology and water-power, its role in industrial advancement and the diversity of its applications from milling grain to crushing ore. The material presented here adds more weight to the idea that the Finley’s model of technological stagnation in the Roman world must now be considered untenable and a reflection of the state of evidence at the time he was writing (though Barbegal was known). There are two main types of watermill that are distinguished by the orientation of the wheel: vertical or horizontal (Wikander 2000b:373-8). The vertical wheel requires a right-angle gear and turns a horizontal wheel shaft on the other end of which is a vertical cogwheel (Figure 11). This cogwheel drives a horizontal cogwheel and the millstones above. Variants of the vertical wheel include the undershot wheel, which is partially immersed in the water and driven by running water, the overshot wheel, which is powered by water conducted through a chute above the wheel and the breastshot wheel, which is powered by water hitting the back of the wheel. The horizontal wheel does not require a gearing system and is driven by water conducted either by a steep chute onto oblique paddles on the wheel or from a nozzle at the base of an arubah penstock (which functions like a tube delivering water to the wheel), or exceptionally, as a turbine (Figure 12).

62 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST

Figure 11: Schematic drawing of a vertical-wheeled mill with a right-angled gear.

CHAPTER ONE

63

Figure 12: Schematic drawing of a horizontal-wheeled mill with an arubah penstock.

The majority of the 25 known watermills from the East seem to be of the horizontal type with an arubah penstock. The arubah penstock is usually a 6-10 m high stone tower that contains a column of water. This column of water is let out through a narrow orifice at the bottom of the tower, so the wheel is driven by a jet of water under pressure (Wikander 2000b:376). This is a particularly efficient design where stream flow rates are limited, as in large parts of this region, because the power developed is proportional to the head of pressure in the drop-tower, which makes efficient use of smaller volumes of water (Wilson 1995b:503; Braemer 2009:42). With the exception of the mills at Amida and Antioch that are dated by literary evidence and the mill at Wadi Faynan, some of the other mills in the region potentially may be later than our period.15 In general, the dating of most of the mills included in this study is tentative as it has not been confirmed by fieldwork. A case in point is the mill complex at Lejjun, which was thought for several years to date to the Roman period, but excavation subsequently revealed that it dates back only as far as the 19th-century AD Ottoman use Amida mills: AD359 (Ammianus Marcellines 18.8.11); Antioch mills: AD73-74: inscription relating to possible fulling mill (Lewis 1997;96-99, Feissel 1985); Antioch mills: 4th century AD (Libanius Orat. 4.29). 15

64 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST of the area (McQuitty 1995:746). In most cases the power from the water wheel in these installations was used to run grain mills, but there are two, possibly three, notable exceptions from Wadi Faynan, Jerash and Antioch; these special cases are discussed in Chapter 2.

INFLUENCING TECHNOLOGICAL CHANGE AND CONTINUITY IN THE NEAR EAST Throughout this chapter it has become clear that there were several agents or agencies of change in the Near East coming from both within and from outside the region, including Herod, the army and of course, the Emperor. In addition to these agents and the move towards change or additional elements, however, there also seems to have been a strong desire amongst the inhabitants for continuing technological practices. This section will explore the evidence of who brought about change in the Near East, looking in particular at the literary and epigraphic evidence for who paid for the construction and maintenance of major elements of the water supply and management infrastructure. Then, there will be some discussion of the reasons behind the decisions to change or not to change water supply and management technologies, which acts as a short introduction to a recurrent theme throughout the book. The initial outlay costs and cost of subsequent maintenance of water management facilities could be exorbitant. It is the generally accepted opinion that the city collectively would manage to pool resources by either the donations of rich citizens or tax levies since the cost was usually beyond that of a single act of euergetism (Eck 1987:74-9; Leveau 1991:153-4; Leveau 1992 :233-5; Wilson 1996 :18). The cost of building aqueducts was immense; C Iulius Secundus, for example, donated 2 million sesterces for the construction of the Bordeaux aqueduct (Leveau 1991:153; CIL XIII, 596). For this reason private aqueduct construction was rare, except in North Africa where c. 30% of the aqueducts whose funding source we know were privately funded (Wilson 1997:146-9). Alternatively, imperial benefactions were granted to cover the construction costs. It is also proposed that on these occasions imperial architects and foremen were used, as is suggested by the use of building techniques such as opus reticulatum outside Rome and Italy (Leveau 1992:234-5; Wilson 1996:18-9). In consequence of these imperial associations, it is commonly thought that aqueducts ‘constitute

CHAPTER ONE

65

physical manifestations of Roman imperialism’ (Hitchner 1995:157. Also see Leveau 1987). The subject of maintenance and upkeep has received less attention, but is believed in the main to have been undertaken by personnel organised at a municipal level (Leveau 1992:236). There is a large amount of evidence for imperial benefaction in constructing aqueducts in the Near East.16 The Acropolis aqueduct at Antioch was donated by Julius Caesar. Later, Antioch received aqueducts from three Emperors: the north and central aqueduct built by Caligula and the southern aqueduct built by Trajan and Hadrian as recounted by Malalas in his Chronicles (216.21-217.2, 243.10-21, 275.22-276.3, 277.20-278.19).17 In addition to these benefactions for the building of aqueducts, Justinian also donated money after the earthquake of AD 526 for the restoration of the aqueducts (Malalas 422.4-5). A fourth aqueduct at Antioch (mountain), probably built in the 2nd century BC, may have had Roman involvement too. A graffito of a name inscribed twice in the wall of this aqueduct seems to name Cossutius who may be the same Roman architect who, according to Vitruvius (VII, praef. 15. 17), supervised the work on the Temple of Zeus in the Olympieion at Athens on behalf of Antiochus Epiphanes (Downey 1938:160, #90). While the graffito may have been written by an Italian craftsman, architectural analysis from the Temple of Zeus suggests that Cossutius was working in the Greek architectural tradition, hence his links with the Seleucid monarchy (Anderson 1997:20-21). Seven inscriptions from the Caesarea High Level aqueduct channel B mention Hadrian and three others were worded to show that the work was carried out on imperial orders and are also evidence for army construction (Di Segni 2002:47). Pilate also ordered On account of the strength of Di Segni’s (2002) research, the literary and epigraphic evidence from Israel and the Occupied Territories for who paid for and built aqueducts in the Near East has received more attention than the rest of the study area (with the exception of Antioch) and will, therefore, have to form the bulk of the evidence for the region as a whole. It is, of course, always possible that practices differed across the provinces in question. 17 Libanius (Orat. 11.125) also makes mention of unspecified rulers conducting water to the city. 16

66 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST the building of an aqueduct at Jerusalem; this raised a disturbance among the Jewish population because he paid for it using ‘Corban’, i.e. sacred temple offerings (Josephus BJ 2.9.4). While the aqueduct was paid for out of expropriated funds, this may have been done under the orders of the Emperor. The use of legions in the construction of the aqueducts may also point to indirect imperial involvement and contributions, if not in actual monetary terms, at least in provision of manpower. Three aqueducts bore inscriptions that show direct legionary involvement: Jerusalem High Level, Caesarea High Level channel B and Beth Govrin (north). Thirty-one inscriptions naming the century of particular centurions have been found on the inverted siphon section of the Jerusalem High Level aqueduct (Vetrali 1967; Di Segni 2002:41). The legion was probably the X Fretensis that was stationed in Jerusalem from Titus’ reign until the late 3rd century AD. Eleven inscriptions (two fragmentary) from the Caesarea High Level aqueduct channel B, indicate that vexillationes of no less than four legions (Legiones X Fretensis, VI Ferrata, II Traiana Fortis and XXII Deiotariana) were involved in its construction (Di Segni 2002:47): Imp(erator) Caesar/ Traianus/ Hadrianus/ Aug(ustus) fecit/ per vexillatione(m) leg(ionis) X Fr(e)te(nsis). ‘The Emperor Trajan Hadrian Augustus made this through a vexillatio of the legio X Fretensis.’18

A centurial inscription from Beth Govrin (north), probably no later than the early 3rd century AD, also points to the work of Roman soldiers (Legio VI Ferrata) on this aqueduct (Di Segni 2002:51). Military units occupied all these sites and Di Segni (2002:52) has pointed out that almost every place where military units had an attested presence also had an aqueduct. What is the connection between military sites and aqueduct building? Is the provision of aqueducts to military sites one of pure need or is it a demonstration

18 Negev (1964:248) believes that these inscriptions commemorate the restoration of the aqueduct rather than its construction, but Di Segni (2002:47) argues that fecit points to the construction.

CHAPTER ONE

67

of strength, power and skill? Were the legionaries principally providing for their own needs or helping urban development? There is some, albeit limited, evidence to answer these questions. The imperial orders instructing the construction of the High Level aqueduct channel B at Caesarea may point to an imperial concern for general urban development. At Maximianopolis, however, the aqueduct conveyed water to the northern area of the town, which is believed to have been the location of the soldiers’ camp, so it could be argued that the military were primarily concerned with their own supply (Di Segni 2002:52). This seemingly self-interested action can possibly be tempered with the idea that the wells and cisterns, which probably provided the main bulk of water for these sites prior to the arrival of the military, would not be able to cope with the sudden influx of people and therefore a greater supply needed to be brought in from elsewhere. The later introduction of an aqueduct by the military to a town has also been noted in France (Février 1983). In addition, the close presence of the military may have provided the security and protection that these costly installations would need. The army, of course, also provided valuable manpower and expertise in aqueduct building. In Numidia, for example, the army was employed in projects for military and veteran towns from the 2nd century AD onwards because they were becoming increasingly idle and there would be an eventual profit from their activities (Fentress 1979:164). It is also possible that the military were involved only in the hardest tasks involving complex technology. The sections with attested military involvement include the Caesarea High Level channel B arcade and the Jerusalem inverted siphon, both of which represent sophisticated technologies. This is similar to the dam-tunnel complex at Seleucia Pieria, where the engineering challenges appear to have been met by the military as well. An army engineer, Nonius Datus, was also involved in a tunnelling project in Algeria; the project was almost a failure as the two crews, also military units, tunnelled past each other from either end, but were corrected by Nonius Datus (CIL VIII.2728. Hodge 1992:13, 128). This example shows that it was the use of military engineers, maybe, rather than military manpower that was the important factor in complex construction. Thirteen aqueducts are attributed to the Herodian period, with four others possibly being Herodian. Two of these aqueducts fed

68 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST Roman camps, which highlights the close relationship between Herod and the army. As mentioned earlier, Herod appears to have had a key role in the introduction, transmission and diffusion of Roman technology in the East, for example using latrines in his private palaces and commissioning bridges and arcades to be constructed on his palace aqueducts. Herod was an effective link between Roman and local and so, may be described and viewed as an influential agent in ‘middle ground’ behaviour in the Roman East. There is limited evidence for municipal involvement in aqueduct construction. The present opinion among Israeli scholars is that it did not take place (Di Segni 2002:54; Hirschfeld 2002b:3878). In Syria, however, the aqueducts to Kanata and Suweida were both dedicated to Trajan, under the governor Aulus Cornelius Palma, who was governor of the province in AD 104/5-107/8 (Di Segni 2002:n. 95, 101). These epigraphic formulae were typical of municipal construction elsewhere in the Empire: dedication to the Emperor and mention of a governor who has gained permission to use civic funds. There is only one attested example from the East for private benefaction: an inscription from a bathhouse in Apamea lists the many and substantial benefactions of C. Iulius Agrippa, including the aqueduct (AE 1976, no. 678; Rey-Coquais 1973:41-46; Leveau 1991:154). While the paucity of evidence for private aqueduct benefaction may be explained partly by the low epigraphic habit in the East, it may also indicate that North Africa was exceptional in this regard. The cost and responsibility of aqueducts did not, of course, stop at construction, but extended into the upkeep and maintenance of these structures. Surprisingly, however, this subject has not received the attention it deserves; for example, it is all but excluded from Hodge’s (1992) seminal work on aqueducts. There were three main areas that needed attention: protection of the aqueducts from vegetation and cultivation; ongoing routine maintenance and major structural repairs. In the case of protecting the aqueducts from vegetation and cultivation, the Theodosian Code (15.2.1) records a ruling from Constantine to Maximilianus, Consular Administrator of the Water Supply that landholders through whose lands aqueducts run must clean the aqueducts when they become choked. Personal responsibility for the protection of the aqueducts is also seen in a post-

CHAPTER ONE

69

Justinianic imperial edict found near Bethlehem (Di Segni 2002:589). The edict forbids anyone from planting or sowing within 15 feet from each side of the aqueduct on pain of death and confiscation of property. Similar edicts are known throughout the empire from the Augustan period onwards, though none was as heavy-handed with the punishment as this example. Ongoing routine maintenance seems to have been in the hands of city officials. According to Libanius (XLVI.21; XXV. 43), plebeian manual labourers overseen by curiales usually undertook the general daily upkeep of the aqueducts at Antioch. In Sepphoris in AD522 an inscription records that a new water installation (νέον ὑδρίον) was carried out in the time of Flavius Orestes, and under the supervision of Silvinus, son of Marinus, comes and protos. During the excavations at Scythopolis, replacement ceramic pipelines alongside older pipelines have been found that attest to this renovation archaeologically as well (Tsafrir and Foerster 1997:105-6). Choricius of Gaza (Laudatio Aratii et Stephani pars. 44-49) tells us that the governor, Stephen, consul of Palestine, improved the Caesarea High Level aqueduct in AD 530–535, removing the old hindrances (τὰ πάλαι κωλύματα) (Di Segni 2002 :61). It seems that the municipal authorities and Stephen’s predecessors had failed to discharge their duties of maintaining the aqueduct and supervising the maintenance respectively. Choricius also suggests that skilled workers carried out the repairs, which may mean that some of the work needed specialist care, for example, replastering the specus. In this case, it would seem that as the city authorities had not cared for the aqueducts effectively, the governor had to be called upon; this is more akin to major structural repairs. Major structural repairs were often caused by several, dramatic, events: earthquakes, seasonal flooding/runoff and subsidence. In some cases this may have resulted in the complete severing of the aqueduct line, such as at Misyaf where a section of the aqueduct was moved several metres to the west by an earthquake. In other cases severe leaks may have occurred, as evidenced by the limescale incrustation on the outside of one of the aqueduct bridges in Antioch (Figure 13). Under these circumstances, the aqueducts were repaired by governors and with imperial aid. Evidence from Bosana in Syria suggests that a spring-house was renovated in AD 365 by a syndikos, who seemed to be responsible for defending the interests of their town in front of the Emperor or

70 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST provincial governor (Le Bas and Waddington 1853:#2239). It is possible that this repair may have been necessary due to a big earthquake that year. As noted above, Justinian donated money after the earthquake of AD 526 for the restoration of the Antioch aqueducts (Malalas 422.4-5). The aqueduct at Bosra, known only from epigraphic evidence, as well as several other public buildings, were renovated under Justinian using money from the provincial funds (παρὰ τῶν δημοτικῶν), after being damaged by the earthquake of AD 526 (IGLS 13.1, 9128-37 ; Di Segni 2002:64; Dumond-Maridat 2008:80). In addition, an inscription from the Caesarea High Level aqueduct records that in AD 385 Flavius Florentius, proconsul of Palestine, renovated both High Level aqueducts from the foundations (Hamburger 1959:189-90; Di Segni 2002:61). The inscription was found in the swampy Kabara area where it would seem the aqueduct had been damaged by being on unstable ground.

Figure 13: View of the limestone incrustation on an aqueduct bridge in Antioch.

So, it would appear that several different bodies undertook the costs and responsibility of aqueduct construction: Emperor, client king, municipal authorities and the legions. The evidence from An-

CHAPTER ONE

71

tioch seems to point to a high degree of Roman imperial involvement in aqueduct building there, but it was no ordinary Roman town. As provincial capital, it is more than likely that Antioch had more Romans amongst its population and had a special status that may have made it subject to more imperial favour than other towns and cities. Antioch was exceptional in the number of imperial benefactions that it received and in the extent to which they were documented. Trajan and, to a lesser extent, Hadrian, seem to have been the most prolific aqueduct builders, which is reflected in the peak of aqueduct construction in the Near East in the 2nd century AD. Herod was also an extraordinary figure in the field of aqueduct construction, being responsible for an early rise in the numbers of aqueducts built in the region, as well as being a catalyst for the use and dispersal of newer technologies. As far as maintenance was concerned, it was the responsibility of the landowner through whose lands the aqueduct traversed to protect and maintain that section of aqueduct.19 Day-to-day upkeep was in the hands of city officials. Larger structural repairs were the responsibility of city officials and provincial governors. Where the structural repairs were caused by a natural disaster, direct imperial help would need to be called upon. In comparison to the amount of data known about the construction and maintenance of the aqueducts, less epigraphic or literary evidence is available about other parts of the urban management system. The tax law (AD 137) of Palmyra tells us the caravan traders were charged 800 denarii per year for the use of the springs outside the city (Matthews 1984:177). At Jerash the ‘devotees’ dedicated a stoa and reservoir (τὸν λάκκον) in the Temple of Artemis (Welles 1938:389, #28). With these exceptions (and the nymphaeum and fountain donations), the evidence relates to the late Roman period. It may not be viable to extrapolate this situation back because the situation in the late Roman period seems to have been quite distinct.

19 Leveau (1991:154) notes the responsibility of local users for aqueduct maintenance, but does not consider how expenses were covered in larger renovation schemes.

72 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST The majority of the evidence for imperial benefactions comes from nymphaea and fountains. We know of a nymphaeum at Suweida dedicated to the city in honour of Trajan, under a governor (either Julius Quadratus or more likely his predecessor Cornelius Palma) (Le Bas and Waddington 1853:#2305 and #2308; Glaser 2000:445; Di Segni 2002:54; IGR III:1273, 1276). The statue of Trajan was placed in a very prominent position: directly above the water outlet in a single storey nymphaeum façade (Glaser 2000:445). The nymphaeum at Jerash was dedicated to Commodus in AD 191 (or AD 190) (Welles 1938:406-7, #69). The link between nymphaeum construction and Emperors is discussed further in Chapter 3. There is also evidence for governors being involved in the donation of water supply and management facilities. An inscription from a fountain in Jerash records that Attidius, the consul designatus, set up fountains in AD 150. In addition, Herod wished to prove his generosity by giving costly fountains (κρῆναι πολεθτελεῖς) to Ashqelon (Josephus BJ 1.21.11). An inscription from an agraria statio in Provincia Arabia also records military involvement in reservoir building: in AD 334 an officer who had seen many of his men killed in an ambush by Saracens while collecting water built a reservoir (Iliffe 1944; Di Segni 2002:52). Evidence from the Hauran villages, in contrast, points to the involvement of villagers in paying for their water infrastructure (Dumond-Maridat 2008:83-4; Braemer 2009:50). There is more information about the upkeep and maintenance of the urban distribution and drainage network. Shopkeepers in Antioch were responsible for cleaning the drains, which Libanius (Orat. xlvi.21) suggests was dangerous work in which a man might be choked to death (Liebeschuetz 1972:219). Work on replacement water pipes in Scythopolis in AD 521 was attributed to the principalis Silvinus son of Marinus and to the governor Flavius Orestes in two inscriptions set in the street pavement (Di Segni 2002:62). Four 6th-century ostraca from Sbeiteh/Shivta point to villagers contributing a period of labour in the public reservoirs. The ostraca show that the reservoirs were cleaned in October, which is appropriate as the rains resume at this time and the reservoirs would have had their lowest water levels. The involvement of the Church is also seen. An inscription from Jerash recorded that a bishop corrected the nuisance caused by bad odours from the sewer (Fisher 1930:9). We also see Church

CHAPTER ONE

73

involvement in the renewal and upkeep of the water supply at Bosra (Dumond-Maridat 2008:80). Euthymius restored two large reservoirs in the vicinity of his monastery (Di Segni 2002, 52 fn. 91; Cyril of Scythopolis Life of Euthymius 51). This is different from the aqueduct examples, where no involvement from the Church was seen. Different again from the aqueducts is that no renovation or maintenance seems to have been carried out under imperial orders, which is to be expected except under special circumstances such as repairing damage after natural disasters. Instead, maintenance seems to have been undertaken under the supervision of municipal authorities using a lower class workforce or under the supervision of the Church.20 So, in this review of the evidence of who paid for the construction and maintenance of the water supply and management infrastructure, we find that there are several individuals and bodies responsible for different elements of the system: individuals such as the Emperor, Herod and governors and bodies, such as the army and its engineers, who are under imperial control, municipal officials and later in the Roman period, the Church. Amongst these individuals and bodies, there is a strong emphasis on the involvement of the agents of empire. This may suggest that they had a heavy involvement and influence in technological decisions. Certainly, this seems to be the case in large infrastructure projects, such as aqueducts and dams. The extent to which these agents of empire had an effective influence on choices of other elements of water technology is, however, less clear and convincing. Herod, for example, seems to have been very influential in introducing aqueduct arcades and bridges to the Near East, but his use of latrines does not seem to have spread with the same enthusiasm. As discussed in Chapter 4, the reasons behind deciding whether or not to use latrines are complex and multifaceted and strongly linked to attitudes and emotions that were embedded in the fabric of Near Eastern culture and society. It is this that seems to lie at the heart of technological choice in the Near East, rather than a sense of choosing what might be deemed technologically ‘superior’. This 20 Di Segni (2002:52) provides evidence about military involvement in the construction of rural reservoirs.

74 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST theme will be developed throughout this book, providing insights into why bathhouse popularity reached its peak a century later in the Near East than in other parts of the Empire, why private houses did not use water for display and why some cities chose to have nymphaea, while others did not.

2 WATER AND THE ECONOMY This chapter will review the evidence for water-related economic activity in the Roman and late Roman Near East. The first section analyses the ways in which water was used in industry and production. The second part of the chapter discusses some of the wider debates into which the evidence from the Near East fits, drawing on the evidence presented in the first section and in Chapter 1. This part is divided into two main debates. Firstly, there will be discussion of the contribution that water supply and management studies in the Near East can add to the debates over whether cities were consumers or producers in the Roman world. It will be argued that neither of these labels is necessarily helpful and that there was a more symbiotic relationship between cities and their hinterlands. This will be followed by looking at the changing economy in the late Roman period and consideration will be given to the nature of urbanization in the late Roman period.

USES OF WATER IN INDUSTRY AND PRODUCTION Water was integral to many industrial processes in the ancient world. This section will look at the processes of milling, tanning, dyeing, fulling and fish breeding, for which the archaeological evidence is best. It will assess what the water was used for during these processes and how the water was supplied. Watermills As mentioned in Chapter 1, most watermills in the Near East were used to run grain mills, but there are two, possibly three, notable exceptions from Wadi Faynan, Jerash and Antioch. These three examples point to mills being used for crushing ore, sawing stone and fulling respectively. At the copper-mining site of Wadi Faynan, the water leaving the mill was not used for irrigation in the surrounding field sys75

76 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST tems, as would be expected in an arid region (see Figure 8) . It is possible that the reason the water was not re-used was that it had been contaminated during an ore-crushing process (Barker et al. 1999:277; Barker et al. 2007:162-4, 317-8). Other examples of orecrushing mills are known from elsewhere in the Empire from gold and silver mines in the Iberian Peninsula and Wales (Lewis 1997:106-110; Wilson 2002b:21-23). Although this interpretation is tentative, if the mill at Wadi Faynan was used for ore-crushing, using rotary crushers, rather than trip-hammers, this would be the first known example of water-power used for this process in the East. Evidence for the existence of sawmills in late antiquity was based for a long time around a passage from Ausonius (Mosella: 359-364) in AD 367 that clearly describes sawmills on the Moselle River (White 1962; Rosumek 1982; Simms 1983; Simms 1985; Wikander 1989; Seigne 2002). A passage from Gregory of Nyssa (In Ecclesiastem III, 656A: referring to Cappadocia) may also allude to water-powered sawmills in the 4th century AD (Wikander 1989:190). There is also a 3rd century relief of a water-powered sawmill from Hierapolis, Phrygia (Ritti et al. 2007). The installation at Jerash now proves archaeologically the existence of such mills in the late Roman period. This sawmill, which was inserted into the pre-existing Temple of Artemis complex, must date between the 5th-century abandonment of the Temple of Artemis and the great earthquake of AD 749; a Justinianic date is likely as that was a period of great prosperity at Jerash (Seigne 2002:212). A 7th-century sawmill may also have been identified at Ephesos, in association with ten overshot watermills, which points to a large-scale operation (Vetters 1984:225; Wikander 1989:190; Schiøler 2005). The Jerash mill was used to saw up column drums for use as veneer. Therefore, not only is this mill an example of industrial complexes encroaching on public space in late Roman Jerash, but also it was actively involved in the spoliation and destruction of the Classical architecture of the city. Seigne (2002) reconstructs the workings of the sawmill as follows. A small cistern, acting as a header tank, in the courtyard of the Temple of Artemis fed the mill, which was located 4 m below in a subterranean chamber. The outlet channel probably exited the building and joined the main street sewer as it would have been contaminated by stone dust. Carved rectangular sockets for shaft

CHAPTER TWO

77

bearings were found in the wheel race, one of which showed signs of circular wear traces. These patterns point to an overshot waterwheel of at least 1 m diameter mounted on a short horizontal shaft with a vertical wheel at each end of the shaft. The most important finds for the identification of this mill’s function were two partlycut column drums which suggest that there must have been four saw blades working simultaneously on each side of the wheel. The saw blades are likely to have used abrasive material and water rather than teeth, to have been 2.2 m long minimum and set in a rigid frame for tension. The saws would have cut vertically down through the stone with blades moving back and forth horizontally to take advantage of their weight. A system with cranks or with eccentric and connecting rods would have been necessary to ensure the transformation of a continuous circular motion (from the millwheel) into a longitudinal reciprocating motion (for cutting the stones with the saw). It is likely that each of the wooden wheels on the wheel shaft would have carried an eccentric pin that was linked to the end of the saw frame via long connecting rods. The sophistication of this installation suggests that this was not a prototype, thus making it probable that other, earlier examples did exist and adding weight to the idea that the mills described by Ausonius in AD 367 were indeed sawmills. Archaeologically, however, this is probably the earliest attested example so far of this kind of technology. Lewis (1997:96-99) has argued that the 1st-century AD ‘fullers’ canal’ at Antioch, known only from epigraphy, must have driven fulling mills. Fulling mills and stocks can be used as part of the beating process of fulling. Traditionally this beating process was done by hand, but at some point it became mechanized; a triphammer, cam and (normally) a waterwheel were harnessed together to create fulling stocks and fulling mills (Lewis 1997:92). The use of a fulling mill means that less manpower would have been needed and potentially more fulling performed. This is often felt to be a medieval progression, but Lewis believes this inscription points to an earlier use of waterpower in fulling. Although one needs a good water supply for fulling, the estimated 300,000 m3 of water at a velocity of almost a metre per second supplied by the Antioch channel would be far in excess of what was needed for regular fulleries, especially when compared to the aqueduct at Pompeii that supplied significantly less (2,500 m3), yet fed 14 fulleries as well as the public

78 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST fountains. It seems possible then that the channel was used to provide waterpower as well as supply water. As the river could turn norias at this location, the channel could turn other undershot water wheels. The channel’s size and the amount of power it could generate means that, potentially, it could have run 42 pairs of stocks. This makes it possible that there was fulling on an industrial scale at Antioch as early as the 1st century AD. Although there are no certain examples of fullers’ mills until the 11th century AD, this is a tempting, but circumstantial, argument (Wikander 2000b:406). The possibility still exists that the water in the canal was used for rinsingon a layer scale than at Pompeii, rather than as a source of power (Miko Flohr pers. comm.). Tanning, dyeing and fulling Two installations were associated with dyeing textiles at Beirut (Bey 008: Curvers and Stuart 1996:229) and Jerash (officina tinctoria in the macellum: Uscatescu and Martin-Bueno 1997:78). A further four may have been fulleries in Antioch (Lewis 1997:96-99; Feissel 1985), Beirut (Bey 006, Area 2: Perring et al. 1996:195-6; Perring 1997-1998:22), Jerash (hippodrome: Glueck 1934:6) and Scythopolis (Foerster and Tsafrir 1992:7). Ten sites have installations that are of uncertain function, but might have been connected with tanning, dyeing or fulling: Dor, Caesarea (2: Raban 1995:298-300; Flinder 1976:77-79; Angert 1994:138-139), Gaza (Ovadieh 1969:197), Sarafand (Pritchard 1971:47-48), Tel Yizre’el (Yogev 1988-9:193), Khirbet Summaqa (Dar 1985:106), Khirbet Ni’ana (Bashkin 1993:59-61), Auara (Oleson et al. 1999:42), Zeugma (2: Abadie-Reynal et al. 1997:355) and Ain Feshka (De Vaux 1960:8; Poole and Reed 1961:114-122). All of these installations are characterised by a series of vats, sometimes connected by channels that occasionally were controlled by a complex system of sluice gates. Textiles were dyed by macerating plants or shellfish in a solution of water that was often heated (Wilson 2000b:144). The purple dye was extracted from the hypobronchial gland of the Murex sea snail (Herzog 1987:22; Wilson 2000b:144). Pliny (NH IX 38, 133-5) records that this was undertaken by heating a mixture of shellfish and salt in water with air piped from a furnace. While the colour develops a garlic smell is exuded, so it is no surprise to hear Strabo (Geography XVI.2.23.575) complaining of the bad smell at Tyre from its dye factories. Tyre was famed in antiquity for its dyeworks,

CHAPTER TWO

79

which came under imperial control under Diocletian (Herzog 1987:19-20; Karmon and Spanier 1987:149). By the late 4th century AD restrictions imposed by the Emperors Valentinian, Theodosius and Arcadius had made commercial production and dyeing of both wool and silk in mollusc purple a capital offence; all such activity had to be carried out in Imperial dyeworks (Bridgeman 1987:160). Circular rock-cut vats, some in clusters that were connected by narrow rock-cut channels, and piles of murex shells were observed at Tyre in the 19th century (Wilde 1840:378-380 - Appendix M, 629-644). The Roman installation at Beirut (Bey 008) comprised a basin that was associated with a large quantity of Murex shells (Curvers and Stuart 1996:229). This seems to have been small-scale local textile production, which reflects the earlier, Roman date of the installation. The so-called officina tinctoria in tabernae 10-12 and exedrae 2-3 of the macellum at Jerash was a larger undertaking and dated to the 6th-early 7th centuries AD (Uscatescu and MartinBueno 1997:77-8). This installation had a furnace, which is the only known example from the eastern empire. Of the six vats, four were rectangular and plaster-lined and had amphorae at the bottom, which were probably used to bring water from a small cistern in taberna 10; these were probably used to immerse the cloths. Two of the vats were semi-circular and made from re-used architectural fragments; these vats were not deep enough to hold much liquid and may have been used for the spreading out of dyed cloths for enhancement with fuller’s earth or dried dyes as described by Pliny (NH IX 35, 198). Two water tanks were also found that were connected via pipes to the central fountain in the macellum. The dyeworks thus involved the renewed use of the water system in this area as well as adaptations to the urban scenery for industrial purposes, e.g. the earlier Roman pavement was broken in order to install the water tanks. An alternative use for this installation may have been fulling, perhaps in combination with dyeing. Fulling, which cleans clothes and prepares textiles for dyeing, involves the removal of animal fats and grease by the trampling of the cloth in tubs that contain an alkaline solution of water and fuller’s earth, dilute urine or other alkaline substance. The archaeological remains of these installations usually consist of small tubs for trampling the cloth and large vats for rinsing them afterwards (Wilson 2000d:143). The late Roman complex from the hippodrome area in Jerash was described as comprising three vats and

80 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST four plastered jars sunk partially below floor level opposite the vats, which is similar to the fulleries at Ostia (Glueck 1934:6). No detailed information is available about the late Roman fulleries from Beirut and Scythopolis, but both were supplied by piped water. The fulleries at Antioch, discussed above, also had their own supply of water. It seems significant that three of the known examples of this type of installation did not have to rely on wells or cisterns, but were deemed worthy of their own specific supply of water, which suggests that there had been considerable investment in these ventures. It would appear, then, that these installations may be compared with the large-scale ventures found at Rome, Ostia and Pompeii. The ten other sites in the Near East of unclear function consisted of complexes of basins or vats with channels, some of which had sluices. The two installations from Zeugma were found in caves on the bank of the Euphrates and seem to have been fed by water percolating through the cave walls as calcite formations were found on the walls of a rock-cut niche above a basin (AbadieReynal and Ergec 1997:355). A system of channels of varying dimensions criss-crossed the floor of the caves; all of them angled down to the mouth of the caves and the Euphrates and some of them had an ashy charcoal-rich fill; the provision of small basins and large vats suggests that they may have been used for fulling. A complex of a vat, basins, pits and channels at Ain Feshka is commonly thought to be associated with tanning or vellum preparation because of the site’s proximity to Qumran where such activities could not be carried out due to religious strictures. Also Ain Feshka had a permanent and abundant supply of water (Harding 1967:200; De Vaux 1959; De Vaux 1960). The low phosphate content and lack of organic traces (hair, plant remains and diatomaceous bodies) in residues from this complex show, however, that animal skins were not processed there (Poole and Reed 1961:122). Tests for tannins were also inconclusive and flax-retting is also unlikely as there were no plant remains. It is possible that these may have been fishponds as the water was not static, but no recesses or jars were found. Fishponds (vivaria) Jars set into the walls, probably the cells described by Columella (De Re Rustica 8.1.3, 17.1-6) as essential in a fish-breeding pond, are

CHAPTER TWO

81

diagnostic for fishponds. Such jars, set horizontally into the walls, were found at Khirbet Sabiya (300 jars: Ayalon 1979:175-177, 179), Caesarea (6th century: 60 jars (Gibson 1991:41) and Tel Tanninim: 16 Gaza amphorae (Stieglitz 1998:63-5)) and Sataf (Mango 2002:325). A similar arrangement was found in the south-eastern reservoir at Andarin, but instead of jars or amphorae rectangular recesses were found at the base of the reservoir walls, possibly numbering 200 (Mango 2002). The example from Andarin was exceptionally large measuring 61 m long x 61 m wide x 3 m deep; this is probably because it also acted as a reservoir for irrigation supplies. As it was fed by freshwater it is possible that catfish were bred there. Catfish were the only freshwater fish bred in fishponds cited in ancient sources, such as Apicius and Pliny (André 1981:109-113); in addition, catfish bones have been identified in the assemblage retrieved from the cistern Andarin bathhouse, though not from the reservoir itself (Mango 2009:75). If these installations were indeed vivaria, it is possible that they were used by the nobility as attested by Varro (Res Rusticae 3.3.5-10; 3.17.2-9). With the exceptions of Khirbet Sabiya, Andarin and Sataf, all of these sites were coastal. The examples from Dor seem to have been fed by seawater (Raban 1995:343). A pipeline branch (Channel E) from the Caesarea High Level aqueduct channel A fed the reservoir for the piscina at Tel Tanninim; the pipeline ran up a ramp attached to the southern wall of the reservoir (Stieglitz 1998:57-8). The reservoir at Andarin was fed by qanat water, which was subsequently channelled out into the fields for irrigation (Mango 2002). This multi-functional use of the qanat water is unsurprising given Andarin’s inland, pre-desert position. Spring water from a spring flow tunnel at Ein Bikura, Sataf fed a pool with two rows of ceramic jars in its sides with mouths towards the pool (Gibson 1991:41). In all of these cases the nature of the water supply means that a constant supply of circulating water was ensured (see Columella De Re Rustica 8.1.3, 17.1-6). The water could also be reused for other activities, for example, irrigation at Andarin and in a possible bathhouse at Tel Tanninim. The installation at Khirbet Sabiya, however, had no channelled or piped water supply and could only have been fed by wells identified in the vicinity of the site (Ayalon 1979:179). This would suggest that the water was not changed frequently on this inland site, which may be related to the type of fish contained within the pond. Palladius (Opus agriculturae, 1.17.2) ad-

82 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST vocates putting eels and river fish, for example, into cisterns in order to create a current. This does seem to be at odds with Columella's advice, but may offer a possible explanation for the curious situation at Khirbet Sabiya. The paired rock-cut tanks from Dor (Raban 1995:343) and Beirut (Thorpe 1998-9a:36-38) had no such recesses or jars. It is possible that these pools were therefore not used for fish breeding, which may also explain why these tanks were found in pairs unlike the other installations. It has been suggested that the Beirut tanks, which were lined with opus signinum, might have been holding tanks for holding the catch after the return of the fishing vessels. This is because the vats were in close proximity to a cove to which they were linked by a flight of steps. A similar function is also possible for the Dor installations since they too were situated very close to the coastline. Alternatively, it is possible that these were vats for the production of salsamenta where the flesh of the fish was cut up and salted. While there is evidence that fish were salted in the eastern Mediterranean, archaeological evidence has been lacking so far, prompting the idea that maybe the process was done in pithoi or dolia (Wilson 2006b; Curtis 1991:112-8, 129-147). Salting vats in the western Mediterranean (in particular Spain, Portugal and Morocco) were remarkably similar and almost universal in construction, though they varied in size and depth (Trakadas 2005:69-72). While the rectangular or square tanks in the western Mediterranean were usually built of brick or rubble, rock-cut examples are known from Portugal at Punta de l’Arenal and Praia de Angeiras. These Portuguese examples were not joined together as was usual elsewhere, which is similar to the possible salting vats from the East. The western examples were faced with opus signinum, as at Beirut. Large, round holes are sometimes found in the ground near the vats, which may have been for holding dolia; no examples of these have been recorded at Beirut or Dor. In addition, the western vats were usually arranged in rows along the inner walls of a room or building. No evidence for buildings was found at Dor and Beirut, but this may be due to large-scale truncation in the vicinity. Finally, in the west, the processing installations usually had a fresh-water supply from wells, cisterns or occasionally an aqueduct; again this has not been recorded in the East. So, the pairs of vats at Beirut and Dor illustrate some, but not all, of these common traits. It is possible, but

CHAPTER TWO

83

not definite, that these installations may be, then, the first fishsalting sites identified from the eastern Mediterranean. The evidence presented here, albeit limited, for industrial uses of water has shown that a wide variety of industries were taking place in the East, particularly involving fabrics and fish. It is clear from this that there is scope for more work on the remains of industrial installations in the East. The East stands out from other areas of the Empire in its high use of arubah penstocks for water-powered grain mills. This is largely due to the climate and limited water supplies of the region, which means that an arubah penstock that functions efficiently on low flow rates is ideally suited. The special uses of waterpower at Jerash and Antioch point to an awareness of how to maximise productivity, interestingly in an urban setting, which will be discussed further below. The connection of the dyeworks at Jerash and the fulleries at Beirut and Scythopolis to the municipal water supply also indicate that these industries were recognisably important and presumably profitable enough to merit the expense of connection. Furthermore, it is likely that if a constant water supply was needed, these installations must have been highly productive. With the exception of the Antioch fulling mill, these installations date to the late Roman period, which may point to a general upturn in industrial processing or an increase in scale making it more visible archaeologically, or maybe it moved into the cities in this period; this is discussed further below.

WATER AND THE ECONOMY: WIDER DEBATES One of the central tenets of Roman economy studies since Finley has been that cities were consumers, were parasitic on their hinterlands and had little or no productive capacity aimed at export. This has come under increasing scrutiny in the past decade, particularly by Wilson (1999; 2002c), Erdkamp (2001) and Mattingly and Salmon (2001). The role of aqueducts and industrial uses of water has a significant part to play in these debates. The focus here is firstly on the debates on the relationship between urban centres and rural areas, where evidence from branch lines off aqueducts in the Near East is put alongside that of North Africa to illustrate that more nuanced views about how cities and hinterlands interacted may need to replace those of Finley and his followers. In the second part of this section, attention is given to a slightly different part

84 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST of this debate, focusing on the increased productivity of rural areas and of cities as shown by the intensification of irrigation schemes and industries that made use of water either in their processes or for power. In particular, this section will demonstrate that there were significant changes in the late Roman economy in the East, which also makes a contribution to debates about the nature of urbanization in the late Roman period. The rural – urban divide Of particular interest for an economy of water in the Near East are the branch lines coming off urban aqueducts. Much previous research into aqueducts across the empire as a whole has been based on the dichotomy of productive rural aqueducts versus consumptive urban aqueducts (eg Shaw 1984; Leveau 1987:96, 98, 104; Corbier 1991; Shaw 1991; Ellis 1997; Hodge 2000b:47); a theory embodied by Ellis in his conclusion: ‘The aqueduct could thus in itself be a sign of domination, removing the productive capacity from the rural hinterland, and enslaving a rural spirit in an urban fountainhead.’ (Ellis 1997:149)

More recently some strong evidence has been presented suggesting that this theory needs to be adjusted and that there was a more symbiotic relationship between the rural and urban water supply (Wilson 1999:328-9; Gazenbeek 2000; Piras 2000:248). The significance of this topic goes beyond studies on water supply as the conclusions drawn here can play a key part in economic discussions on the relationships between urban centres and their hinterlands. The bulk of the water supply evidence so far has come from the western provinces and North Africa (Wilson 1999), but more can now be added from the Near Eastern provinces, including work in the Hauran that has demonstrated that the social organization of water supply and management was village-based (Braemer et al. 2009). A total (minimum) of 39 branch lines was found in the study on 18 aqueduct lines, 10 of which were urban lines [Caesarea High Level channels A and C and Low Level, Banias, Beth Govrin [north], Samaria [Shechem/Sebaste], Sepphoris – e-Reina, Tiberias, Beirut and Sbeiteh] (Porath 2002a:28; Porath 2002b:105-117; Conder and Kitchener 1882II:22-23; Siegelmann 2002:134-5; Pa-

CHAPTER TWO

85

trich and Amit 2002:17; Negev 1964; Peleg 1991a; Everman 1992:183; Porath and Yankelevitz 1989-1990:130-131; Hartal 2002; Conder and Kitchener 1883:110, 227; Hartal 1993:1-2; Hartal 1996:5-7; Sagiv et al. 2002; MacAlister 1901; Hedges et al. 1996:197; Frumkin 2002:270-276; Tsuk 2002a:281-284; Winogradov 2002; Conder and Kitchener 1881:419; Davie et al. 1997; Tsuk 2002c:7779) and 8 of which were rural [Cypros [long/lower], Hyrcania [long/upper/southern], Jericho palaces – south-west and north, Qumran, Urtas, Machaerus and Umm Ratam] (Patrich and Amit 2002:17; Meshel and Amit 2002; Garbrecht and Peleg 1994:167168; Conder and Kitchener 1883:205, 206, 227; Patrich 2002; Glueck 1951:404-409; Dauphin 1984a:35; Netzer and Garbrecht 2002:367-372; Ilan and Amit 2002; Amit 2002b:261-262; Glueck 1937-1939:134; Linder et al. 2000:554). Twenty-six of these branch lines were probably used for agricultural or industrial purposes, fifteen of which were on urban lines. Around the top of many of the access holes through the top of the vault of the Caesarea Low Level aqueduct a wide ring of mortar was found to support a superstructure that sometimes survives as several courses of cut stone and sometimes a single course of worked stones (Everman 1992:186). It is thought that these may have served to draw off water for nearby agricultural land, most likely using water-lifting devices, possibly a shaduf (Wilson 1999:328). On the Banias aqueduct a steep chute descended 2.4 m to the west to water nearby fields (Hartal 2002:90). A similar chute led to fields from the Sbeiteh aqueduct (Tsuk 2002c:77). A branch line leading to a settling tank and rectangular pool was found on the Sepphoris aqueduct (Tsuk 2002c:284-5). The excavators believed that the pool was used seasonally for swimming and bathing, but it seems equally likely, if not more so, that the pool was a storage reservoir for irrigation. A branch line drawn from the tunnel of the Beirut aqueduct probably had agricultural purposes (Davie et al. 1997:247). Saarisolo (1927:51-53) found several branches on the late Roman Tiberias aqueduct leading down to the lakeshore (Winogradov 2002:299-300). Six storage reservoirs were also found in the agricultural zones near this aqueduct, all lined with the same plaster as the specus. The best-preserved reservoir seems to have been fed by terracotta pipes and part of a channel leading water from the pool to the fields also survived. In addition, at least six mills seem

86 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST to have been fed by the aqueduct. The same aqueduct almost certainly supplied a probable dyeing complex of uncertain date consisting of a series of interconnected pools. Two other features on urban lines may have had rural or agricultural uses. On the Beth Govrin aqueduct (north) a grooved stone lying in topsoil is believed to have been part of a chute feeding agricultural zones (Sagiv et al. 2002:181). Secondly, on the Samaria (Shechem/Sebaste) line a secondary channel was found, but it is unclear if this channel irrigated the slope or was a by-pass channel for the settling pool near by (Frumkin 2002:272). Seven of the branches probably fed rural buildings, five of which were taken from urban lines. Three of these branch lines came off the Caesarea High Level aqueduct (Porath 2002b:116-7). One branch piped water west from channel C to a reservoir at either a villa or a monastery at Tell Tadwira. Two branches came off channel A; the first branch (known as Channel E) seems to have fed a pool at a late Roman site on Tell Tanninim and the second piped water to the ‘Christian building’ where it fed a bathhouse and fishpond. Finally, the Tiberias aqueduct fed two branch pipelines to bathhouses at Beth Yerah and Hammath (Saarisolo 1927:53; Winogradov 2002:300-302). We know from water laws that the drawing of water direct from aqueduct lines, rather than from reservoirs, was illegal, but numerous archaeological examples suggest that it was common practice (Cod. Theod. 15.2.4-6; Frontinus 2.97). Permission could, however, be sought to draw off water with a regulated calix. This rule was often flouted as shown by an inscription from Ephesos that refers to the damage caused to an aqueduct by illegal tapping inside the city (Wiplinger 2004-6:21-23). In addition, the Theodosian Code (15.2.3) records specific rules about how much water may be drawn off for private bathhouses. The Theodosian Code (15.2.2) also records that pipes off the aqueduct at Daphne were too large. Some of the branch lines listed above may have been illegal, but some of them may have been granted exemptions. It seems, for example, from Italy and North Africa, that the practice was common. In the cases of the seven branch lines to rural buildings, it would seem reasonable to assume, given the high status of the buildings being fed (villas and bathhouses), that these lines were paid for by the wealthy owners of these rural ‘estates’. This also

CHAPTER TWO

87

may have been the case in the Roman Campagna; Pliny (NH 31.42) for example complains that the waters of the Aqua Marcia and Aqua Virgo were diverted to villas and suburban gardens (Wilson 1999:315). In addition, we know from Frontinus (Aq. 78-86) that about two-sevenths of the aqueduct water for Rome was delivered outside the city itself (Wilson 1999:315). Villas and bathhouses close to the Nîmes and Arles aqueducts also may have benefited from a channelled supply (Benoit et al. 1994:160-62; Wilson 1999:323, 326). It is more difficult to know who owned and paid for the branch lines used for agricultural and/or industrial purposes. There is evidence from the East for the following landholders: the state (fisc or Emperor), the Church, large-scale landowners (such as Libanius who owned entire villages peopled by tenants), small to medium landowners and lessees (free-farmer communities, who were not coloni) and the military (veterans and soldiers (Kraemer 1958:20; Jones 1964:415-6; Liebeschuetz 1972, passim; Mango 1984:312, 322, 409; Decker 2001:38-45). Unfortunately inscriptions do not survive on the branch lines or installations that might suggest which of these landholders owned them, nor do we have any other evidence concerning landownership in the vicinity of these branch lines, such as the two maps from near Rome, which detailed the names of estate owners and the numbers of pipes they owned (CIL VI 1261; CIL XIV 3676; Wilson 1999: 315-16; De Kleijn 2001). Several scenarios are possible. The installations may have belonged to wealthy landowners who were entitled to and/or paid for their water supply, such as the state, the Church (in the late Roman period) or large-scale landholders. Alternatively, a group of lower class people may have co-operated to pay for the water supply. While co-operation in agricultural projects specifically is not attested to, there is evidence for groups of farmers pooling money in other contexts. At Qabr Hiram, near Tyre, a group of farmers (georgoi) paid for a church mosaic pavement, and at Zahrani, near Sidon a group of farmers donated money for marble revetment of a church ambulatory (Chehab 1957:101; Decker 2001:42). In the Hauran, georgoi set up a statue to Nike at their own expense (Macadam 1983:113; Le Bas and Waddington 1853:#2479). In these two scenarios the cost of the branch line could probably have been easily borne out and recouped by the increase in profit from a higher agricultural yield. The lines may also have tapped the aqueducts

88 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST illegally. In this case, while it may have been possible that this was the action of a wealthy, but miserly landowner, it seems more likely that it would have been done by poorer, peasant farmers. Where actual channels or pipelines left the aqueduct, as on the Caesarea High Level aqueducts, their visibility would suggest that they were legal (or the channel-builders foolish). In terms of the rural versus urban debate, it makes good economic sense to use water that is travelling through agricultural land for agricultural purposes as well, especially if it is making previously badly-watered land more productive. This would not only have saved money that would have to be spent on alternative irrigation schemes, but also brought in money. After all, urban centres could not consume what was not being produced. The use of water from urban aqueducts in a rural environment makes them both consumptive and productive. It has also been noted that the channel on the Tiberias aqueduct narrowed and carried less water as it reached the city (Winogradov 2002:299). A similar set-up has been found on the aqueduct to Zabi, Algeria, which has twin, parallel channels measuring 0.6 m and 0.2 m wide, the wider one of which had branch lines feeding land adjacent to it (Payen 1864:11; Gsell 1902:76-77; Wilson 1999:321). This led Saarisolo (1927:15 n.2) to hypothesise the following: ‘The survey is of the opinion that this great aqueduct was built only for the mills and irrigation purposes…But why should the Romans in Philoteria (Kerak), Sennabris (Sinn en Nebrah) and Tiberias not have procured a supply of good drinking water also…?’

Chronological detail would be useful here: were the branch lines and related structures contemporary with the main lines and part of the primary conception or added at a later date? The branches off Caesarea High Level aqueduct channel A fed significantly later buildings, so, while probably legal, were probably a later addition to the network. The similar construction of the reservoirs to the main Tiberias aqueduct line, however, may point to them being contemporaneous and therefore the rural supply was part of its original conception. While it may go too far to suggest that the urban centres were an afterthought in the building process, it is certainly significant that some of the branch lines seem to have

CHAPTER TWO

89

been planned from the outset. In such a scenario, rural areas achieve a higher status and importance in aqueduct planning than has sometimes been thought, which should maybe not be a surprise. After all, the urban centres were little without their rural hinterlands, which in turn could not function effectively without a good water supply. The large size of the irrigation reservoirs in comparison to the urban reservoirs (see Chapter 1), seems relevant here as it again highlights the importance of water supply and storage in rural areas. In the case of branch lines with granted exemptions, it can be argued that Ellis’ (1997) theory of using water for social control is actually enhanced by this interpretation. If water was brought deliberately to rural areas, not only was a higher yield of crops almost guaranteed, but also the rural populace were placated and would not begrudge their urban counterparts. This would go against popular views of disgruntled Mediterranean peasants griping that they ‘would have the citizens wash less so as to leave more water for the crops’ (Leveau 1991:159).21 In the case of illegal lines, we might see an opportunistic flouting of Roman law for the farmers’ ends. Although he is talking about trade, it is interesting that Libanius (Orat. 11.230) says that the Antiochene villages had little need for the town; this does not sound like a disgruntled rural population that is being taken advantage of by an urban centre, even one as powerful and important as Antioch. So, it would seem that, as argued for other parts of the Empire, the relationship between rural and urban environments does not respect the productive:consumptive dichotomy, but was more sophisticated and finely balanced (Wilson 1999:328). Contributing to this idea are four ‘urban’ aqueducts that fed gardens within or on the outskirts of urban and suburban settlement: Herodion (Porath 2002a:29; Mazar 2002b:243; Amit 2002b:256-261; Amit 1994:572), Daphne (Lassus 1983:210), Jericho palaces (south-west: Patrich and Amit 2002:17; Netzer and Garbrecht 2002:367-372) and Homs (Conder and Kitchener 1883:205; Seyrig 1959:189). In the case of the Homs aqueduct, of the 1800 21 Also see Bruun 2000, 215-6 who has difficulty finding explicit evidence for conflict over water between urban and rural populations.

90 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST litres per second discharge, only 300 litres per second was directed towards the city; the rest was dedicated to the fertile gardens of Homs. It is tempting to interpret this as a concern for both productive land and consumptive settlement as the whole system seems to have incorporated both urban and rural needs from the outset. A similar situation appears to have held outside Rome, where almost all the water from the Alsietina aqueduct, for example, seems to have been used for irrigation outside the city (Frontinus Aq. 11.1-2, 85; CIL VI.31566 = XI 3772a = ILS 5796; Taylor 1997, 488-92; Wilson 1999:317). The watering of gardens near Antioch and Homs is also interesting, especially when taken with the evidence from Rome (De Kleijn 2001:34), as these important cities would have provided lucrative markets for the extra produce that could be grown with increased irrigation and watering. This was paralleled in North Africa and Italy where an intensification in rural hydraulic infrastructure seemed to go hand in hand with proximity to urban markets (Wilson 1999:323). Furthermore, these data provide a handle for understanding whether aqueducts were a symbol of an imperial authority. In both the legal and illegal scenarios this was arguably not the case: in one, we see a lenient, broad-minded authority and in the other an authority that could be ignored and taken advantage of. Six rural aqueducts that fed rural settlements were also used for irrigation and agricultural purposes: Alexandrion (lower/Ras Kunetra), Cypros (long/lower), Jericho palaces (north), Auara, Phasaelis and Khirbet Ayun Ghuzlan (Patrich and Amit 2002:17; Amit 2002c:307-310; Garbrecht and Peleg 1994:167-168; Meshel and Amit 2002; Conder and Kitchener 1883:206, 227; Oleson 1991; Eadie and Oleson 1986; Eadie 1984:217; Oleson 1986; Glueck 1935:65; Porath 2002a:33; Glueck 1951:416). Four of these six rural aqueducts that have two functions date to the Hasmonean or Nabataean periods. The Alexandrion aqueduct dates to the Hasmonean or Herodian period and the Cypros aqueduct is Herodian. In the case of the Jericho palaces (north) aqueduct, branch lines off the aqueduct seem to have been constructed in a similar manner to the main channel and may therefore have been contemporary. Is it possible, then, that multi-functional aqueducts were a feature of rural communities in the Near East prior to the Roman conquest? If this is the case, what we see in the Near East is a continuation and adaptation of a well-known practice that forms a sensible re-

CHAPTER TWO

91

sponse to the dispersal of much needed and sought-after water and possibly even a practice that was then adopted by the Romans. The two aqueducts associated with Herod may suggest again that he was one of the forces driving this continuation. What we appear to see here, then, is a more symbiotic and nuanced relationship between rural hinterlands and urban centres. The water being taken to cities and towns in aqueducts appears to have multiple other uses along its course that must cast doubt on the idea that cities were depriving rural areas of water. Furthermore, there is evidence that some urban aqueducts were used to water gardens close to the cities. Finally, the evidence from Hasmonean and Herodian aqueducts suggests that these multifunctional practices may have had earlier origins in the Near East, related to a common-sense approach to water management in an unpredictable and arid climate (see Chapter 4). The rural economy in the late Roman period There have been strong debates on the fate of the late Roman economy. Most of these seem to focus on a perceived ‘decline’ in urbanism and economic production in advance of the fall of the Empire (see Lavan (2009) for a summary of arguments in the Eastern provinces). In the Near East, however, there was an observable peak in the agrarian economy in the late Roman period (Decker 2001:337; Butcher 2003:139; Decker 2007; Oleson 2009). Several reasons have been put forward for this upturn. While the single largest landholder in the region was the state (in the form of the fisc or the Emperor), the Church also obtained and maintained significant areas of land for its ecclesiastical and monastic properties, for example around Dara and Antioch (Mango 1984:409; Decker 2001:39). Monasteries were centres of production and developed marginal areas, hence the cistern-fed gardens noted in Chapter 1. A growth in rural settlement and population, which had increased throughout the period of Roman rule, was most striking in this period, as attested, for example, by the Dead Cities in the limestone massif of Syria (Butcher 2003:140, 146). The amount of settled land in the hinterland seems to have been directly proportional to an increase in urbanisation, for example at Dara, Anasartha and Maurikopolis (Decker 2001:340). In this increasingly crowded landscape agricultural intensification must have been necessary to meet

92 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST personal needs as well as urban and export demands. We know from literary sources, for example, that grain was frequently sold on the open market in cities.22 In addition, Antioch and Apamea were exporting olive oil (Decker 2001:341). Libanius (Ep. 709) also sold wine in Cilicia (Liebeschuetz 1972:45; Petit 1955:305 n. 5). As noted above. this phenomenon has been observed elsewhere in the Empire, for example in North Africa and Italy, where an intensification in rural hydraulic infrastructure seemed to go alongside proximity to urban markets (Wilson 1999:323). It has been suggested that this intensification of rural production seems to be at odds with the literary evidence, which emphasised the crushing tax burden in the late Roman period (Liebeschuetz 1972:73). It seems possible though that it may have been just this tax burden that prompted an increase in agricultural production. The supply of the late Roman army and the annona militaris must also have exerted a significant amount of pressure on the production capabilities of the rural landscape. The late Roman army was divided into small legions of c. 1,000 men (unlike the large legions of the principate and early Empire), some of which were mobile field army units (comitatus) and some of which were permanently based. By the 4th century AD, with the exception of the III Cyrenaica at Bosra, most of the legions had moved to bases in remote areas (Table 7) (Butcher 2003:414). It seems likely that these legions were supplied locally and that the provinces were responsible for soldiers within their borders. Libanius’ letters (Ep. 21 (358)), for example, show that the provision of supplies for the army at Callinicum on the Euphrates and further east in Mesopotamia placed a heavy burden on 4th-century Antioch (Liebeschuetz 1972:163). In general, Egyptian grain was not used to supply the

Julian regulated grain prices in Antioch, which were higher than in Egypt, in AD 362-3. Grain prices were also regulated during the famine of Edessa. The Life of St Spyridon also attests to permanent markets in cities. See Liebeschuetz 1972, 128; Trombley and Watt 2000, 39; Decker 2001, 299; Butcher 2003, 167. 22

CHAPTER TWO

93

eastern armies, except under exceptional circumstances (Josh. Styl. 70; Decker 2001:302, 304; Liebeschuetz 1972:76).23 Legion Base Province IV Scythica

Taibe

Syria

XVI Flavia

Sura

Syria

III Gallica

Danaba

Syria

I Illyricorum

Palmyra

Syria

X Fretensis

Aila

Palestine

IV Martia

Betthorus (Lejjun)

Arabia

Table 7: Bases of late Roman legions in remote areas of the East.

The locations of large-scale irrigation works and the presence of the late Roman army seem to tally remarkably well. Several of the irrigation channels, for example, flowed within the territoria of late Roman fortresses such as Barbalissos, Callinicum and Circesium. In addition, there seems to have been a marked tendency for qanats to be employed in the vicinity of military sites. This may be because many of the military frontier sites were in the desert, and so qanats may have been the only viable option. Similar militarybased explanations have also been proposed to explain the seeming increase in numbers of field systems in the later Roman periods (Newson 2002:260). This can also be tied to the pattern of dams located along the Strata Diocletiana. It seems plausible that military labour and engineers would have been employed for a number of these irrigation projects. This highlights the nature of the impact of a Roman authority on the irrigated landscape of the Near East. While there is no incontrovertible evidence for any technological impact in terms of changes, improvements or additions to the existing methods, it seems that administrative changes and the presence of the late Ro23 On Julian supplying Antioch with corn from Egypt in AD362, see Liebeschuetz 1972, 130.

94 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST man army may have contributed to the intensification of agricultural production and irrigation. In this way we can see a reciprocal relationship between an imposed imperial authority and the existing population. That said it might go too far, or even be naïve, to suggest that the presence of the late Roman army did not prompt negative reactions. Although the authorities may have been content to respect the native knowledge found in the East, the actual effort involved in order to fulfil their requirements must have put considerable strain on the resources available to local populations. Urban productivity and industry in the late Roman period Many of the arguments on ‘decline’ in the late Roman period focus on the perceived ‘decline’ of urban life (see discussion in Mattingly and Salmon 2000; Lavan and Bowden 2001; Lavan 2009). As observed above, many of the large-scale industrial installations date to the late Roman period, which may point to a general upturn in industrial processing, larger-scale facilities that are more visible in the archaeological record, or the movement of industrial processing into the cities in this period. In Beirut alterations in the mid 4th century saw one shop become a fullery (Perring 1997-8:22; Perring et al. 1996:196; Perring 2006:32). Further ovens, sunken storage vats and mortar floors were inserted into shops in the area during the mid 5th to 6th centuries. This supports the latter argument (and possibly the former as well). Late Roman Jerash possessed not only two of the larger and more sophisticated installations, but also saw substantial changes in the area and buildings surrounding these installations in order for them to function effectively, for example the dyeworks intruded on a previously public space (the macellum) (Uscatescu and Martin-Bueno 1997). Furthermore, the sawmill in the Temple of Artemis complex was involved in the spoliation of the city’s Classical monumental architecture (Seigne 2002). This evidence seems to point to a change in emphasis in the use of public space in the city during its later history. Monumental space appears to have been used more frequently for utilitarian purposes, a trend that also seems to have occurred at Apamea (Balty 1981; Balty 1988; Millar 1993:256-263; Ball 2000:159-161; Butcher 2003:passim; Lacoste 1941; Shahada 1957; Rey-Coquais 1973:41-46; Balty 1983; Schmidt-Colinet 1984; Balty 1987; Leveau 1991:154; Neudecker 1994). The aqueduct, for example, appears to have continued in use, but its water was stored in

CHAPTER TWO

95

reservoirs in the centre of the city. Reservoirs were inserted at the cardo/decumanus intersection and in the northern area of the agora. Also, there seems to have been a clear encroachment on public space in the agora as a latrine was inserted into its eastern entrance, off the colonnaded street. It is likely that a combination of factors was responsible for this change in emphasis at Apamea. It is possible that there was a population increase in the late Roman period caused by the use of the city as the winter quarters of the Second Parthian Legion, possibly as a response to the looming threat from further east. Its position as an important bishopric may also have brought higher numbers of visitors to the city, which may have prompted the construction of the latrine and the possible bathhouses. The seeming disregard for previously important elements of the city, such as the agora, appears to be part of a wider phenomenon at Apamea. This is illustrated starkly by the treatment of inscriptions related to C. Iulius Agrippa, the major 2nd-century AD benefactor (Butcher 2003:365). The inscription describing his benefactions, such as the aqueduct, was re-employed as a lintel for a window, ironically in his portico. In addition, a dedication to the same man was reused as a kerb stone in the colonnaded street. Finally, the fate of various buildings at Scythopolis has a complex history with some structures, such as the nymphaeum being chosen for restoration, while others such as the basilica were not maintained (Tsafrir and Foerster 1997). The transformation of the propylon connecting the temple with the acropolis makes an interesting example. It has been argued by Tsafrir and Foerster (1997:115) that this area was used exclusively for religious processions and pagan ceremonies. When the temple was destroyed, this area subsequently lost its religious importance and was rebuilt, in the second half of the 4th century AD, as an industrial complex with pools, water pipelines and 'other installations'. The shift from paganism to Christianity must have played a key role in making this change in use and outlook possible. In addition, this example attests again to the increasingly industrial nature of cities in the late Roman period and their economic vibrancy. The monumentality of cities, then, no longer appears to have been a major concern for their inhabitants. This seems to be related to changes in the power structure of cities that saw power held increasingly by individuals, in particular bishops, rather than by the boule. In this, notably Christian, atmosphere the monumentality of

96 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST cities no longer seemed to fit with cultural ideals, where wealth was used to enhance the status of the individual rather than that of the community (Butcher 2003:269; Liebeschuetz 1972:101-105, 136, 259-260, 262), rather collective enterprises were focused towards religious expression (Millar 2008:229, 240). Public entertainment came under state control in the 5th century, which also diminished the sense of a city’s identity. The insertion of the vats into the Jerash hippodrome may reflect such a change in control over public entertainment. The assumption, however, that these changes were indicative of an urban downturn, does not seem to follow. The increased industrial capacity of cities points rather to busy, bustling cities that were full of activity. The cities undoubtedly changed, but should not necessarily be seen as going into decline. Furthermore, the 'Christianisation' of cities should not necessarily be seen as bringing about decline. As shown in Chapter 1, the Church was actually responsible for significant contributions to the upkeep and maintenance of the water supply and management infrastructure, which would have been vital to urban life and living. The issue of decline is more about our expectations of what an ancient city should be and informs us more about our own preconceptions and values than about the situation in the ancient world. Indeed, this evidence can work with that provided by aqueduct branch lines and bring us closer to the picture of a productive city. In this case, what we can also show is that there was significant production capacity, some using innovative and sophisticated water technologies, within cities in a variety of different industries, most notably in fabric-related industries. Also, this example highlights that the chronological element may be very important to this debate as cities are not monolithic, unchanging spaces and places, but rather adaptive and ever-changing.

CONCLUSIONS The evidence and discussion presented here suggest that there have been significant moves away from Finleyite views of the ancient world, and in particular of how cities and rural areas functioned both separately and in relation to each other, as well as adding further weight to the argument against Finley's ideas of technological stagnation (see Chapter 1). Furthermore, the somewhat ingrained idea of ‘decline’ in the late Roman period is hard to support with evidence from the eastern provinces. The rural economy appears to

CHAPTER TWO

97

have seen an intensification of production and cities appear to have moved towards more industrial activities. These changes may have been, at least in part, forced by serious military needs and challenges, but also, particularly in the cities and towns, by a radical religious shift towards Christianity that had a deep impact on urban life, ways of living and attitudes towards urban space.

3 ATTITUDES TOWARDS WATER AS A RESOURCE IN THE ROMAN NEAR EAST ‘As it would never occur to the maidens who went to fetch water from the spring to turn it off, so the system in the town was not turned off.’ (Peleg 2000:241)

Traditional research on Roman water supply and management has emphasised the idea that the system worked on a constant-offtake principle (in particular Hodge 1992:3, 79, 89, 279, 280, 296, 303, 322. Also Forbes 1964:172; Leveau and Paillet 1976; Shaw 1984; Cotterell and Kamminga 1990:51). This is the idea that water flowed constantly through the system and was not stored in significant quantities at any point, as epitomised by Peleg’s statement above. In effect, this means that there would have been no reservoirs, no dams and no way of preventing runoff of unused water from nymphaea, which points to an intrinsically wasteful and lavish approach to water supply. This model has come under criticism from Wilson (1997; 2001) for North Africa where he cites evidence of large, covered reservoir-cisterns for water storage (also see De Kleijn 2001:37). In an area, such as the Near East, that still struggles with an unpredictable rainfall regime and hot, dry summers, it is easy to imagine that water conservation would have been an issue in the past as well. Two ways of approaching this issue are taken in this chapter. Firstly, did the Roman and late Roman Near East have the technological capacity to conserve water, by using dams and urban storage facilities, such as those found in North Africa? Secondly, what were the social attitudes towards lavish water supply? This latter section will look at evidence from domestic contexts and the provision of urban public facilities, such as fountains and bathhouses.

99

100 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST

TECHNOLOGY AND WATER CONSERVATION Dams As one of the most important functions of a dam is its role as a reservoir, the presence of dams strongly contradict the constantofftake principle. Even Hodge (1992:79), who is one of the strongest supporters of the constant-offtake principle is forced to admit that dams were ‘at variance with the traditional Roman aqueduct principle of constant-offtake.’ Dams were not a Roman innovation in the Near East. They had formed a significant element in the water supply of the area from the Bronze Age onwards, particularly in areas with low rainfall. Dams became increasingly popular in the periods directly before the Romans, particularly in Nabataea, where 15 well-dated dams are known.24 The zenith of dam-building, however, occurred in the Roman and late Roman periods, where 45 dams have been identified. As discussed in Chapter 1 the real art of dam building in the Near East lay in the ability to choose the right location and to design a dam that would suit the needs of that location most economically. The Harbaqa dam, near Palmyra, for example, was located in a particularly good position because its reservoir was fed not only by the Wadi al-Barde, but also by other large wadis that flow into the al-Barde a few hundred metres further up. In addition, the ring of mountains that almost entirely encircle the area around the dam must also have contributed large quantities of runoff water. In this way, the dam-builders were able to conserve and make-use of the seasonal and sometimes meagre resources available. This concern for water conservation should not necessarily be equated with concern for water supply ethics and the environment. As modern controversy has shown over recent dam projects, particularly in the Middle East, the building of a dam does not only have positive aspects and to many it is not viewed as a valuable means of providing and conserving more water. In many cases the construction of a dam involves the displacement of large numbers 24

On the dating of dams, see Kamash (2006b).

CHAPTER THREE

101

of settled and itinerant people, including the loss not only of their homes, but also of their agricultural land.25 Furthermore, the effect of impeded water supply downstream has in modern times led to heightened tensions between neighbouring countries, including threats of air attacks. There is also evidence from the 8th century BC for conflicts arising between authorities over the availability of river water: the king of Assyria, Shamshi-Adad, ordered the release of water from the Balikh River to irrigate Tuttul, but a local authority blocked the river to irrigate another city (Wilkinson 1998b: 151; Bagg 2002:228). The impact of dams on the environment is also of concern, for example the creation of large lakes in south-eastern Turkey has seen the introduction of malaria into a previously nonmalarial region. While evidence for such effects is hard to identify in the archaeological record, one can assume that similar considerations would have affected inhabitants of the Near East during the Roman period. Urban water storage The constant-offtake principle debate focuses on the lack of storage facilities for aqueduct supplies. Aqueduct supplies can be stored in three types of storage installations: reservoirs (open ‘pools’), cisterns (covered storage with a volume less than 1,000 m3) and reservoir-cisterns (covered storage with a volume greater than 1,000 m3). Wilson’s criticism of the constant-offtake principle in North Africa was centred on the presence of significant numbers of aqueducts ending at reservoir-cisterns. At the outset of this research it was expected that due to similarities in environment and climate that the situation in North Africa would be mirrored in the eastern provinces and that buffer reservoirs would have been provided in towns and cities, which could balance supply against usage. While there are some large reservoir-cisterns, such as those at Resafe, the picture does not seem to be so straightforward as, nor to mirror the situation in, North Africa. Whereas almost all North African aqueduct-fed sites were associated with storage facilities, only 18 aqueduct-fed sites in the 25 See Métral 1987 for an interesting discussion on the modern implication of the Tabqa-Thaoura dam in Syria.

102 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST East had large storage facilities, leaving 23 urban sites fed by aqueducts that do not seem to have storage facilities. While partial excavation and publication bias may account for some of this disparity, there are further differences, which make this disparity warrant further investigation. North Africa, for example, had markedly more installations with capacities greater than 2,500 m3 than the Near East (Table 8). In terms of capacity, over 72% of the urban water storage installations were in the ranges below 2,500 m3 (19 installations); there was a marked drop-off in numbers in the larger capacity ranges. In general, with the exception of the massive open reservoir at Capitolias, larger volumes (2,500 m3 and over) were, unsurprisingly, stored in covered installations rather than in open ones, which is probably due to lower evaporation and less risk of water contamination in covered installations. The largest reservoir (Capitolias: 15,500 m3: Tsuk 2002a:293; Harding 1967:56; Schumacher and Le Strange 1890:162-166) and reservoir-cistern (Resafe: 14,600 m3: Brinker 1991) in the East were significantly smaller than the largest North African reservoir-cisterns from Carthage, Cirta and Zama. The Near East had no cities with reservoircisterns in the over 20,000 m3 category, whereas North Africa had three: Carthage (Bordj Djedid: 20,000 m3 and La Malga: 50,000 m3), Zama (28,000 m3) and Cirta (30,000 m3). Even without including these exceptionally massive reservoir-cisterns, a large discrepancy between approaches to urban aqueduct water storage in the two areas of the Empire still appears. The fact that a large majority of the dated storage installations were not contemporary with aqueduct construction, but were later additions to the water management system further compounds this difference (Table 9; see discussion below). Capacity (m3) East North Africa 500-999 6 1 1,000-2,499 3 4 2,500-4,999 4 7 5,000-9,999 1 8 10,000-19,999 2 5 20,000+ 0 4 Table 8: Comparison of urban aaqueduct-fed storage installations by capacity (over 500 m3) in the East and North Africa. Data on North African reservoir capacities is based on Wilson (1997:79-80, table 4).

CHAPTER THREE Date Roman Late Roman

Reservoir 4 13

Cistern 2 8

103

Reservoir-cistern 2 5

Total 8 26

Table 9: Aqueduct-fed urban water storage installations (including in bathhouses) by date.

So, did the constant-offtake principle, where aqueduct water was not stored in significant quantities along its course, apply to the Near East? Firstly, I will address the question of to what extent the towns and cities in the Near East relied on an aqueduct supply. This will include an analysis of whether there were complementary resources. An assessment will then be made concerning the extent to which there was a shift in approaches to water management in urban contexts in the late Roman period, how this might affect the constant-offtake principle debate and what may have been the motivations for this shift. In order to answer the question of the extent to which cities and towns in the East relied on their aqueduct supplies, this section will consider how many sites were supplied only by aqueducts and whether this is a reliable picture. Analysis will also be made of which sites were supplied by a combination of sources, such as aqueduct water, well water and rainwater. Aqueduct-fed storage Site Acco Antioch Apamea Ar-Rabbah Ashqelon Ba’eij Beirut Beth Govrin Bosra Burqa’ah

A X XD X

R

C

X

X

RC

Rainwater/ Other runoff storage R C W X X

X

X X

X X X

X X

X X X

104 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST Aqueduct-fed storage Caesarea Capitolias Dafyaneh Dara Deir al Kahf Dor Dura Europos Eboda Elusa Homs Horvat Castra Horvat Zikhrin Irbid Jerash Jerusalem Kurnub Lejjun Nessana Palmyra Petra Qanawat Qasr alHallabat Resafe Saadi Sabkha Sabkhiya Sbeiteh Scythopolis Sepphoris Sumaqa Tiberias

XD X

X X

Rainwater/ runoff storage

X

Other X

X X

X

X X

X

X

X

X X

X

X X

X

X X

XD X X X X X

X X

X

X

X

X X X X

X X X X

X

X

X

X X X

X X X

X

X

X X X X

X X X X

X X X

X

X X X

X

X

X X X X X

X

CHAPTER THREE Aqueduct-fed storage Umm alJimal Umm alQetein

X

105 Rainwater/ runoff storage

Other

X X

Table 10: Water sources and storage installations in the East. (A: aqueduct, D: dam, R: reservoir, C: cistern, RC: reservoir-cistern, W: well).

Although the archaeological data for some installations, such as wells, are sparse, several observations can be made usefully about the complementary use of water sources and storage installations across the East. Table 10 presents the known data across 41 sites in the East. This does not include the 17 sites only fed by aqueducts without known public storage facilities or other complementary water sources.26 Of these 12 have not been subject to extensive excavation or survey, in most cases because of the overlying modern town, such as Damascus, Aleppo, Lattaqia and Tyre, making the negative absence doubtful. The aqueduct at Lattaqia, for instance, is known only from literary evidence (Josephus BJ 1.21.11). More is known about Emmaus and Susita/Hippos, but it is of course always possible that the other facilities have not been published or yet been found. Eight sites only had aqueducts and aqueduct-fed storage: Antioch, Apamea, Bosra, Capitolias, Dara, Homs and Umm al-Jimal. The lack of other resources at Capitolias and Dara is probably due to the low level of archaeological investigation at these sites; attention has focussed, there, on the larger, more monumental remains. More research has been undertaken at the other sites, which makes the lack of rainwater storage installations at these sites difficult to explain without suggesting that the aqueduct supplies were sufficient. 26 These are: Aleppo, Banias, Damascus, Edessa, Emmaus, Lattaqia, Neapolis, Samaria, Samosata, Si, Sidon, Susita, Suweida, Tripoli, Tyre, Umm Qes and Zeugma.

106 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST Information on rainwater storage and wells from Antioch and Homs may be lacking due to the overlying modern towns, which restrict archaeological investigation. Antioch and Homs also appear as two of the sites that did not have urban storage facilities. While this may be due to excavation bias, the fact that they both had dams may be significant because this may have obviated the need for large-scale storage facilities inside the city. Although Caesarea did have storage facilities inside the town, they were small in scale: the known reservoir capacities were just 36 m3 and 26 m3 (Porath 2002b:124). Dammed water storage may also explain this apparent anomaly. The lack of wells at these sites may be due to environmental factors. Apamea, for example, was on a plateau above the Orontes valley and therefore digging wells may have been undesirable due to the depths necessary. Similarly, topography also dictated the water supply at Dura Europos on a plateau above the Euphrates, though unlike Apamea an aqueduct supply was not possible at this site. Both Bosra and Umm al-Jimal were in basalt areas, which would have made digging wells exceptionally difficult. The underlying basalt geology may also explain several of the sites that relied on stored cistern supplies: Ba’eij, Burqa’ah, Dafyaneh, Deir alKahf, Sabkhah, Sabkhiyeh, Umm al-Qetein and Irbid. Covered cisterns, rather than open reservoirs, must have been used on these sites because they were supplied by seasonal runoff water supplies, so long-term storage and prevention of evaporation would have been a high concern. Evidence for the complementary use of aqueduct water, well water and rainwater only came from three sites: Dor, Petra and Tiberias. There were, however, 16 sites that showed a combination of two of these water sources. Seven sites are known to have used aqueduct and rainwater supplies: Acco, Jerash, Jerusalem, Qanawat, Sbeiteh, Scythopolis and Sepphoris. Given the propensity of other Negev sites for wells, one might expect to have found wells at Sbeiteh. As the site has been subject to thorough survey, it is likely, therefore, that Sbeiteh was atypical because it was the only Negev town to have an aqueduct supply, which may have rendered wells unnecessary. Five sites made use of aqueduct supplies alongside wells: Ashqelon, Beirut, Beth Govrin, Caesarea and Palmyra. As Palmyra was in an area with low rainfall under the 200 mm isohyet (the guideline

CHAPTER THREE

107

threshold for classing semiarid zones as being outside the limit of rain-fed farming), the lack of rainwater cisterns should maybe be expected, but may also reflect under-reporting of cisterns. In addition, as an oasis site, well water would have been easily accessible near the former lake. Conversely, it would be reasonable to expect that the coastal sites (Ashqelon, Caesarea and Beirut) would have taken more advantage of the higher rainfall, especially Beirut, which would have benefited from the effects of the Lebanon and AntiLebanon mountain range. The action of the mountains is brought into sharp focus by the Homs-Tripoli gap: the region west of Homs is noticeably better watered than areas to the north and south that are in the lee of the Lebanon and Anti-Lebanon mountain ranges. Beirut has, however, been subject to only limited excavation due to its troubled recent history. In the case of Caesarea, the use of complementary resources may be tempered by the idea that well water was only used when there was no aqueduct supply or when the aqueduct supply was in decline. A further four sites used wells alongside rainwater storage to guarantee a stable supply: Eboda, Elusa, Nessana, and Saadi. Notably these sites are all in the Negev, none of which had an aqueduct supply due to their desert location (the single exception in this area was Sbeiteh, see above). Rainwater storage may also not have been a reliable, perennial source, so wells would have been a necessity. This is brought out clearly in a letter from Jerome to Procopius, which includes a complaint that the well water at Elusa was brackish and barely palatable (Mayerson 1983:251. Procopius Ep. 2). This suggests that well water was relied upon when reservoir water was unavailable, for example in the middle of summer, or during drought years. There is limited evidence for the use of river water in urban centres. It is unlikely that rivers would have been used to supply urban drinking water, but it is possible that river water may have been lifted to supply industrial installations. While the lack of large-capacity aqueduct-water storage facilities in urban contexts seems to suggest that the constant-offtake principle applied in the East, i.e. that there was a wasteful approach to water management, this would seem to be an overly simplistic analysis. Firstly, dams creating large storage reservoirs outside the cities may have rendered further large-scale storage inside the city unnecessary. In addition, many of the sites under consideration did

108 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST not rely solely on aqueducts for their perennial water supply, indeed several had no aqueduct on which to rely. In the majority of cases complementary resources would have covered seasonal shortfalls in supply. Reliance was rarely placed on a single water source, be it an aqueduct or other source, the choice of which was often determined by topographic, geological and climatological factors. There is also evidence from taps, at Sepphoris (Tsuk 2002a) and Auara (Eadie and Oleson 1986), that, actually, the aqueduct supply could be turned off. In addition, the system at Capitolias that combined a large open reservoir with a tunnel reservoir-cistern, which could be used for back-up storage, suggests that the aqueduct supply could be more finely balanced than the constant-offtake principle would allow (see Chapter 1). The development of water storage and distribution in urban centres over time also reveals some interesting patterns. One would expect to find that aqueduct and storage facilities were contemporaneous and this is the case for four sites: Sepphoris, Tiberias, Beth Govrin and Resafe. There are, however, several sites in which this is not the case. The late Roman period at Apamea, for example, in water supply terms, was typified by the building of small reservoirs, whereas the earlier period was concerned with the aqueduct (Balty 1987). A similar pattern of earlier aqueduct and later storage is also attested at Jerusalem, Acco, Caesarea and Dor. The fact that aqueducts and storage on the same site were not necessarily contemporary has obvious consequences for any theory of buffer reservoirs. In the East there seems to have been a shift in approach to urban water management in the late Roman period (Table 9). If the constant-offtake principle was ever used in the East, it seems not to have been the principle on which water management functioned in the late Roman period. This shift did not just affect how water was managed and, in particular, stored, but also where the new water management installations were placed in the urban landscape, such as the latrines placed in the entrance to the agora at Apamea and the change in use of the macellum to a dye works at Jerash. It would seem that not only were there shifts in how water was managed in this period, but also in how urban centres functioned, so we need to ask what brought about these changes. As shown by the encroachment of installations making use of water in public spaces, the monumentality of the city no longer appears to have been a major concern for its inhabitants (see Chap-

CHAPTER THREE

109

ter 2). Individuals were keen to gain heaven’s approbation, thus bringing about a lowering in the number of donations made for public buildings and an increase in the wealth of the Church. Indeed, the number of water storage facilities, both reservoirs and cisterns, associated with church buildings strongly suggests that the Church was one of few institutions that could afford their construction and upkeep. It is also possible that the strong links between Christianity and water, as shown by the placing of Christian holy sites near springs and aqueducts, may have made it appropriate for the Church to concern itself with water provision. These changes in urbanism may explain the small-scale reservoirs at Apamea (Balty 1987:22) and Caesarea (Porath 2002b:124). These reservoirs are very different in scale from the massive, late, open reservoirs at Constantinople and Ptolemais.27 What the Eastern reservoirs appear to represent is a more fragmented, less centralised approach to water management in this period. This seeming lack of any overarching water management plan may be attributable to the extensive administrative changes of the period. The decline in civic self-government with a lack of effective power in the council and the rise in power of the Church, or more specifically the bishop as city leader, did not just affect public munificence, but also appears to have had a significant impact on wider water management strategies. In the case of these small storage installations, it would seem that they might have been the result of private or small group enterprise. This may have been the only available option when municipal and provincial governing bodies were tied up with other more pressing issues, such as the Saracen revolt in Caesarea, which led to the disruption of the aqueduct systems. The increase of storage inside the city walls in this period may also point to a fear of siege. The construction of a string of frontier cities, such as Dara, Resafe and Khan al-Manqoura, must be viewed Constantinople reservoirs: Aetius (5th century): 197,000 m3; Aspar century): 220,000 m3; Mocius (6th century): 250,000 m3: Ceçen 1996; http://longwalls.ncl.ac.uk/Water/Constantinople.htm. Ptolemais reservoirs: buildings 11 and 12 (52.5 m x 37.5 m and 118 m x 126 m (26 million gallons) respectively): Kraeling 1962:71-2. 27

(5th

110 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST as a deeply-felt concern over the increasing power from Persia. In addition, is it mere coincidence that Resafe and Dara were two of the cities with the largest storage facilities in the eastern empire? An inscription from an agraria statio in Provincia Arabia tells us of an officer in AD 334 who had seen many of his men killed in an ambush by Saracens while collecting water and so built a reservoir, which lends weight to the idea that water storage was closely allied to protection from external force (Iliffe 1944; Année Epigraphique 1984:136; Di Segni 2002:52).28 These examples seem to indicate that the water management system worked effectively when ordered directly from the Emperor, but otherwise strategies were implemented on a rather ad hoc basis. Environmental factors may have governed how water was managed in the late Roman period as well. Although we know little about widespread climate changes in the Roman and late Roman periods, there is evidence about the effect of earthquakes and the imperial provision of emergency aid in their aftermath. Justinian, for example, donated money after the earthquake of AD 526 for the restoration of the aqueducts at Antioch (Malalas 422.4-5). In addition, in response to a drought in AD 520 in Jerusalem, we are told that: ‘…the archbishop, worried at the unrest of the population, began a survey of the more humid places, putting a large number of hands to digging pits, expecting to find water, but could not find it. He went down to the Siloam valley near St Cosmas’ Cave, beside the road to the Great Laura, and with the help of an engineer and a large number of labourers dug down to a 28 CVM PERVIDISSET VINCENTIVS PROTE/CTOR AGENS BASIE PLVRIMOS EX AGRA/RIENSIBVS DVM AQVA SIBI IN VSO TRANS/FERERENT INSIDIATOS A SARACENOS PE/RISSE RECEPTACVLVM AQUAR EX FVNDA/MENTIS FECIT’ OPTATO ET PAULINO VV CC CONSS. ‘Vincentius, who was acting as chief of the bodyguard of Basius, observing that many of the outlying pickets had been ambushed and killed by the Saracens while fetching water for themselves, laid out and constructed a reservoir for the water. He did this in the consulship of Optatus and Paulinus, both distinguished officials.’

CHAPTER THREE

111

depth of 40 fathoms [c. 70 m]. But the archbishop could not find any water, and was in great distress…’ (Cyril of Scythopolis Life of Sabas chap. 54)

Not only does this illustrate that alternative solutions, such as wells, were looked for when there was a shortage of supply, but also highlights again the power of the Church in this period: it is an archbishop, not a city official, who responds to the crisis and takes on the responsibility of finding a solution in order to placate the population. It appears then that several factors may have explained the move towards increased water storage in the late Roman period. These included internal changes to how cities functioned and were governed, which saw the rise of the Church and the decline of civic self-government, instability in the region (and so protection of the water supply in the event of a siege), and environmental factors that prompted emergency action. So, while the East was similar in many ways to other areas of the Empire with aqueducts (occasionally terminating at castella divisoria), nymphaea and storage installations and water being distributed around the city via a network of channels and pipelines, supplying bathhouses, smaller storage installations, street fountains and houses, there were significant differences. In particular, there was not a total reliance on aqueduct supplies. Most cities and towns employed complementary water resources, such as wells and cisterns. Despite the low numbers of large capacity storage installations, this suggests, alongside the evidence from the large numbers of dams, that the basic wasteful premise behind the constantofftake principle was not at work in the East. The late Roman period in the East saw a change in approach to water management, with an apparent increase in water storage facilities. These changes seem to have been related to the general changes in urban life in the late Roman period, which saw the decrease in the monumentalisation of cities with the rise of the Church. This was concurrent with the decline of city and town governing bodies. Further pressure on urban water management was exerted by the threat from the East, as well as natural environmental disasters such as earthquakes and droughts.

112 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST

SOCIAL ATTITUDES TOWARDS WATER CONSERVATION The technological evidence suggests that water conservation was an important issue across the Near East for a number of different reasons. This preoccupation can also be traced in public facilities, especially nymphaea and bathhouses, as well as in the private sphere in domestic contexts. In general, the volumes of water used in these contexts seem to have been fairly restricted in comparison to other parts of the empire, which suggests that these concerns were widespread and consistent over time. In this section, the evidence from nymphaea and bathhouses will be considered, before looking at how water was used in private houses across the Near East. Nymphaea Nymphaea were ornamental structures comprising a public drinking fountain with a pool as part of its façade (Segal 1997:151). In general, these were more lavish and monumentalized than a regular street fountain, which had much simpler designs. This made them key points for euergetic display across the empire, including in the East. The nymphaeum from Pella, for example, is depicted on coins under Elagabalus, where it is shown as having three stories and a façade with statues (Dvorjetski 2001-2:503). Its architectural style would have made it similar to the façade nymphaea found in Asia Minor at Miletus, Aspendos, Perge and Side, though without such impressive dimensions. The basin at Caesarea (15 m long x 3 m wide) was also ornamental with marble-facing and a life-size marble statue (Porath et al. 1998:45). The relationship of fountains to the Emperor is also instructive here. We know of a nymphaeum at Suweida dedicated to the city in honour of Trajan, under a governor (either Julius Quadratus or more likely his predecessor Cornelius Palma) (Le Bas and Waddington 1853:#2305 and #2308; Glaser 2000:445; Di Segni 2002:54. IGR III:1273, 1276). The statue of Trajan was placed in a very prominent position: directly above the water outlet in a single storey nymphaeum façade (Glaser 2000:445). Such an iconographic link between water supply and empire was also made in Volubilis, North Africa where the Arch of Caracalla (AD 216/7) had two fountains incorporated into two niches on each face of the arch (Wilson 1997:154). Statues of the Imperial family also featured in the elaborate 2nd-century nymphaeum of Herodes Atticus at Olympia,

CHAPTER THREE

113

which emphasised the donor’s close link to the family (Glaser 2000:440-1). Similarly in Argos a colossal statue of Hadrian was placed in the nymphaeum, probably in honour of the Emperor’s visit in AD 125 (Glaser 2000:442). The nymphaeum at Jerash was dedicated to Commodus in AD 191 (or AD 190) (Welles 1938:406-7, #69). The link between Emperor and nymphaeum appears to be used for propagandistic purposes and demonstrations of loyalty to Rome. There seems to be an implicit connection between the Emperor and the life-giving waters in the East, as well as elsewhere in the empire. Presumably, the governors who donated fountains similarly wished to be viewed as magnanimous providers, as shown by the public display of their benefaction. An inscription from a fountain in Jerash records that Attidius, the consul designatus, set up fountains in AD 150. Finally, according to Josephus (BJ 1.21.11) Herod gave costly fountains (κρῆναι πολεθτελεῖς) to Ashqelon. It must not be forgotten, however, that as well as being impressive focal points, nymphaea would also have been used as main water-drawing points in the city. Indeed, we know from Josephus that when the fountain of Siloam in Jerusalem failed, prior to the arrival of Titus, water had to be sold in amphorae (Josephus BJ 5.410; Adan 1979:100). In total, 22 structures claimed as nymphaea across 17 sites have been recorded across the region (Tables 11 and 12). It is more likely that some of these structures, such as the Amman example, should be classed as kalybe, which has no equivalent in other parts of the Roman Empire (Ball 2000:291-2). The kalybe seems to be related to the nymphaeum, but was solely for display with no water provision element. This type of structure has also been identified at Bosra and Shohba (Figure 14). Therefore, 17 structures across 13 sites have been identified convincingly as nymphaea. The sites were distributed evenly across the region and no regional bias could be discerned. In contrast, 8 of the 12 dated nymphaea were constructed in the 2nd century AD. The construction dates must suggest that nymphaea were a Roman introduction to the region. The construction peak in the 2nd century AD must be related directly to the peak in aqueduct construction in the same century. It also suggests that the euergetic culture intensified during this period, which would form part of a pattern of increased monumental construction, for example colonnaded streets and gates, in eastern cities from the

114 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST end of the 1st century AD (Segal 1997:3). A similar boom in nymphaeum and fountain building occurred in 2nd-century Greece and Asia Minor (Glaser 2000). The lack of nymphaeum building in the East in the following centuries reflects a general downturn in building after the Severans.

Figure 14: View of the kalybe in the forum at Shohba/Philippopolis.

CHAPTER THREE

Date

Site name Antioch

3rd

Antioch (Daphne) Apamea Apamea Banias

2nd century 1st century AD

Beirut Bosra

late 2nd century

Caesarea

1st century AD

Gadara Jerash

mid 2nd century late 2nd century

Kanawat Kanata Pella Petra Petra Scythopolis

Sidon Suweida

century

115

Reference Will 1997:103; Lassus 1983:217; Leblanc and Poccardi 1999:100; Lassus and Stillwell 1972:44-47. Leblanc and Poccardi 1999:120 fig. 13. Balty 1983:259; Balty 1987:20. Balty 1987. Tsaferis and Avner 1989/90:3; Tsaferis and Israeli 1995:5. Saghieh-Beydoun et al. 1998-9:114. Segal 1997:157; Mougdad and Makowski 1983:41; Dentzer et al. 2002, 116-125. Porath 2002b:122; Porath et al. 1998:45; Porath 1996:112-3. Segal 1997:154-5; Weber 1988:349 Segal 1997:160-2; Harding 1967:95; Horsfield 1926:2; Fisher 1938:21, 54; Seigne 2008. Segal 1997:152-3.

Smith 1969:6; Dvorjetski 20012:503. Segal 1997:164; Bachmann et al. 1921:34. Bachmann et al. 1921:36. 2nd century Segal 1997:157-160; Foerster and (renovated after Tsafrir 1988-1989:18-21, figs 16mid 4th century) 18; Tsafrir and Foerster, 1989/90:122; Bar Nathan and Mazor 1992:36; Tsafrir and Foerster 1997:109, 113. nd 2 century Dunand 1967:41. early 2nd cen- Segal 1997:154; Brunnow and tury Domeszewski 1904:90.

Table 11: Nymphaea in the East.

116 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST Site name

Na’arran

Zeugma

Comments

Reference

2nd half of Barrel-vault over 2nd century stream Possibly not a nymphaeum

Amman

Susita Hippos

Date

Segal 1997:162-4; Harding 1967:68; Northedge 1992:58; Almagro 1983:631 figs 25-27; Waheeb and Zu’bi 1995. 3rd century Not a nymphaeum Dauphin and Schonaccording to later field 1983:197-204; interpretation Dauphin and Gibson 1991:177. Underground cis- Segal 1997:153-4; tern. Probably not a Ben David 2002:205nymphaeum 6; Segal 2000:13*. 2 doubtful nymphaea Early et al. 2003:15, 49-50.

Table 12: Doubtful nymphaea in the East.

The nymphaea at these sites, however, were on a notably less grand scale than those in Asia Minor, for example the elaborate and costly façade nymphaeum at Miletus. It seems possible that this difference in scale may be linked to a cultural difference in the perception of water. There does not appear to have been a problem with ostentatious architectural display. Procopius of Gaza (c. AD 500), for example, describes a water clock at Gaza that was crowned by an eagle and had 12 doors, out of a different one of which Herakles appeared every hour to perform one of his labours (Wilson 1983:31; Sprague de Camp 1963:257). The building (c. 8 ft wide and 19 ft high), which had a marble barrier with spikes at the front, marble pillars at the front and several large statues, was highly decorative and elaborate, not dissimilar in fact to a façade nymphaeum. The display element did not, however, come from the water itself (though it was powered by water). As noted above, the kalybe was like a variation of the nymphaeum without water, i.e. it fulfilled the same display functions, but did not make ostentatious use of water. Is it possible then that nymphaea were not maybe as elaborate as one might expect because large water displays seemed overly ostentatious and wasteful in a semi-arid climate?

CHAPTER THREE

117

This might explain why they seem to be missing from some sites from whose architecture one might expect such a grandiose display, such as Palmyra. Where they are found, the historical background of most of the sites appears to provide the reason why they chose to put concerns for water conservation aside and construct nymphaea regardless. Three of the cities were regional capitals: Bosra, Petra and Antioch. Five of the sites were Decapolis cities (Kanata, Gadara (Umm Qes), Scythopolis, Jerash and Pella), which have strong Hellenistic leanings. Significantly, recent work on the fountains at Jerash has demonstrated that the four fountains in the north tetrapylon were never connected to the water supply (Seigne 2008:37). So even in a city that did have a nymphaeum, water was still not used for display at all times and venues. Of the other sites Apamea, Caesarea and Beirut are renowned for their Roman-style architecture and propensity to take on Roman-style monuments. That is, there seems to have been a conscious decision in these cities to use water lavishly in order to identify themselves with the rest of the Greco-Roman world, which suggests that the concern for not making lavish use of water reflects a more Near Eastern mentality. The renovation of the nymphaeum at Scythopolis after the mid 4th century, when other buildings such as the basilica were not maintained (Tsafrir and Foerster 1997:113), suggests that these associations could be long-term and transcend other shifts in priorities. Bathhouses The use of water in bathhouses has often been cited as an example of the luxurious use of water in the Roman world, particularly in urban centres. While bathhouses did, eventually, become popular across all parts of the Near East, about a century later than in other parts of the empire, there is evidence from the numbers of bathhouses and their water management strategies to suggest that they did not make use of such large amounts of water as other areas of the empire may have done. At least eleven sites in the Near East had multiple bathhouses (Table 13). While this shows that multiple bathhouses were not uncommon, in comparison with other Mediterranean sites these are not high numbers; for example 17 are known from Ostia, 13 from Timgad, 8 from Pompeii, 7 from Athens, 7 from Ephesos and 5 from Olympia (Nielsen 1993, Vol. I:38, 93, 96; Vol. II:32, 34). The sites from the East seen to fall into one or both of the

118 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST following categories: those with strong Roman influence (e.g. Antioch, Caesarea, Apamea, Bosra, Philippopolis) or strong military associations (e.g. Bosra, Dura Europos, Athis). At Dura Europos it has been argued convincingly that the baths were almost solely for the use of the military (Pollard 2004:132-143). This is strikingly similar in pattern to that suggested above for the distribution of nymphaea. While bathhouses were supplied by aqueducts in the Near East, one of the most conspicuous features of the water management in Near Eastern bathhouses was their use of a wider variety of sources for supply, such as wells and cisterns, and concomitantly the installation of small basins and tubs, rather than pools. In the late 1st century BC to 1st century AD, when all bathhouses (with one possible exception of a public bathhouse in Petra) were private and mostly associated with Herod’s palaces, small rainwater cisterns seem to have fed the bathhouses at Masada (Netzer 1991: 76101, 124-5, 164-9, 251-263) and Jericho (Pritchard 1958:9-10). In the 2nd and 3rd centuries AD, cisterns probably fed the baths at En Gedi (Mazar and Dunayevsky 1967:142-3) and Dura Europos (Brown 1936:49-77; Nielsen 1993, vol. II: 44; Perkins 1973:27; Pollard 2004:132-143), both of which had either very small pools or small basins for washing. In the 4th to 7th centuries AD cisterns provided water for bathhouses at Serjilla (Butler et al. 1907 II.B.118: 121 ills 134-6; Nielsen 1993, vol. II: 46), Ashqelon (Dauphin 1996:47-72; Stager 1991:45-47; Israel 1993:104) Rama (Tsaferis 1980; Nielsen 1993, vol. II: 41) and Andarin, where a well with a saqiya also formed part of the supply (Mango 2009, 73). Rainwater formed at least part of the supply at the Scythopolis bathhouse in this period as a covered channel collected rainwater from gutters (Foerster and Tsafrir 1992:38-41). This period also has an example of a bathhouse fed solely by a well at Zenobia (Lauffray 1991:125; Orssaud 1991:248). Public baths fed by rainwater cisterns are very unusual and comparanda are difficult to find. Only one public bathhouse in North Africa, at Tiddis, seems to have been supplied primarily by cisterns (Wilson 1997:128). The use of rainwater cisterns as the only water supply must have constrained the usage of bathhouses, as year-round supply could not be guaranteed. Bathhouses fed by wells were also unusual and are known again only from North Africa: Sabratha (unpublished theatre baths), Banasa (Fresco Baths) and Bu Njem (military baths) (Wilson 1997:127-

CHAPTER THREE

119

128). In some cases, the location of the bathhouse may explain its use of cistern water. Due to topographic difficulties, Dura Europos, for example, had to rely on cistern supplies. Also Hazeva and Yotvata were in the Negev, which may have meant that again cisterns or wells were the only viable water resources. In other cases, cultural forces may have been at work, for example Serjilla, Rama and Brad probably did not have the associations with Rome that were illustrated at Jerash, Bosra or Antioch. This may suggest that bathhouses were not such an integral part of life that their continual use was deemed necessary. Available resources may also have been a governing factor as sites, such as Serjilla and Brad, though commonly referred to as the ‘Dead Cities’ were more akin to small towns or even large villages and therefore may not have been able to raise the capital to construct an aqueduct. In addition, if there was no other pressing need for an aqueduct, it may have seemed overly luxurious to build one solely for the use of a bathhouse. Site

Jerash

No. of baths 5 (c. 20 from literary evidence) 6

Bosra

4

Dura Europos

4

Apamea Caesarea

3 3

Scythopolis

2

Antioch

References Fisher 1934:5-31; Nielsen 1993: vol. 2, 45; Campbell 1936:1; Levi 1947:260-261; Stillwell 1941:8-9. Browning 1982:165, 168; Harding 1967:86, 91, 99-100; Nielsen 1993:vol. 2, 41; Fisher 1938b:265269; Lepaon 2008. Butler and Prentice 1901:II.A.4, 260, 264-5; Nielsen 1993: vol. 2, 3. Brown 1936: 49-77, 86-100; Nielsen 1993:vol. 2, 44; Perkins 1973:27-8; Pollard 2004:132-143. Balty 1981:53; personal observation. Levine and Netzer 1986:149-160; Hoss 2005:45; Horton 1996:177189; Porath et al. 1998:42. Mazor 1988-9:22-24; Foerster and Tsafrir 1992:38-41; Mazor and BarNathan 1998:20.

120 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST Site Umm Qes

No. of baths 2

Sepphoris Jerusalem

2 2

Athis

2

Philippopolis 2

References Lux 1966:64-73; Andersen et al. 1993; Holm-Nielsen et al. 1986:220229; Nielsen 1993:vol. 2, 41. Weiss and Netzer 1994:34, 43; Baruch 2002:75*-76*; Reich and Shukron 1994:95. Harper and Wilkinson 1975:329; Nielsen 1993:vol. 2, 46. Dumond-Maridat 2008:82.

Table 13: Sites with mulitple bathhouses in the Near East (not including Herodian palaces).

The use of water in domestic contexts: status and display Recent work on Pompeii, in particular, has focussed on the use of water for display in houses (Jashemski 1996; De Haan 2001; Jansen 2001; Sear 2004; Jones and Robinson 2005). Until recently it was thought that with the advent of a piped supply in the Augustan period, several houses developed innovative methods of display involving water. After the introduction of the Augustan aqueduct, gardens, for example in the House of Polybius, were decorated with fountains, pools and plants that required more water (Jashemski 1996:53). Inside houses, there was a clear link between the availability of a sufficient water supply system and the building and use of private baths in 40-20 BC. In particular, large bath suites (with separate rooms for the apodyterium, tepidarium, caldarium and frigidarium) located next to luxurious, public areas of the house were only constructed after this time (De Haan 2001: 42, 46-7). Ohlig’s work on the castellum divisorium at Pompeii, however, has shown that Pompeii had an earlier (possibly Sullan) aqueduct (Ohlig 2001; also see Wilson 2006a). On being connected to the Serino aqueduct network in the Augustan period, Pompeii’s water supply was actually reduced, which casts doubt on a simplistic link between the Augustan aqueduct and water display in Pompeiian houses. It is possible that the Augustan aqueduct did, however, provide a more seasonally reliable supply, which may explain the observed pattern. Regardless of these complications, Jansen (2001:37) has illustrated that the piped supply in Pompeian houses

CHAPTER THREE

121

was used mainly to feed fountains in atria and courtyards. Work areas of houses, i.e. kitchens and latrines, were very rarely connected to the piped supply.29 This suggests that piped water was used primarily as a tool for display and ostentation. This has also been argued for North Africa where the installation of a piped water system was not to obtain a better drinking supply, but to indicate wealth and social position (Wilson 1995). In his description of Antioch, Libanius (Orat. 11.244-248), in a passage strikingly similar to Strabo (5.3.8) describing Rome, makes analogous claims: Ἕξεστι δὲ τὸν μὲν τῶν πηγῶν πλοῦτον τῷ πληθει τῶν οἰκιῶν σκοπειν, ὄσαι γὰρ οἰκίαι, τοσαῦται κρῆναι, μᾶλλον δὲ καθ'ἑκάστην πολλαι, καὶ τῶν γε ἐργαστηρίων τὰ πολλὰ τούτῳ φαιδρύνεται...ἡμῖν δὲ διὰ τὸ εἴσω θυρῶν ἑκαστοις εἰναι κρήνην αἱ κοιναὶ ἐπίδειζιν ῥέουσι. ‘One can judge the wealth of our waters by the number of houses, (sc. in the city), since there are as many fountains as there are houses, or rather there are many fountains in each house, and indeed the majority of the workshops are also adorned in this way…With us, since everyone has a fountain within his house, the public fountains flow merely for display.’

While this seems like good evidence for a similar situation in the East as in other parts of the empire, the archaeological evidence for domestic water supply (largely from high class houses in Antioch, Zeugma, Dura Europos and Sbeiteh, reflecting the present state of excavation and research) suggests that Libanius was actually talking about higher class housing in Antioch and overstating the number of houses that had fountains. Most of the houses in the East had relatively simple water management systems that focussed on supply from cisterns and probably wells. Only the wealthier houses in cities such as Antioch and Zeugma appear to have been connected to an external piped supply. It was in these houses that additional features, in particular fountains, made use of the piped water to display the wealth and status of the inhabitants. The desire to use water for display was, in defence of Libanius, particularly 29

Also see Sear 2004 on the Casa del Granduca.

122 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST strong in Antioch and where there was no connection to a piped supply, pools were used for ostentation and as status markers (see below). The use of domestic, as well as public, fountains for display is illustrated by their locations within the houses. Many of the fountains were in courtyards and peristyles, with one example in a triclinium in Antioch- Seleucia Upper City (Stillwell 1941:5). These would have been highly visible and public areas of the house. The importance of fine views of fountains in peristyles and triclinia has also been noted in North Africa (Wilson 2001a:92). In Pompeii, the power of fountains to display wealth and social position was taken a step further as they were often placed so as to be visible to passers-by in the street (Jones and Robinson 2005:700). It does also seem to be true that more houses in Antioch than in other cities were furnished with fountains (see Table 14). This pattern is maintained even when compared with other sites that had relatively high numbers of excavated houses (Sbeiteh, Dura Europos and Zeugma). No fountains have been found in the houses at Dura Europos and Sbeiteh, which makes a striking contrast to Antioch and (to a lesser extent) Zeugma. In the case of Dura Europos, which had to rely on cistern supply due to its topographic location, a piped, pressurised supply necessary for spouting fountains would not have been available. Basin Pool Fountain Baths Latrine Antioch 3 13 14 2 2 Beirut 1 1 2 1 Zeugma 6 3 1 Dura 3 1 2 Europos Petra 2 2 1 Apamea 1 1 2 Sepphoris 1 1 Caesarea 1 Kurnub 1 Table 14: Luxurious or decorative water features in urban houses across all periods.

It is arguable, though that Dura Europos could have supported basins and pools, but with the exception of the Palace of

CHAPTER THREE

123

the Dux Ripae at Dura Europos, which is responsible for all of the features attributed to the site in Table 14, no such features have been excavated at this site. It is possible, then, that as well as there being a general paucity of water on this site, there was also less willingness to use precious water in a frivolous manner. In the Palace of the Dux Ripae display seems to have been restricted to three brick basins and to the bath suites. These display elements were in many ways restrained in such a large and important house, in contrast to the Yakto Complex at Antioch, for example (Levi 1947:279-280; Lassus 1938:99, 110, 121-2, 130-3). Also, the fact that two bath suites were installed as a priority over fountains may suggest that by the 3rd century (when public bathhouses were flourishing across the East) these were almost deemed a necessity by the owner. The presence of two latrines also points to a degree of luxury, as has also been suggested for residential district 1 in the terraced houses in Ephesos (Wiplinger 2004-6, 41). A similar situation may have pertained at Sbeiteh. Although the site was supplied by an aqueduct, this water must have been highly prized in such a badly-watered landscape. Again the conspicuous absence of ostentatious domestic water features suggests that decorative uses of water were not high on the list of priorities for Sbeiteh’s inhabitants. These examples may suggest a more reserved and cautious attitude towards water in areas where water was difficult to obtain. Zeugma also shows some interesting and nuanced patterns in relation to water for display (Abadie-Reynal 2008). Like Antioch the water features were placed in the most visible parts of the houses and sometimes in places that could be seen from the street. The visual effect of these features was amplified by the decoration used in their vicinity, for example, the nymphaeum in the House of Poseidon was accompanied by a mosaic depicting Poseidon, Oceanos, Tethys, marine animals, dolphins, fish and crustaceans. These effects may have been further enhanced in some houses by the use of rotary jets evidenced by two small bronze discs that were found on the site. Not all houses, however, made lavish use of water. Rather than making use of actual water, some houses made use of décor that represented watery themes. The richly-decorated House of the Synaristosai, for example, had an apse painted to look like the marble used in a fountain (room 8), a mosaic depicting 42 species of fish in the triclinium and other marine depictions in room

124 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST 17. Furthermore, the basins in the houses were often relatively small and only c. 0.15 m – 0.2 m deep, which points to small volumes of water and may even suggest that they were not used continuously. This suggests that while watery associations did create an impression of wealth and luxury, it was not always deemed necessary or preferential to use actual water. Attitudes towards housing also may explain why there were more fountains in Antiochene houses. In general, houses in the East were introverted and the entrance was positioned so as to avoid the public gaze (Butcher 2003:302-3). This is in clear contrast to Italian housing. It seems plausible, then, that the low numbers of display features were also related to the perceived privacy of the house i.e. the house was not an area regularly used for public display. To use one’s house for display then may have been a conscious attempt to behave in a Roman manner. Although pools were used in houses that were not connected to the piped supply, such as the early phase of the House of Iphigenia in Antioch, and so may be seen as holding a lower display value, there were still more in Antioch than elsewhere (see Table 14) (Stillwell 1941:10; Levi 1947:119).30 This suggests that while they may not have been as high status as fountains, they were still associated with ostentation and display. This is confirmed by the fact that they were often in the same rooms as the fountains and were, therefore, also visible elements. Similarly, basins were located in courtyards, peristyles and bath suites within the houses.

30 On the idea of ‘passive’ pools see Jones and Robinson 2005:705 with reference to the House of the Vestals, Pompeii.

CHAPTER THREE

1st BC 1st AD 2nd AD 3rd AD 1st BC –3rd AD 4th AD 5th AD 4th AD – 5th AD

125

Basin 1 2

Pool ?2 1 1 (?3) 2

Fountain 1 2 2 4

Baths 1 3 1

Latrine 1 3 -

1 2 2

1 1 1

4 3

1 -

1 3

Table 15: Luxurious or decorative water features in urban houses by century (not including unphased houses).

The chronological developments of water usage for ostentation are not so clear in the East as in, for example, Pompeii, but there does seem to have been a slight increase in decorative water features over time (Table 15). There was some evidence for individual houses changing their use of water in response to a change in supply at the Houseof Iphigenia, Antioch where fountains were added in the 3rd century AD (Stillwell 1941:10; Levi 1947:119) and in Caesarea, Insula W2S3 where the well was backfilled after connection to the municipal supply (Porath 2000:37*). At Zeugma, Abadie-Reynal (2008:106-7) suggests that the addition of water features in later periods reflects an occidentalisation of the peristyle at the site. The installation of private baths also seems to have become more popular alongside the increase in numbers of public baths. An increase in private baths in the 3rd century is also paralleled in North Africa where it seems to be a consequence of elites distancing themselves from public life (Wilson 1997:133). Interestingly, private latrines became popular earlier than public ones. This may be because a private latrine would not have posed the issues of modesty that seem to have discouraged public latrine use (see chapter 4). The numbers of private latrines were low, however, in comparison to Ephesos, for example, where there were at least three latrines in residential districts 1 and 2 in the terraced houses (Wiplinger 2004-6:38-45).

126 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST There seem to have been relatively strong differences between status and display in urban and rural housing. The first is a general lack of a running water supply for rural housing. Houses in towns simply (though not without some expense) needed to be connected to the existing urban network in order to make decorative features feasible. In rural areas one would need to build an entire aqueduct, at considerably greater expense. Rural housing also lacked pools, basins, fountains, wells or latrines in many of the examples. While this may indicate a lower standard of living or lower amount of disposable wealth among the non-elite in rural areas than in urban centres and shows a potential division between rural and urban communities in the Near East, it may also be that the archaeological record is more evenly balanced between elite and non-elite housing in rural areas. If a more representative sample of housing from a range of classes were known from urban areas, a similar pattern may occur, i.e. a higher proportion of houses may have had only basic facilities. The house at En Ya’el, near Jerusalem is clearly different from the other known rural houses. This house does not stand out just in terms of its water supply, but also in its mosaic decoration, which has been described as displaying an ‘intermingling of eastern and western iconography’ (Roussin 1994:41). Furthermore, it has been suggested on the basis of roof tiles stamped with legio X Fretensis that the villa may have been a soldier’s residence or at the very least had a close connection with the Roman army (Edelstein 1990:40). Tiles stamped with LEG X FRET have also been found in a 2nd/3rd-century AD private bathhouse at Ramat Rahel outside Jerusalem, near the High Level aqueduct (Aharoni 1962:26).

CONCLUSIONS Drawing together these varied sets of evidence from dams, storage facilities, nymphaea, bathhouses and private houses reveals several facets to attitudes towards water conservation in the Roman and late Roman Near East. The default position of the everyday, reasonably wealthy inhabitant of the Near East seems to have been to conserve water and not use it for luxurious, and arguably wasteful, status display as shown by the majority of the evidence from the private housing. The obvious exceptions to this rule are the inhabitants of the provincial capital in Antioch where their close association with Rome (both actual and desired) led to a more ‘frivolous’

CHAPTER THREE

127

use of water in the private domain. This pull of Rome and of wanting to behave in a Roman manner is more prevalent in the public than the private sphere of display through water as evidenced by the more widespread take up of nymphaea and bathhouses. This suggests that there was more resistance to change privately than was expressed publicly and also possibly at higher levels of society (maybe also due to more disposable wealth). The evidence from the public facilities is not ‘black and white’, however, but still shows an underlying concern with water conservation outside the strongly Roman centres. In these areas, innovative responses to the problem are created in the form of water-less kalybe structures to take the place of lavish fountains and bathhouses with smaller pools and basins that are better suited to the lower volumes supplied by cisterns and wells. The bathhouse evidence also hints at the low perceived need for such Roman ways of living; they can be used as added extras when the supply allows, but need not be supplied as a priority. This idea of balancing what one has with what one needs is also found in the technological solutions provided for urban supply and storage. Rather than conforming to the general Roman ‘ideal’ of a continually flowing system, which would not be practical in the Near Eastern climate and regime, more emphasis is placed on dams and complementary water sources and storage facilities. This response was increased in times of stress, for example in the late Roman period, which saw serious social changes, military challenges and natural disasters. Water conservation was not, then, undertaken necessarily only for its own sake, with the overtones of ecological and environmental mindfulness that it may be associated with in the modern western world, but was also motivated by, and indicative of, political, ideological and cultural positions and sensibilities.

4 WATER, HYGIENE, CLEANLINESS AND PURITY

In this chapter the extent of hygiene and cleanliness in water systems in the Roman Near East will be critically assessed. Concepts such as these are clearly loaded with modern preconceptions and standards of what it means to us for something to be classed as clean or hygienic, a concept which can change radically over time (Ashenburg 2008). In particular, the idea that water is always considered as a purifying, life-giving resource has been challenged recently (Strang 2004; Kamash 2008). Instead, water can often have multivalent meanings, and complex associations with life, death, pollution and cleanliness. The discussion here will focus on what past attitudes to water and cleanliness may have been, using a variety of archaeological, epigraphic and art historical evidence. I will begin by looking at how waste water was dealt with, and what this might tell us about levels of hygiene and, most importantly, attitudes towards it. This will consider drainage systems in cities and from private houses. The second part of the chapter focuses on how water was used and perceived in public bathhouses and latrines, starting with a discussion of Jewish stepped pools (miqva'ot). Public bathhouses and latrines are often held up as ‘civilising’ effects of the Roman colonial effort (eg Robinson 1992:113118). We need to consider here to what extent this was actually the case, how welcome an introduction they were, how they were perceived in terms of hygiene, cleanliness and purity, and how such responses might be related to identity.

DRAINAGE SYSTEMS The various elements of the drainage networks found in urban sites and private houses have been categorized here according to a five129

130 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST order classification (Table 16). This classification is based on the four-order classification of Bodon et al. (1994:391), but sees the division of the third order into two. The following discussion will consider the classification from the smallest (1st and 2nd order) to the largest (5th-order collectors), starting with a brief case study from Dor, which shows how the different elements of the system could work together. Order Description 1 Drain within a building at the entry point to a system that leads waste water away; for example: gutters from roofs; channels from latrines. 2 Drain that combines 1st-order drains and forms a single exit from a building. 3 Drain that combines 1st or 2nd-order drains and runs under minor streets and alleyways. 4 Principal drain that combines 1st, 2nd or 3rd-order drains and runs under principal streets. 5 Large collector that can combine all previous orders (usually 4th) and discharges outside settlement. Table 16: Orders of drainage.

The quantity of excavation and publication at Dor means that we can reconstruct how the drainage systems may have worked across this site as a whole (Berg et al. 2002:164-7). This case study also suggests that there was not always concern about the disruption caused by putting in drains and changing the drainage systems after laying out the streets (see Hodge 1992:340). Drains were found under all the excavated streets, and nearly every house appears to have been connected to the system. The water collected on the house roofs flowed down ceramic downpipes (1st and 2nd order). These fed 3rd-order drains in minor streets, which were usually 0.1 m - 0.2 m wide and deep, built from masonry with plastered bottoms and capped with stone. In turn, these 3rd-order drains fed 4th-order drains beneath major streets. These drains were 0.4 m wide and 0.4 m – 0.8 m deep, constructed from fieldstones and ashlar and capped by street flagstones. Finally, the vaulted 5th-order drains, ‘large enough for men to walk through’, were fed by the 4thorder drains and discharged outside the settlement (at an unspecified location). It has been proven stratigraphically that the construction of the drainage system at Dor actually preceded the aque-

CHAPTER FOUR

131

duct; this was also the case at Timgad in North Africa, but is otherwise deemed rare across the empire. In addition, the drainage system went out of use before the aqueduct did. In the period prior to the aqueduct, 3rd-order drains fed by roof gutters discharged into local sumps; it was not until the next phase (AD 138-230) that Street I had a 4th-order drain installed. It appears, therefore, that at least at Dor drainage systems existed before water was supplied by aqueduct, but became more sophisticated with the provision of aqueducts. At other sites only three urban private houses have been recorded in sufficient detail to make it possible to reconstruct their 1st and 2nd-order drainage systems: the palace of the Dux Ripae in Dura Europos (the supposed palace of the 'Roman commander of the riverbank') and in Antioch the Yakto complex and S-18-K. These houses were large complexes, and so it is possible that smaller houses did not also use both 1st and 2nd-order drains. In the palace of the Dux Ripae 1st-order drains, sometimes made of stone, led from the basins and latrines and fed into 2nd-order drains (Detweiler 1952). In the Yakto complex small 1st-order drains, 0.08 m to 0.09 m wide, were collected by a 2nd-order drain in a neighbouring room. Another drain also ran under a basin; its larger size (0.4 m wide x 0.75 m deep) suggests that it was a 2nd-order drain that also undertook some of the functions of a 1st-order drain (Lassus 1938). Finally, 1st-order drains in Antioch S-18-K originated in the triclinium and were collected in a 2nd-order drain along the southern edge of the eastern colonnade; another 2nd-order drain was also found in the northern colonnade (Stillwell 1941:33). Triclinia needed drains in order to remove wastewater from washing down the floors that had been strewn with food debris. Other triclinia provided with drains are known from North Africa, for example the Maison de Dionysos et des Quatre Saisons, Volubilis (Etienne 1954:65; Wilson 2002a:481). Khirbet al-Mureq, Khirbet Mansur alAqab and Ramat Hanadiv were the only rural houses with recorded provision for drainage (Damati 1972:173; Hirschfeld and Birger 1991:93, 103; Hirschfeld 2000:27). Khirbet Mansur al-Aqab was furnished with gutters in the paved courtyard in the western wing and in the stabling area. Ramat Hanadiv had two plastered drains, one draining the tower area and one draining the possible stable area.

132 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST What is not clear from these examples, with the single exception of Dor, is whether these drains, especially from urban houses, fed into the wider urban drainage system (i.e. into 3rd to 5th-order drains). It is possible that they drained into cess pools or discharged into the streets. This has obvious consequences for any consideration of hygiene, but more detailed recording needs to be made at future excavations for firm conclusions to be formed. What is clear though is that some effort was made, in larger houses, to move waste water away from the interior of the house. One of the Scythopolis latrine drains (in the West Baths) and the drain from the latrine in Caesarea Area CC were connected to the main external drainage system, though latrines may have been subject to different considerations as they would have been carrying sewage as well as waste water (Mazor and Bar-Nathan 1998:23; Porath et al. 1998:55). Indeed, latrine drains could also be constructed in slightly different ways to other drains. Wilson (2000b:169) suggests that drains, with particular reference to ‘large drains’, were not lined with waterproof mortar and sometimes had an unpaved floor in order to enable water percolation into the soil. While this is generally true, there are exceptions where percolation into the soil may not be desired, for example the latrine (i.e. 2nd-order) drains were lined at Zeugma and Dura Europos (Brown 1936:71; Early and Humphrey 2003:56). The depth of latrine drains also seems to have been regulated according to the amount of waste they would have to handle. While most were under 1 m deep, the drains at the Atrium latrine in Apamea and Scythopolis East Baths were 2 m and 2.5 m deep respectively (Balty 1981:76; Schmidt-Colinet 1984:1412; Mazor and Bar-Nathan 1998:14). Presumably the greater depth of these examples was due to the higher numbers of people using these facilities: the Apamea example could seat between 80 and 90 people, and is one of the largest known in the empire, and Scythopolis could seat approximately 57 people. This is significantly higher than the average number of seats in latrines, which tends to have varied from 12 to 15. As the Ramat Hanadiv and Dor examples show, however, it was not just latrine drains that were lined. The underlying geology, for example on a basalt site, may cause the development of unwanted cesspools. In addition, there seems to be good evidence from levels data to suggest that the 3rd-order drainage system out-

CHAPTER FOUR

133

side a building at Rujm al-Malfuf was designed to take groundwater away from the foundations (Boraas 1971:36-7). This drain was plastered, presumably to facilitate the removal of water; percolation in this instance would negate the primary purpose of the drain. Another possible scenario, especially in an arid zone, is that the water may have a further purpose outside the settlement. The reuse of water from drains has been noted as a striking feature of Roman drainage (Wilson 1997:217). It was proposed, for example, that the 5th-order collector in the via Praetoria at Auara discharged into the fields (Oleson et al. 1999:420). There also seems here to be a concern with making the best use of the available water (see Chapter 3). The final destination of waste water can be traced if the location and route of 5th-order collectors are known, as at Caesarea (three), Dor (two), Jerash, Zeugma, Jerusalem, Auara and Horbat Castra (Harding 1967:94; Wiemken and Holum 1981:34-8; Gawlikowski 1986:110; Stern and Sharon 1987:209-11; Stern and Gilboa 1989-1990:117; Stern et al. 1991:49; Abadie-Reynal and Ergec 1997:392, figs 13-15; Porath et al. 1998:42, 55; Reich and Billig 1998:90; Oleson et al. 1999:420; Yeivin and Finkielsztejn 1999:23*; Abadie-Reynal 2001:251; Berg et al. 2002:164-7). In general 5thorder drains were substantially deeper than any other drains on site, ranging in depth from 1.05 m to 4 m. The only exception to this is the example from Auara that measured only 0.3 m wide x 0.25 m deep, which may be explained by the smaller settlement size being drained (Oleson et al. 1999:420). This order of drain was usually stone-built and vaulted or capped with stone slabs; one exception at Jerusalem was rock-cut and capped with stones (Reich and Billig 1998:90). These large collectors discharged into a variety of contexts. The Zeugma example discharged into the Euphrates, Caesarea into the sea, Horbat Castra into a cesspool at the edge of the settlement and Auara possibly into the fields (Oleson et al. 1999:420; Porath et al. 1998:42; Yeivin and Finkielsztejn 1999:23*; Abadie-Reynal 1998:392; Abadie-Reynal 2001:251). This variety points to a flexible approach to how and where to discharge waste water, an approach that is largely dependent on the local landscape and geology. This also reflects attitudes towards water sources. The major source for aqueducts in the Near East was springs (53, ie 46%), which follows the pattern seen across the Roman Empire. This was

134 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST followed by runoff (19, ie 16%) and rivers (9, ie 8%). This shows that there were very low numbers of river-fed aqueducts. The area west of the River Jordan and the Dead Sea presents an interesting case. This area saw the highest number of aqueducts fed by runoff, in conjunction with aqueducts fed by springs, whereas no aqueducts seem to have been fed by the River Jordan, a rather surprising lacuna. While further survey may reveal an aqueduct, it seems possible that this lack of use of a significant water provider in the region may reflect concerns about purity and pollution from river water. This is understandable if rivers were deliberately chosen as an environment able to be polluted by waste water. Some sites, of course, took slightly different approaches to drainage. At Umm Qes (Gadara), for example, a street was inclined to the north so that the unwanted water ran along the sidewalk. A channel was only provided through the north tower to avoid banked-up water in front of its threshold (Kerner and Hoffmann 1993:363). In addition, at Petra a depression in the middle of the street diverted runoff water, presumably through a grating, into the drainage system under the street (Zayadine and Farajat 1991:286-8). While these techniques may seem ‘primitive’, they also represent a simple, but efficient solution to drainage issues. This is comparable to the situation at Pompeii where the pavements were raised high enough to avoid the surface water on the streets, and stepping stones were provided so that one did not have to step in the water when crossing the road (Hodge 1992:335). These examples strongly indicate that it would be inappropriate to judge concepts of Roman Near Eastern hygiene by our own modern, western standards. While walking through waste water is probably avoided by the majority of westerners, this was clearly an acceptable and deliberately chosen method of drainage in some areas of the Near East. The unwillingness to use river water for drinking provision via aqueducts, particularly in an area with minimal year-round supply sources, however, displays a concern with dirty water, and a level which cannot be transgressed easily. The plastering of some channels to move waste water to ‘acceptable’ places also gives rise to interesting nuances about where is able to be polluted and where not. This reflects some of the key themes that will be brought out in the following discussion of ritual baths, latrines and bathhouses.

CHAPTER FOUR

135

JEWISH RITUAL BATHS Jewish ritual baths (miqva’ot (pl.); miqveh (s.)) are stepped pools that are used for the purification of the body by immersion in water. A substantial bibliography has built up around these installations and their use in antiquity, particularly at Qumran, though much of this is in Hebrew.31 Miqva’ot first seem to appear in the Hasmonean palaces and have been identified at Masada (Yadin 1965:91), Jericho (Netzer 1977:figs 3, 6 and 7) and Herodium (Corbo1967:figs 7, 18 and 19; Foerster 1970:pl. 16b). Other examples have also been discovered at Samaria (Crowfoot et al. 1942:122, 132, 134-5), Sepphoris (Galor 2006), Jerusalem (Mazar 1975:146-7; Shanks 1977:21; Reich 1998), Qumran (De Vaux 1973; Wood 1984; Hidiroglu 2000; Galor 2002), near Legio (Zissu et al. 2006) and near Ramat Rahel (Meyers 2002). In many cases these pools seem to have been concentrated in residential areas at these sites (Galor 2002:4), which Miller (2007:218, 225, 228, n. 18 and 20; 2010:236-7, 241, n. 71) has suggested attests to the role of domestic life in Jewish society and also may have had an effect on whether these pools were only used for ritual purposes, or whether more mundane or even profane activities may have been carried out such as rinsing dishes and clothes or swimming (see also Galor 2006:266). An exception to this is Jerusalem where stepped pools were found in a public setting on the Temple Mount. Reich (1997:431) has claimed that after the destruction of the Temple in AD 70, there was a decline in the number of miqva’ot from two to three installations per house to one or two installations per village or neighbourhood. There is evidence, however, from literary sources that ritual purification involving immersion continued into late Antiquity (Miller 2010:2323) and the majority of the stepped pools at Sepphoris appear to have stayed in use until AD363, with two continuing into the 7th century (Galor 2006:267). There are two main types of miqva’ot that have been identified archaeologically (Hidiroglu 2000:37). The first type comprises two basins; one basin functions as the immersion pool and one for wa31 The major recent works in English are: Wood 1984; Reich 1997; Reich 1998; Hidiroglu 2000; Galor 2002; Galor 2006; Hoss 2005; Miller 2007; Miller 2010. For references to the works in Hebrew see Galor 2002.

136 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST ter storage. The second type has just one basin, the immersion pool, which is often deep (over 3 m). It appears that the first type was often used in arid areas where water storage was necessary. The capacity of a miqveh must be above 40 seahs (40 seahs is roughly equivalent to 1 m3) (Reich 1997:430). The installations identified as miqva’ot at Qumran (68, 69, 144, 138) had relatively low volumes (c. 10 m3), though two others (117, 118) that might also be miqva’ot were larger (50 m3 and 40 m3 respectively) (Hidiroglu 2000:28-9). It would not be necessary, therefore, for these installations to be filled to capacity all year round. Furthermore, it was possible to find mixtures of types on the same site. At Sepphoris, for example, only one miqveh had a reservoir of the type prescribed in rabbinic halakah. Another stepped pool was supplied by a channel leading from a nearby cistern. The remainder of the pools (c. 18) could also have been filled like this or maybe by pipes bringing rainwater from rooftops (Miller 2007; Miller 2010:236). According to Jewish law, water for miqva’ot should not be drawn. Preference was given to rainwater, though if drawn water was brought into contact with rainwater, it too became acceptable (Hidiroglu 2000:38). At Herodium and Jericho pairs of miqva’ot were found that were connected by a pipe or channel. Both miqva’ot were originally filled with pure rainwater, but only one was used for immersion. This miqveh could be re-filled using drawn water that was brought into contact with the pure water fed through the connecting pipe or channel (Reich 1997:430). Aqueduct water from a spring was also permissible. Six of the installations at Qumran were fed directly or indirectly by water channels, presumably supplying either rainwater or spring water. In addition, 19 of the 20 Hasmonean and Herodian miqva’ot at Jericho were fed by aqueduct water (Reich 1997:430). Provisions for draining miqva’ot were often not provided as it was always necessary to maintain the 40 seahs of pure water (Wood 1984:51). Two main distinguishing features can be used to identify these installations as miqva’ot rather than cisterns or other cavities. The first is the steps, which usually extend across the full width of the pool, rather than being confined to one corner as is usual in a cistern. The width of the steps points to a ritual, rather than a functional purpose; if these pools were intended for water storage, this would be an inefficient design because the steps take up a large percentage of the potential volume of the tank (Wood 1984:47);

CHAPTER FOUR

137

though as mentioned above, it is possible that some of these pools had multiple purposes. In addition, the steps provide easy access for several people. On the Temple Mount in Jerusalem there were 37 cavities of different shapes and sizes, and 2 of these (36 and 6) have been claimed as miqva’ot on the basis that they may have had steps (Reich 1998:63-4). These cavities were similar in plan, both resembling a capital T with the ‘bar’ across the top aligned eastwest. In cavity 6 the bottom of the southern extension was 4.8 m higher than the east-west aligned cavity, which suggests that the southern extension originally may have contained a staircase. This staircase may have been removed in subsequent remodelling of the cavity as a cistern. The second feature is a method of dividing those who are entering from those who are leaving. This can be in the form of double doors or more frequently low walls or partitions running down the steps (Wood 1984:49; Hidiroglu 2000:38). At Jerusalem there was just one partition running down the steps (Shanks 1977:21), but at Qumran there were two (Wood 1984:49). The extra partition at Qumran may have been used to direct water into the pool. A final method of distinguishing miqva’ot from cisterns is to analyse whether the stepped pools are located close to unstepped pools, which would suggest that they have different functions. At Jerusalem, Sepphoris and Jericho stepped pools were often found in close proximity to other unstepped plastered installations that have been identified as cisterns (Galor 2002:42). This theory, in combination with the two other identification methods, has been used to prove convincingly that the plastered installations at Qumran were a combination of ritual pools and cisterns (Wood 1984; Galor 2002). While this summary suggests that there were clear rules about the amount of water to be used and how it should be drawn, there is also evidence that the rabbis may have disagreed amongst themselves about some ritual purification practices (Miller 2010:233, 235). Furthermore, as is clear from this survey of the features of ritual pools across several sites, there were similarities, but there were also several instances where their design differed in details. These additional layers of complexity may be representative of what Miller (2010) calls 'complex common Judaism' where commonness can suggest similarities while allowing for differences and complexes and does not necessarily imply unity; a concept that will

138 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST be important in the distribution and use of latrines discussed below. Miqva’ot were predominantly private ventures rather than public. Most importantly, miqva’ot were for ritual cleansing of the body, rather than physical washing and cleaning of the body. This means that the roles and purposes of miqva’ot were distinctly different from those for Roman baths. The lack of drainage facilities would mean that the water in the miqveh was stagnant and probably very dirty. Although we refer to these installations as ‘ritual baths’, they were not used for bathing in terms of getting physically clean (Galor 2002:44): pure and impure were not equivalent to clean and dirty, which is a crucial concept when considering how water was perceived.

LATRINES With the advent of the Roman period in the western world came also the spread of public toilet facilities. Although private toilets were known before this time, the concept of public latrines did not seem to have been widely used. In contrast, public latrines became a regular feature of cityscapes in the Roman world. One of the most significant features in Roman public latrines is their communality, i.e. there were no cubicles or screens that provided any privacy to the users, who would all be sitting together. This, together with the use of water for the disposal of the waste, seems to have had a profound impact on the spread of latrine use across the Roman Empire. In large parts of the Empire, latrines reached their peak in popularity in the 2nd century AD, which coincides with the mass construction of many other types of public buildings and facilities, such as aqueducts. In the majority of Roman cities, latrines were located within or next to bathhouses. This is probably because bathhouses would have a good, reliable water supply, drawn from the city aqueduct. This water was needed to wash through the drain under the seating bench, and supply the gutter and also any hand basins or decorative fountains that may have been present. When not located in the vicinity of bathhouses, the latrines tended to be placed close to other big water sources such as monumental fountains (nymphaea). This pattern is repeated across the Near East. The general design of latrines across the empire is surprisingly homogeneous, and the majority have similar basic design elements,

CHAPTER FOUR

139

including those in the Near East (Figure 15). The facility often had an entrance that opened onto the street. Seating, in the form of wooden or stone benches, was arranged around three sides of the room. The surviving stone examples of these benches often had keyhole-shaped apertures on which each user would sit. It is thought that the shape of these apertures may have permitted easier access for cleaning oneself after going to the toilet. While most latrines could accommodate c. 12-15 users at a time, some were considerably larger, for example the 2nd-century AD latrine in Apamea, Syria, had space for c. 80-90 users. The seating benches were placed above a large drain, which seems to have increased in depth according to the number of potential users. As discussed above, this drain was plastered to prevent leaking and promote the movement of solid waste through the system. In front of the seats by the users' feet, a small gutter (c. 0.1 m wide) ran along the floor of the latrine. This gutter contained fresh water and was most probably used to rinse sponges-on-sticks (xylesphongia), which are also thought to have been communal and shared between users.

Figure 15: View of one of the latrines at Apamea, showing the large drain beneath the seats, the gutter in front of the seats and notches in the wall for fixing the seats.

140 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST The individual decorative details of latrines were more variable than the basic plan and design. Some latrines were a lot more luxurious than others, with elaborately carved seats and mosaic floors, but generally the décor seems to have been quite utilitarian. Mosaics are known from latrines in the Baths of Placcus, Jerash, in Bath E at Antioch, the East Baths in Scythopolis and in the Apamea atrium latrine (Fisher 1938b:266; Neudecker 1994:157; Balty 1981:76; Schmidt-Colinet 1984:141-9; Levi 1947:260; Campbell 1936:1; Mazor and Bar-Nathan 1998:14). The latter also had marble seating, though this was made from reused architectural fragments. A possible hand-basin was found set onto the floor of a latrine at Zeugma (Early et al. 2003:55-6). If this is a hand-basin (which may be in doubt given its low height), this would be the only recorded example in the Near East, which is in marked contrast to North Africa, Turkey and Italy where they are more common features. No fountains inside latrines are known from the Near East. Finally, the ventilation of latrines was probably quite variable. The majority of the latrines were probably roofed, but 'peristyle' latrines, such as the large one in Apamea, were unroofed. Evidence for windows in the roofed examples is quite limited and to a large extent is dependent on the height to which the walls have been preserved. The presence of lamps, for example at a latrine in Caesarea, may point to the necessity of artificial light due to the lack of windows, though it may also simply point to night-time use. It is also possible that latrines may have been placed away from the prevailing wind as noted by the excavators of a latrine at Toprak an-Narlidja (Stillwell 1941:21); it is unclear though whether this was just happy coincidence or not. In the Near East, the uptake of latrines was significantly later than in other parts of the Empire, with the majority being constructed in the 4th century AD (Table 17). Furthermore, there were particularly low numbers of urban sites with latrines in Judaea at any point during the Roman and late Roman periods; the only known examples are Caesarea Maritima and Scythopolis (Figure 16). This low frequency tends to be explained and discussed solely in terms of religious taboos over nudity and impurity that would have made using latrines problematic for religious Jews (eg Hoss 2005). There is no written evidence to support such an explicitly religious explanation for the distribution in Syria, but the slow uptake does suggest that the inhabitants of Syria found Roman-style

CHAPTER FOUR

141

latrines similarly problematic, and therefore that religious taboos, while probably playing a large role, may not have been the sole influence on the choice to use a latrine or not. The spread of the communal bathing habit across the Near East helps to tease out this pattern. Again, the peak of construction of Roman bathhouses is somewhat later in the Near East than in other parts of the Empire, which may again be tied, at least in Judaea, to religious strictures concerning nudity and graven images. In spite of this, though, bathing does seem to have become an integral part of urban daily life across the Near Eastern provinces (see below). Where latrines were introduced, however, they significantly lagged behind and were not introduced alongside bathhouses as one might have expected given that this was a general preference for locating a latrine. The fact that bathhouses were more easily incorporated into urban daily life than latrines, suggests that there were broader, more deeply embedded, sets of concerns specific to latrine use that need to be explored.

Figure 16: Map of the Near East showing the location of public latrines (grey triangles).

142 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST

Location

Date

Antioch Bath E

4th century AD

Antioch Bath F Apamea

Roman

Apamea Apamea Athis/ Dibsi Faraj – Public Athis/ Dibsi Faraj Principia Athis/ Dibsi Faraj – Principia Bosra – South Baths Bosra - Central Baths Caesarea Caesarea Dura Europos F3 Dura Europos C3

No of seats 16 - 18

References

80 - 90

Late Roman 4th century AD

Neudecker 1994:157; Balty 1981:76 figs 767; Schmidt-Colinet 1984:141-149. Personal observation. Butcher 2003:263; personal observation. Harper and Wilkinson 1975:329.

Late Roman

Harper and Wilkinson 1975:328.

Late Roman

Harper and Wilkinson 1975:328.

6th century

Dentzer et al. 2002, 97.

5th/6th century 6th century AD Late Roman pre 165 AD

Dentzer et al. 2002, 100. Horton 1996:183.

3rd century

Brown 1936a:95.

mid 2nd century AD

Levi 1947:260 pl. CXb, fig. 100; Campbell 1936:1. Stillwell 1941:8.

Porath, Raban and Patrich 1998:54. Neudecker 1994:158; Brown 1936:71.

CHAPTER FOUR Location

Date

Herod’s 2nd Palace Jericho Jekmejeh (near Antioch) Jerash – Baths of Placcus Palmyra

Herodian

No of seats

143 References Netzer 2001a:211-2.

2nd or 5th century AD

15

Stillwell 1941:25.

454 AD

12 - 15

Fisher 1938b:266.

Diocletianic

Scythopolis - West Baths Scythopolis – East Baths Scythopolis – West Baths Serjilla

6th century Late Roman Pre- 6th century 473 AD

57

Toprak en Narlidja Zeugma Zeugma

mid 5th century AD Roman Roman

10 7

Bounni and As'ad 1989:71; personal observation. Mazor and Bar-Nathan 1998:23. Mazor and Bar-Nathan 1998:14. Mazor and Bar-Nathan 1998:21. Butler et al. 1907:II.B.3, 118, 121. Stillwell 1941:19. Early et al. 2003:55-6. Early et al. 2003:56.

Table 17: Latrines in the Near East.

The history of an area and site, and its overall response to Roman colonialism, seems to have played a key role in the dissemination of the latrine concept. Scythopolis and Caesarea, the two cities in Judaea with latrines, are amongst a small number of sites in the Near Eastern provinces to display all or most of the public water supply facilities and technologies associated with a Roman way of life. Herod, a famous Romanophile, founded Caesarea as a deliberately Roman and 'Romanised' city in Judaea. It is maybe not surprising then that this is one of the cities whose public 'face' of water supply had a pronounced Roman character, eventually extending as far as the construction of public latrines. The Romanness found at Scythopolis may be related to its Greek associations.

144 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST Although it had Jewish inhabitants and neighbours, the city and, notably, its occupants wished to be viewed as Greek, as was made clear on a 2nd-century inscription which referred to the city as ‘(one of) the Greek cities of Coele Syria’ (Foerster and Tsafrir 1986-7:53; Millar 1993:6, 378; Tsafrir and Foerster 1997:87, 91). As mentioned above, it has been argued that we should think of a 'complex common Judaism' (Miller 2010), rather than a monolithic, unified Judaism. This more fluid identity may also be part of the explanation for the high level of acceptance of Roman public latrines in the city, which stood in stark contrast to other towns and cities in the vicinity. In other words, it may be that members of the Jewish community in these two cities had different understandings of what was considered pure and impure or of what impurities could be tolerated.32 While this neatly explains why latrines were eventually used in those cities, the reluctance to take them up needs further exploration and explanation. After all, if these inhabitants were so proRoman, why did they not seize the opportunity to go to the toilet in a Roman style at the same time as avidly taking on other aspects of Roman public life? This brings us back to looking behind the cultural and religious taboos in search of the motivations behind them. I suggest that two main areas may have been problematic: the use of water for the disposal for excrement and the communality of the latrines. There were clear Jewish religious strictures against the depositing of excrement in water. Rather Jewish law dictated that human excrement should be buried in a field. Latrines, therefore, posed a clear problem as they ‘flushed’ excrement into the urban drainage system. The early (private) latrine in Herod’s Second Palace seems to have presented an innovative solution to this problem. This latrine was particularly unusual as although it had a channel underneath the probable seating area, it had no connection to the bathhouse water system despite its proximity in the building (Netzer 2001:211-2). The excavators have proposed that baskets may have been placed in the channel under the seats and their contents subsequently buried. This practice of using baskets to remove excre32

On tolerated versus prohibited impurities see Miller 2010:226-7.

CHAPTER FOUR

145

ment from latrines is attested in two Syriac texts. The Chronicon Ecclesiasticum tells the story of Patriarch Athanasius I Camelarius of Constantinople who went out at night to the spot that the monks would go to out of necessity (i.e. to relieve themselves), and with nobody looking on, he carried off the waste in a basket on his back (Bar Hebraeus Chronicon Ecclesiasticum col. 263). A similar story is also told about monks in the Monastery of Qartamain (Nau 1907). While these Christian stories do not include the requirement of burying excrement, the use of baskets in both does point to an alternative method of removing human excrement without the use of a water-based latrine structure, a method which was probably widespread across the eastern provinces. In addition, there is an emphasis in both stories on the event taking place at night, without anybody watching. It is possible that this secrecy was to avoid being accused of striving for public approbation from undertaking an unpleasant task. It may also be related to a need to perform this task away from the social gaze (see below). While attitudes towards water and its associations with life and death appear to have been entangled in a complex and multifaceted set of negotiations in the ancient world, and there were concerns about the cleanliness of the water in eg bathhouses, the deliberate deposition of excrement in water seems to have been a step too far in some areas. Ethnographic evidence suggests that this may have been because the deliberate contamination of water can be viewed as the contravention of basic order (Strang 2004:77). This essentially evolutionary concern could then become a moral problem, expressed in terms of ritual purity and impurity, as the person who pollutes could be deemed a pollutant themselves. Indeed, FrymerKensky (1983:403) has demonstrated that there was a close relationship between 'danger beliefs' and 'contagious pollutions' in Judaism, to such an extent that an individual or a community could then be isolated from holiness (see also Miller 2010:225). The act of going to the toilet is one that is intimately tied up with our bodies and senses, and so too our emotions. Recent research on the so-called 'forgotten' emotion of disgust has demonstrated that certain elicitors of the emotion are found crossculturally, which may be useful when trying to envisage and evoke the experiences of people so far removed from us in cultural space and time (Philips et al. 1998; Curtis and Biran 2001). By undertaking sensual and embodied analyses of latrines, I have suggested that

146 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST there were a variety of potential disgust-elicitors drawn out through negotiations with smells, sights, sounds, tactile sensations, social taboos, notions of privacy and nudity and aesthetics (Kamash 2010a; Kamash 2010b). There would have been, for example, plenty of potential for other people to have been aware of one’s animalistic behaviour. Although one’s visual privacy may have been maintained through physical and moral barriers, the basic act of going to the toilet, along with the accompanying smells and sounds, would have been performed in the presence of, and experienced by, others. As disgust can be triggered by ideational concerns about what something is or where it has been (Haidt et al. 1997:109), sitting on the same seat and therefore leaving oneself open to contact with other people’s body products, such as urine and excrement, or even just their sweat, may have elicited feelings of disgust. Of course, other sensory cues like dampness and the smells inside the latrine may also have contributed to people viewing latrines as potentially disgusting places. In addition to the disgust elicitors outlined above, it is highly likely that shame (pudor or verecundia) would also have played a key role in determining behavioural choices as in the ancient world it governed what was appropriate behaviour. So, to behave animalistically and flaunt human social rules, was to behave shamefully. Possessing a sense of shame was also closely tied to having a sense of honour; in Jewish areas this was also associated with what was considered to be under the social gaze. It is notable, for example, that the only early latrine known in Judaea is the private one in Herod's palace, described above. This is important because, being private, the risk of bringing shame on oneself through gazing on others and being gazed on by others would have been significantly reduced. Furthermore, it meant that Herod, who was both a Romanophile and a Jew (two identities that did not sit easily together and often clashed), was still able to express his Roman sympathies and desire to be Roman, while not compromising his Jewish identity. In other words, as an individual who was acting as an individual away from the social gaze, he was able to bypass the collective choices, which serves to highlight the power of the social controls in the public arenas. This need to be outside of the gaze also seems to have been present in the Christian, Syriac stories mentioned above.

CHAPTER FOUR

147

So, it would appear that Roman public latrines triggered a variety of socio-cultural, moral and religious responses in the Near East. Where we have the evidence, these social rules and concerns seem to have focused on concepts of cleanliness and purity, particularly on awareness of oneself or others as behaving animalistically. These issues were closely interrelated with moral and religious taboos on impurity. It would appear that to break these taboos would have been to render yourself impure and potentially disgusting. Overall, this seems to have resulted in the use of latrines being severely restricted in certain times and areas of the Near East. This analysis also suggests that what impurity meant varied significantly in different parts of the Near East, at different times, and between, and even amongst, different groups. So, while it would appear that the majority of the inhabitants of the Near East felt very strongly about using latrines in the earlier part of the period, including to some extent even Herod, by the later period certain groups of people found the practice more acceptable. It is particularly interesting and important to note that this must break down any conception of Near Eastern people acting in the same ways throughout their history, and even of identifiable religious groups acting in unison, as shown by the contrast in latrine use between Scythopolis and Caesarea and the rest of Judaea.

BATHHOUSES Roman bathhouses provide a useful body of data from which to explore issues of cleanliness, hygiene and pollution. Traditionally, Roman bathhouses have been held up as paragons of hygiene, and indicative of Roman promotion of public health (eg Robinson 1992:113-118; Tsarfir and Foerster 1997:132). More recently, however, the discussion has been widened to assess how hygienic bathhouses actually were and, possibly more importantly, how clean and pure they were perceived to be (eg Kamash 2006a:202-208; Dunbabin 1989; Jansen 2000; Fagan 1999; Fagan 2000). While much of the evidence presented here points to positive attitudes towards bathing and bathhouses, it also exposes an underlying current of concern about pollution, death and disease against which one needs protection. As I have argued these methods of protection are indicative of a hazard-precaution system, which may have been called into action due to the inherently ambiguous nature of water (Kamash 2008).

148 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST In a recent architectural study of bathhouses in Roman Palestine, Hoss has shown that until c. AD 70 the bathing culture was predominantly private (Hoss 2005). This seems to continue from the Hellenistic bathing tradition, evidence for which points largely to private bathhouses, with only one possible public bathhouse, at Gezer. The public bathing culture began in the late 1st - 2nd century AD, which saw the construction of large, central city bathhouses in Scythopolis, Sepphoris (East Baths) and Jerash, which were comparable in decoration to thermae of other provinces (Mazor and BarNathan 1998:20; Weiss and Netzer 1996:34; Browning 1982:165, 168, 208; Harding 1967:86, 91, 99-100; Nielsen 1993, vol. II:41.). The reasons for this relatively late reception of the public bathing culture will be discussed further below. This culture flourished in the 3rd to early 4th centuries AD, when some of the larger cities were furnished with a second bathhouse. This phase also seemed to herald the fashion for thermal baths, for example at Hammat Gader. The number of bathhouses continued to increase into the 7th century AD (Nielsen 1993:98, 112). As a thermal spa, Hammat Gader also raises issues about healing and health and whether bathing was thought to improve the quality of life. Hygieia, goddess of hygiene, was popular in the East in connection with bathing, as well as across the rest of the Empire. Coins depicting Hygieia seated on a rock from which water is flowing are known from the Tiberias mint, in particular under Trajan (AD 99/100 and AD 108/109) and to a lesser extent under Commodus (AD 188/189) (Dvorjetski 1999: 128; Dvorjetski 20012:500). Later coins from Tiberias, under Caracalla (AD 211-217) and Elagabalus (AD 218-222) depict Hygieia alongside the god of healing, Aesculapius. In addition, intaglios and cameos from Umm Qes (Gadara), dating to the 1st or 2nd centuries AD, were devoted to Hygieia, Aesculapius and Sarapis (Henig and Whiting 1987; Dvorjetski 2001-2:506).33 Coins from Umm Qes, under Elagabalus, featured Herakles, to whom it was customary to dedicate spas. In addition, coins featuring the Three Graces are known from Umm Qes under Caracalla, Elagabalus and Gordianus (AD 238-244). The 33 A 2nd-century AD date may be preferable for these gems given the date for the rise in bathing.

CHAPTER FOUR

149

Three Graces or Χάριτες had a multifaceted relationship with water. As well as also being known as the Θαλλάσιαι ‘daughters of the sea’, which connects them with the perceived source, they were also seen as a symbol of health and are known to have been associated with Aesculapius on a relief in the Vatican (Dvorjetski 20012:501). A silver ring from Umm Qes, dating to the first half of the 3rd century, also featured the Three Graces as well as Zeus seated in a temple. Depictions of the Graces are known from mosaics in baths elsewhere, for example in Korykos, Cilicia; Concordia, Italy; Cherchel, Mauretania Caesariensis (Dunbabin 1989:21-2). The dates of these coin issues, gems and the silver ring tally comfortably with the rise of public bathing in the East, and the sites they were associated with were notably famous thermal spas. Hygieia was also invoked to bestow health on the restorer and decorator of one of the Jerusalem baths and on the founder and all who bathed in the Herkleides bath at Umm Qes (Avi-Yonah 1932:175 no. 146; Lux 1966). This bathhouse also had a mosaic that depicted this deity (Lux 1966:65). In addition, an impressive statue of Hygieia was found in Antioch. This bathhouse also featured a mosaic depicting a bust of Tethys surrounded by fishes, which were probably allusions to water and fertility. From the 4th century onwards allusions to Hygieia continued. At Andarin, for example, an inscription recorded that the bath was called Hygieia (Prentice 1922:48-9, no. 918; IGLS IV: 232-3, no. 1685; Dunbabin 1989:13). The bathhouse at Athis had a mosaic depicting Hygieia (Harper and Wilkinson 1975:329). In one of the late baths at Caesarea, a tessellated pavement wishes Χάρις ὑγία τοῖς ὧδε ‘Grace (and) health to those here’ (Lehmann and Holum 2000:105 no. 97). Although, given the conversion to Christianity, Hygieia (and ‘grace’) must be a concept rather than a deity in this period, the general association of bathhouses with well-being still holds. Other personified concepts connected with well-being are known from the 4th century onwards, such as a personification of Apolausis on a mosaic in Toprak an-Narlidja (Dunbabin 1989:20). Another mosaic in this bathhouse was inscribed to the personification of Soteria, which seems to allude to the healing powers of the baths (see below) (Stillweel 1941). Wishes for apolausis were also made in the bathhouse at Ashqelon, where a tub was inscribed with the (fragmentary) exhortation: εἴσελθε ἀπόλαυσον καὶ ('Enter, enjoy and…’) (Stager 1991:45).

150 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST Images from Bath E in Antioch of Ge reclining on a sphinx, accompanied by the karpoi (as named on the mosaic) bringing a garland of wheat as well as images of rivers (Eurotas and Lacedaimonia) were strongly associated with fertility and the gifts of prosperity and joy that water brings (Dunbabin 1989:29-30). The 5thcentury AD poem to the Empress Eudocia from Area D in Hammat Gader combined many of these themes (Green and Tsafrir 1982). The ocean was described as being the life source that brings about sweet streams and springs (see above on the Three Graces coins). In addition, the River Clibanus, the source of the hot stream, was called Paean, the physician of the gods, because of the healing power of its waters (Dvorjetski 2001-2:505). Also a list of what may be statues or names of the baths was given, which included Hygieia and Galatia, a nymph associated with springs and water sources (see Dunbabin 1989:16). This raises some of the issues concerning the medicinal properties of bathing, in particular thermal spas. Evidence for the use of thermal springs for health before the Roman period is restricted to Herod, who bathed in the springs at Kallirhoe for pain relief before his death (Weber 1999:440; Hoss 2005:11). In a similar vein, Naaman was ordered by Elisha to bathe seven times in the River Jordan to cure his leprosy (2 Kings 5). The regular use of thermal springs for health seems to have been a Roman development, which also seems to have been the case in North Africa (Wilson 1997:134). Unsurprisingly, information from the East on ancient attitudes towards the relationship between healing and thermal baths focuses on Hammat Gader (see also above on coinage from other sites). The Pilgrim of Piacenza (Itineraria 132 (163, 12-16)) recounts that in the 6th century AD lepers visited the Hammat Gader baths in the evening. They were provided with lamps and incense and spent the night in the pool; if they had a dream vision during the night, they would be healthy within a week (Krug 1985:187). Epiphanos of Salamis (Panarion II: 7, 5) also commented on healing the sick at Hammat Gader (Hoss 2005:82). Interestingly, he suggested that this was a trick of the devil because God could not be present where men and women bathe together. This gives an additional Christian dimension to some of the religious reservations as to bathing observed above, as well as confirming that mixed bathing took place at least in this period.

CHAPTER FOUR

151

Healing in baths seems not to have been restricted to thermal spas.34 A benevolent moneychanger from Antioch, for example, spent his free time caring for patients in the bathhouses of men; his wife did the equivalent for women (Berger 1982:26; Hoss 2005:82). Also in Antioch, in AD 387, the Emperor Theodosius closed the bathhouses as a punishment after the Riot of the Statues, when a mob, unhappy over heavy taxation, overturned and disfigured pictures and statues of the Emperor and his family. John Chrysostom (Ad populem Antiochenum PG 49, Hom. XIV:6) criticised this punishment because the sick, old, children and pregnant women needed the bathhouses to get well (Hoss 2005:83-4). John Chrysostom (De Lazaro conciones PG 48:973) also commented that the urban poor and homeless used public bathhouses as a place to warm up in the day and sleep in at night. These examples may be seen as evidence in the transition of bathing ‘from munificence to benefit fellow citizens [in the Classical world], towards charity designed to succour Christ’s poor [in the medieval world], (WardPerkins 1984:140; also see Tsafrir and Foerster 1997:117, 119, 1312). Although the medicinal virtues of such baths were lauded in the ancient world, this use of bathhouses by the sick raises questions about how hygienic bathhouses were. It would be useful to know, for example, whether the water was changed and the pools cleaned between sessions. Information on pool volumes, waterstorage capacities and aqueduct delivery rates to the bathhouses would, of course, be necessary in order to calculate whether this would have been possible in theory. Fagan (2000), using data from bathhouses in Ostia and Pompeii, has estimated that some of the bigger bathhouses there could have accommodated between 400 and 700 people per day (equivalent to 5 hours of use) all of whom would have been sharing and

Cf also John 5.1-9: The Pool of Bethesda in Jerusalem, not a bathhouse, was used for healing the sick ‘for at intervals the angel of the Lord came down into the pool, and the water was disturbed, and the first person to enter the water after this disturbance was cured of any ailment he suffered from.’ 34

152 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST dirtying the same bathwater.35 His assessment also brings to our attention the lack of chemicals for treating water, the fumes from hypocausts, the humidity and flies encouraged by it, and the presence of gloios (the matter scraped of the skin after being massaged with oil) in the water. While there were no warnings against bathing with the sick in ancient medicinal writings (Fagan 2000:251-2, 256), a Hadrianic law did reserve special bathing times for the sick in bathhouses at Rome (Köhler 2002:303; SHA Hadr. 22.7). In the East, an inscription from Scythopolis refers to lepers’ baths: Θεόδωρος ὁ ποιμὴν λουτρὰ καινουργῶν νέμε[ι] τοῖς τὴν ἄκραν νουσοῦσι τῆς λώβης νόσον ἐν χρ(όνοις) ἰνδικτονος ζ ἔτους χκβ' ‘Theodore the shepherd (bishop) allots, renewing them, the baths to those sick with the very grievous disease of leprosy in the time of the 7th indiction in the year 622 (AD 558/9).’(AviYonah 1963:325)

This seems to suggest that here lepers bathed separately from the rest of the population and so bathing in this particular city could not spread the disease.36 This inscription, therefore, may suggest indirectly some awareness of health issues and how to prevent the spread of disease. This example, and the bishop's involvement, again demonstrates how bath building can be seen as part of the Church's commitment to the welfare of the community (Tsafrir and Foerster 1997:117, 119). In all these cases there seems to be a simple and direct relationship between water and purity and health. There are, however, some indicators that a darker undercurrent ran alongside this, almost propagandistic, ‘healthy bathhouse’ message. The association The numbers are calculated on the basis of people staying in the pool for 15 minutes; when increased to 30 minutes, the number of people using the same bathwater drops to 200 – 380. 36 Recent research by Matsuoka et al. (1999) in Indonesia has shown that water for bathing and washing is an infectious source of leprosy. 35

CHAPTER FOUR

153

of Hygieia and Sarapis on the coins, for example, requires explanation. The later incarnations of Sarapis in the Roman West, as a transcendent god, conservator and invictus, with a link to Helios, initially make for an obvious relationship. In his earlier Egyptian form, however, Sarapis was a god associated with the Underworld. This dualism in Sarapis' character adds a twist to his alliance with Hygieia and provides an initial hint that the conceptual link between baths and cleanliness was not as clear-cut as it first appears. An inscription from the baths at Serjilla says that their construction brought χάρις (grace) and ὄλβος (happiness) to the country, but also warns of the dangers of φθόνος (envy) with a wish that δόξᾰ (honour) may drive out such envy (Butler and Prentice 1901:62-76; Dunbabin 1989:17). A tabula ansata in a mosaic floor from one of the bathhouses in Caesarea said πθόη ἐπάτωσε ; if the second word was a misspelling of ἐπάτησε, this might be a recollection of mortality: ‘Decay trampled’ (Lehmann and Holum 2000:106 no. 98). These examples illustrate that bathhouses did not just provoke feelings of pleasure, cleanliness and fertility, but may also have held a darker side. In connection with this, protective symbols, for example phallic symbols, swastikas, hederae and grotesque figures, are found in bathhouses across the empire in Italy, Egypt, North Africa and Spain (Dunbabin 1989:38-43). There also seems to have been a belief that ghosts and demons lived in dirty water in bathhouses (Eitram 1915:119-124; Hoss 2005:20-1). Ammianus Marcellinus (29.2.28) records that in Antioch in AD 371, for example, a young man was executed for sorcery because he had tried to heal his stomach pains by putting his hand on his body and on the wall of the bathhouse alternately, and speaking the seven vowels of the Greek alphabet aloud (Dunbabin 1989:37; Hoss 2005:89). This was, of course, in a period when there was much fear of magic, and accusations of practising magic were commonplace; Libanius, for example, was accused five times (Liebeschuetz 1972: 5, 14, 32, 33, 106; Libanius Orat. I.43, 62, 98, 194, 239). Pollution, disease and drowning seem to be likely explanations for these fears and superstitions (Kamash 2008). While water itself, in its original form, for example when it issues from a spring, may be clean and pure, it has the potential to carry pollution and disease, just as it has the potential to flood, drown and destroy. The

154 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST transmutability of water was identified in Strang's Dorset study, for example, as one of the reasons it evokes fear (Strang 2004). It seems plausible that it is this transmutability, this potential of water to be something else and to act in ways that stray from its original character that make it an object of fear. In addition, evidence from clinical and evolutionary psychology and anthropology has shown that humans have a hazard-precaution system, such as that displayed by the epigraphic and art-historical evidence presented above, that responds to potential (as distinct from actual) contaminants (Boyer and Lienard 2006). (Dulaney and Fiske 1994; Fiske and Haslam 1997). It has also been argued that things are seen as pollutants when they confuse classifications and that they are dangerous when they have a half-identity and are recognisably out of place (Douglas 1966: 37, 161). If water is inherently confusing through its transmutability, this may make it seem like a pollutant and so out of place in bathhouses, which, if the multiple images of and invocations to Hygieia are to be believed, were supposed to be clean. The identity of water might then become ambiguous and bewildering, and so a source of fear that may be viewed as a potential hazard. It seems then that while there may have been a wish for water in bathhouses to be associated with cleanliness, there was also a recognised need for protection from its capricious and potentially dangerous qualities.

CONCLUSIONS This overview of ways of looking at attitudes to hygiene and cleanliness in the Near East shows that the situation was complex, nuanced and cannot be inserted simply into simple categories of ‘clean’ or ‘dirty’, ‘hygienic’ or ‘not hygienic’. In many cases the choices that seem to have been made were contingent upon a series of influences and considerations. In the case of drains, flexible approaches were taken to cope with specific needs, with different solutions being used according to what the drains were being used for and where they were located. The idea that certain rivers were deliberately polluted, and so not used for drinking water supply, suggests that different bodies of water should be considered individually. Looking at miqva'ot introduces the idea that what is physically clean may not be the same as what is considered ritually pure. It also suggests that even within a particular group of people tied together by a single identity, there can be variations and different

CHAPTER FOUR

155

interpretations of concepts such as cleanliness and purity and how these should be executed. The problems with accepting latrines in the Near East seem to be multi-layered, comprising responses to colonialism as well as being woven into deeper concerns about what it means to behave socially and what is thought to be disgusting or impure. The case of the latrines also demonstrates that these boundaries are fluid and can change and shift over time. The introduction and use of bathhouses was similarly nuanced. While baths became more commonly accepted than latrines, there was some initial reluctance to use them. Once they had become widespread, concerns over the potential dangers that might be encountered while using them are indicated by undercurrents of fear and the need for protection, which accompany a surface story of cleanliness and purity to be found in the decoration and exhortations associated with bathhouses.

5 WATER AND PAGAN RELIGION IN THE ROMAN AND LATE ROMAN NEAR EAST A link between water and religion had a long history in the region prior to the Roman period. Particularly prominent was a god of storms who took on various names and guises across the region (see Green 2003). In many cases this storm god was the premier deity within the pantheon. While one must be cautious when talking about the 'nature' of these gods, in many cases it would appear that this god had a dual characteristic of violence and gentleness, personifying both the destructive floods and the fertility guaranteed by the storms (Green 2003:282). In particular, the identification of Baal (western Syria) with Hadad (inland Syria) brought together the fecundity of the former with the stormy and fearsome character of the latter (Green 2003:284). This double-edged conception of water persisted into the Roman and late Roman periods, affecting the associations and rites involving water, as will be discussed below. We have already seen elsewhere in chapters 2 and 4 the ways in which religion and religious behaviour affected uses of water, particularly in the slow and restricted uptake of baths and latrines, and how religion affected water supply and infrastructure, for example the greater control of these areas of life undertaken by the Christian church in the late Roman period, in both urban and rural contexts. In this chapter, the focus will be more explicitly on the place of water in the pagan religious cosmologies of the Near East and religious rites that involved water. In the first part, attention will be paid to the characteristics of the gods who had watery associations, which watery contexts (springs, rivers, wells etc) were deemed sacred and why and what evidence there is for religious iconography with watery associations. The second part will look in more detail at the main rites that involved water, including the Cult of Atargatis, the Maiumas and other festivals, votive deposition in watery locations, water used in oracles and divination and water 157

158 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST used for healing and purification. This discussion will demonstrate some of the themes in the role of water in Near Eastern religion, including some long-term continuities, the particularly Eastern flavour of some rites and at the same time the cross-cultural similarities of other activities.

THE COSMOLOGICAL ASSOCIATIONS OF WATER Gods with a watery element to their character While it is maybe the sun cults that had prominence in the Near East, the cults of Hadad and Baal did continue into the Roman period, possibly becoming Jupiter Heliopolitanus, the god of the great sanctuary at Baalbek. The epithet ‘thunderer’ (κεραύνιος) is also added to the name at et-Tayyibe and in the Hauran. His cult is also very prominent at Hierapolis (Teixidor 1977:53-55). The cult of Doliche, specifically Jupiter Dolichenus – another giver of storm and rain – is attested at Doliche-Dülük and Dura Europos. We also know of an Edomite storm god, Qos, who was depicted with bulls and a thunderbolt (Teixidor 1977:90; Butcher 2003:340). In the 4th century AD in Gaza we hear of continued pagan practices related to a type of storm-god. When Porphyrius, Bishop of Gaza in AD395-420, reached his new see, there had been a severe drought. In response to this, pagans in the city were appealing to Marnas the ‘Lord of the Rains’, making sacrifices and vows at his temple during a seven-day ritual and organizing processions outside the town to a place of prayer (Mark the Deacon V Porph. 19, 7-12; Belayche 2004:8). While no explicit mention is made of the rituals involving water, it is clear that the concept of a storm or rain god still held sway. In addition to these storm gods, worship was also given to Yarhibol, a Palmyrene god (possibly with an Akkadian etymology) who was known first of all as ‘Lord of the Spring’ (Kaizer 2002:145). Yarhibol was curator of the Efqa spring (discussed in more detail below) and identified as god of the spring/source in Dura Europos; additionally an altar found at Arak, 28 km northwest of Palmyra, identifies the god as ‘the irrigator of the earth’ (Kaizer 2002:145-8). Yarhibol also came to be an accompanying sun-god for Bel and depicted with a solar nimbus (Ibid:155). This acts as a fitting reminder that the watery aspects of religion went closely in hand with other factors such as the sun and the earth.

CHAPTER FIVE

159

Indeed, Bardaisan, a Syriac Gnostic in AD154-222/3, suggested that the world was brought into being when the four elements or substances – earth, water, fire and light – were mingled with darkness (Drijvers 1965:1). This association with water and sun, in the case of Yarhibol, need not be contradictory or incongruous, rather it could be seen as fitting for a god associated with irrigation to be related to the sun as both are equally necessary for a good crop and harvest. Atargatis, who was also closely associated with water and fertility, is discussed in more detail below. The main gods associated with water, then, to some extent maintained the weather aspects of the earlier gods, also maintaining the dual nature of destroyers and bringers of fertility. This type of god has a long-term role in the Near East, continuing into the late Roman period as seen at Gaza. This is thought to be typical of settled regions where deities tend to be concerned with cycles of nature, regeneration, water and earth (Butcher 2003:342). As well as rain gods, we also see a god of springs, Yarhibol, who subsequently gains a stronger solar aspect, which may be fitting for a god associated with irrigation and agriculture. In what follows, the link between gods and particular water sources is explored in more detail. Cosmological associations of watery locations: sacred springs, rivers, lakes, pools and wells The phenomenon of sacred springs and rivers etc is well-known across the Roman world, so it is not surprising to find that it existed in the Near East as well. The current information on these sacred watery sites is given in Table 18. Several themes arise from these data that have the potential to provide deeper understandings of this phenomenon. In what follows, I will discuss the types of evidence we have for these sacred sites, the date range they cover, the areas for which we have evidence, the types of water they represent and the aetiologies associated with them. These data also provide insights into some of the rites and rituals associated with water, such as votive deposition, oracles and divination and festivals, which will be discussed in more detail below.

160 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST

Name Hierapolis

Type Lake

Comments Lake with fish sacred to Atargatis; altar in centre; circular. Observed by travellers in 18th and early 20th centuries (though the feature may actually have been a football pitch). Derceto angered Aphrodite, so thrown into lake where metamorphosed into a fish. Cult of Atargatis.

References Lightfoot 2003; Goossens 1943; Sourdel 1952; Drijvers 1980; Kaizer 2007; Maundrell 1749; Cumont 1917. Lucian De Dea Syria [DDS] 45-48.

Ascalon

Lake

Edessa

Pond

With fish sacred to Atargatis. Pond was shown to Egeria/Etheria in 5th century AD. Today sacred to Ibrahim.

Drijvers 1980:79-80. Itinerarium Aetheriae 19,7.

Amrith

Lake

6th century BC complex to Milqart and ?Eshmun. Monumental naos on base in centre of lake, possibly with fish. Visible remains.

Lightfoot 2003:491; Dunand and Saliby 1985, 14-20, pls I, III, LXI-LXIV.

Goossens 1943:62. Diod. Sic. II.4; Eratos. Cataster. 38

CHAPTER FIVE Name Si’a

Type Pool/ pond at temple

Comments Staircase leading into pool or pond. Cult centre of Baal-Shamin, 'lord of the heavens'. Staircase leading into pool or pond. Visible remains. Maioumas festival. Inscription from AD535. Maioumas festival. Inscription.

Mushannaf Jerash/ Gerasa

Pool/ pond at temple Reservoir

Tyre

?Pool

Antioch

?Pool

Maioumas festival.

Aphaca (also known as Afqa – above Byblos)

Pool

Fijeh

Spring and pool

Pool and grove. Sacred to the Urania or Astarte. Focus of votive deposition. Goddess descended into pool in form of fiery star. Oracle. Sanctuary of Fijeh in Barada Valley. Channel leads to pool (c. 8 m x 6 m). Coins with temple and water minted under Macrinus and in name of Salonine in reign of Gallienus.

161 References Sourdel 1952:98; Lightfoot 2003:491; Freyberger 1998:55. Lightfoot 2003:491; Freyberger 1998:55. Welles 1938:no 279; Belayche 2004:16. Robert and Robert 1978: no 522 and 599; Belayche 2004:16. Liebeschuetz 1972:230-231; Belayche 2004:16. Robertson Smith 1889:155, 162. Butcher 2003:349. Sozomen HE ii, 4/5; Zosimus I.58.

Robertson Smith 1889:156; Aliquot and Piraud-Fournet 2008.

162 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST Name Yammoune Chalos

Type Spring and pools River

Comments Associated with temple.

References Butcher 2003:460.

Syrians consider the fish to be deities and will not allow them to be harmed. River flows with blood each year because Adonis was wounded on Lebanon. Alternatively, according to a Byblian, the high winds make the soil stain the river. Hera (?Atargatis) bathed in the river after her marriage with Zeus (?Hadad); balmy scents and tame fish mark the spot.

Lightfoot 2003:65; Robertson Smith 1889:160. Xenophon Anab. i. 4,9.

Adonis

River

Aborrhas (Khabour)

River

Belus

River

River reddens and is associated with Adonis’ tomb.

Robertson Smith 1889:155, 159. Eusebius Vit. Const. iii, 55. Josephus BJ ii.10.2. Robertson Smith 1889:155. Robertson Smith 1889:155.

Asclepius

River

Ares

River

Kishon

River

Near Sidon; named after god. Possibly the River Lycus; named after god. Named after god.

Lightfoot 2003:317; Maundrell 1697:46; Robertson Smith 1889:155. Lucian DDS 8; Eusebius Vit. Const. iii. 55.

Lightfoot 2003:493; Robertson Smith 1889:157. Aelian Nat. An. 12.30; Pliny NH 31.37, 32.16.

Robertson Smith 1889:155.

CHAPTER FIVE Comments Named after god who was worshipped under form of lion at Baalbek, which ‘stands at true source of river’. Source of Jordan; grotto was sacred to Pan. Carved out by a dragon that disappeared into the earth at its source. Coins with river god Barada/ Chrysorrhoas.

163

Name Leontes

Type River

References Robertson Smith 1889:155-6. Damascius Vit. Isidore §203.

Paneas/Ba nias

River

Orontes

River

Barada

River

?

Rivers and fountains

Those which are sacred to Baalshamin are those which received his blood when he was mutilated by his son.

Joppa

Spring

Flows red with blood after Perseus had slain Andromeda’s monster

Lightfoot 2003:327; Robertson Smith 1889;159. Pausanias 4.35.9.

Yarmouk

Springs

Sulphur springs associated with the Cult of Asclepius.

Sourdel 1952:47; Jalabert 1906:157161.

Beersheba

Spring

Robertson Smith 1889, 156. Robertson Smith 1889:156. Strabo xvi, 2, 7; Malalas 38. Robertson Smith 1889:156; Aliquot and Piraud-Fournet 2008. Robertson Smith 1889:159. Eusebius Praep. Ev. i.10.22 (Fr. Hist. Gr. iii. 568).

Robertson Smith 1889:158.

164 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST Name Efqa, Palmyra

Type Spring

Comments Dedications to Yarhibol or BaalShamin and altars to the so-called 'anonymous god'; limited evidence for a sanctuary. Dedicatory and building inscriptions.

References Kaizer 2002:143148; Teixidor 1977:110-3.

Ras el-Ain

Spring

Near Baalbek

Aliquot and PiraudFournet 2008:96; Van Ess and Weber 1999:9.

Ain el-Jouj

Spring

Near Baalbek. With cult building and votives.

Aliquot and PiraudFournet 2008:96; Van Ess and Weber 1999:52-55.

Ain elQabou

Spring

Mount Lebanon

Aliquot and PiraudFournet 2008:96; Jalabert 1907:301-2.

Ain Houchbay

Spring

Beqaa Valley

Aliquot and PiraudFournet 2008:96; Jalabert 1907:286; Van Ess and Weber 1999:54.

Temnine al-Faouqa

?Spring

Beqaa Valley

Aliquot and PiraudFournet 2008:96; Krencker and Zschietzschmann 1938:138-140.

CHAPTER FIVE Name Daphne

Type ?Spring

Abraham’s Well, Mamre

Well

Mabbug, Hierapolis

Well

Sumatar Harabesi

Wells

Desert beyond Bostra

Cleft

Comments Water and surrounding groves. Inscription mentions sacred fountain ‘in the mountain’. Daphne was sanctuary of Heracles/Baal before temple of Apollo built. Oracle. Mamre sacred festivals celebrated around a tree and this well, Focus of votive deposition. Magicians practising near well where unclean demon lived; sea water is poured into well to kill the demon. Area dedicated to the Cult of Sin. Inscriptions; wells visible. Stygian waters – where a great cleft received a lofty cataract. Witnessed ordeals by water.

165 References Robertson Smith 1889:158. Waddington no. 2370; CIS no. 3, 1.17; Malalas 204.

Belayche 2004:9; Roberston Smith 1889:158, 160-1. Josephus BJ 4, 533. Sozomen HE ii, 4. Drijvers 1980:93-4; Lightfoot 2007. Apology of PseudoMeliton §9.

Drijvers 1980:122.

Robertson Smith 1889:154, 163-5. Damascius Vita Isidori §199.

Table 18: Evidence for sacred lakes, pools, rivers, springs and wells across the Roman Near East.

A mixture of watery contexts is represented here, with the main divisions being between those that are ‘natural’, such as

166 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST springs and rivers, and those that are ‘man-made’, such as artificial lakes, ponds and wells. Whether this distinction was important at the time, however, is less convincing. Artificial lakes, ponds and pools seem to have been associated predominantly with temples and sanctuaries, but the activities, practices and associated aetiological explanations do not seem to show any clear patterning according to a ‘natural’ v ‘man-made’ dichotomy. The majority of the evidence we have for these sacred sites is literary and epigraphic. Given the likelihood of votive deposition at some of these sites (see below), it seems possible that there could be more archaeological evidence if attention was paid to the objects found at these sites. There appears to be no clear patterning in the dates and there is a relatively even distribution of evidence across the period, which suggests that many of these pagan practices associated with water continued following the rise of Christianity, which may be expected and has already been noted at Gaza. What is more surprising is the seeming prevalence for these sites and locations in Syria, with remarkably few from Arabia. What is not clear is whether this represents a genuine pattern with such ways of manifesting religious belief being more popular in Syria or whether this represents a bias in our data. The focus on the cult of Atargatis in both the literature and scholarly research may account for some of this bias. It is possible that there is more literary evidence for this area than other parts of the region, though this seems less convincing for Judaea than for Arabia. In the case of Judaea, it may be that if miqva’ot had been included in the table, the distribution of sites would appear more balanced between Syria and Judaea, though miqva’ot cannot be considered quite equivalent as they did not represent foci for veneration and played a rather different role in religious behaviour. The reason for the lack of sites in Arabia may be the general lack of watery sites, in particular rivers, that could make foci for religious activities and veneration; for example we do know that the one major river, the River Jordan, was used for healing, so it is not that there were restrictions against such practices. Several interesting themes come out of an examination of the various aetiological stories associated with many of these sacred watery sites, which may provide insights into attitudes towards and perceptions of water during this period. The predominant themes are blood, floods, and fish; the latter will be discussed more fully

CHAPTER FIVE

167

below. The focus on blood and floods again seems to draw on a fascination with the destructive, lethal capacities of water. The redness of the River Adonis, the River Belus, rivers and fountains sacred to Baalshamin and the spring at Joppa are all attributed to blood. The redness of the Rivers Adonis and Belus is attributed to the blood of Adonis when he was killed by a wild boar while out hunting on Mount Lebanon. The rivers and fountains of Baalshamin are said to run with his blood after he had been mutilated by his son. Finally, the spring at Joppa is supposed to have been reddened by the blood of the monster Perseus slew when saving Andromeda. As blood can also be seen as representative of both life and death, it may actually be highly appropriate as a counterpart to water. Other negative connotations of water are also found in the references to a demon living in a well at Hierapolis and a dragon creating the Orontes River. The dragon story, where it disappears into the earth at its source, has overtones of the flood stories. Similarly, the unclean demon in the well near Hierapolis is supposed to personify the flood (Drijvers 1980:93-4). The story of the demon in the well is interesting not only because it demonstrates the belief that dangerous things were thought to dwell in water, but also that waters with different mineral properties have different powers (see below). Watery themes in religious iconography Much of the religious iconography with watery associations features fish and is associated with the Cult of Atargatis. This includes Glueck’s ‘dolphins’ from the Nabataean temple at Khirbet Tannur and the Triton relief now in Urfa museum. As well as this body of iconography, there is also an interesting relationship between representations of city tyches and water (Table. 19). Like the sacred watery locations, these tyche representations predominate in Syria. In all the examples, the goddess has a figure at her feet who is interpreted either as a river-god, or in the case of Palmyra, as the Efca spring. The so-called Aphrodite-tyche from Dura Europos is slightly different as the goddess appears as Aphrodite, but like a tyche has a swimming figure at her feet. The dates of these representations also suggest that this is a long-standing relationship and one which endured for at least six centuries, though this does not appear to continue into the late Roman period. Other gods, such as Poseidon, also appear on coins, for example from Rabbathmoba,

168 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST where the god is accompanied by a dolphin (Spijkerman 1978:271, 273, 275, #29, 34-5, 37). City Palmyra Antioch Edessa Harran

Object type Relief

Date

Reference

Dedicated AD158.

in Drijvers 1980:6768, 84-85, 98 and pl. 33.4. rd Bronze c. 3 century BC Drijvers 1980:67statue + and Roman coins. 68, 84-85, 98 and coins pl. 33.4. Coins Elagabalus, Drijvers 1980:67Severus Alexander 68, 84-85, 98 and and Julia Mamaea. pl. 33.4. Coin Drijvers 1980:6768, 84-85, 98 and pl. 33.4.

Nisibis

Coin

Drijvers 1980:6768, 84-85, 98 and pl. 33.4.

Resh Aina

Coin

Drijvers 1980:6768, 84-85, 98 and pl. 33.4.

Dura Europos

Plaster sculpture

Downey 1977:161-2.

Abila

Coins

Marcus Aurelius, Spijkerman Commodus, Ela- 1978:51, 53, 57, gabalus. #1, 7, 12, 26, 29; Aliquot and Piraud-Fournet 2008:95.

CHAPTER FIVE City

Object type Coins

Adraa

Date

Reference

Commodus, Septimius Severus/Julia Domna, (?)Elagabalus, Valerianus. Elagabalus.

Spijkerman 1978:61, 63, 65, #6, 11, 14, 15.

KanataKanatha Capitolias

Coins

Dium

Coins

Gadara

Coins

Marcus Aurelius, Lucius Verus, Julia Domna.

Gerasa

Coins

NysaScythopolis

Coins

Marcus Aurelius, Lucius Verus, Commodus, Elagabalus. Lucilla, Septimius Severus, Julia Domna, Caracalla, Geta, Elagabalus, Aquilia Severa.

Pella

Coins

Coins

Rabbathmoba Coins

169

Marcus Aurelius, Lucius Verus, Caracalla, Macrinus. Julia Domna, Elagabalus.

Lucilla, Commodus. Septimius Severus, Julia Domna, Caracalla, Geta, Elagabalus.

Spijkerman 1978:95, #15. Spijkerman 1978:99, 101, 103, 105, #3, 7, 18, 22 Spijkerman 1978:119, 121, #4, 13 Spijkerman 1978:139, 141, 147, #42, 45, 50, 71. Spijkerman 1978:161, 163, 165, #9, 11, 16, 17, 21, 22, 32. Spijkerman 1978:193, 195, 197, 199, 203, 205, #15, 24, 28, 29, 33, 35-8, 4951, 55. Spijkerman 1978:213, #5, 7. Spijkerman 1978:267, 269, 271, 273, 275, #14-17, 20-1, 28, 33, 39.

Table 19: Representations of city tyches and water.

170 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST

USES OF WATER IN RITUAL These sections focus on activities involving water in various rituals across the Near East. Some of these activities appear to be particular to certain deities and cults, such as the Cult of Atargatis and the Maioumas festival. Others seem to reflect broader activities and behaviours that cut across the lines of particular sets of religious beliefs, for example uses of water for oracles and divination and for purification, votive deposition in water and water-pouring rituals. Indeed, some of these activities, such as votive deposition, have striking similarities with ritual activities in other parts of the empire and may point to very broad, even cross-cultural, beliefs and associations with water. Cult of Atargatis Atargatis has a well-known and well-discussed association with water, in particular with fish. Several cult sites are known in the East (Hierapolis, Palmyra, Dura Europos, Besechena, Ashqelon, Edessa, Harran, Hatra) and further afield, for example at Delos, mainland Greece and Macedonia, Asia Minor, Rome, the south-eastern coast of Spain, the Danube provinces and Hadrian’s Wall (Lightfoot 2003:10). One of the best known of these sites is Hierapolis as many of the details about the cult of Atargatis are known from Lucian’s De Dea Syria (DDS), which devotes much of its discussion to Hierapolis specifically. This is a text that is not without its problems, but is generally agreed to be useful nevertheless (Kaizer 2007:452-3) and is now available in an invaluable edition by Lightfoot (2003). The discussion here will focus primarily on the watery aspects of the cult, though it is recognised that the cult has more facets. The main watery features of this cult to be discussed here are: the physical presence of water in the form of a pool or lake, the focus on sacred fish, and water-pouring rituals. Clear mention is made by Lucian (DDS 45-7) of a lake at Hierapolis. This lake is said to contain the sacred fish, be ‘more than 200 hundred fathoms deep’, have an altar in the centre which is ‘always wreathed and burns spices’ and at which ‘very great festivals’, involving the cult images and called ‘descents to the lake’, take place. Goossens (1943:106-120) has reconstructed the archaeology of Hierapolis using travellers’ descriptions from the 18th to early 20th centuries (including: Pococke 1745; Chesney 1850;

CHAPTER FIVE

171

Maundrell 1749; Cumont 1917; Stocks 1937; Volney 1787). From these it appears that the lake may have been to the west of the temple, near the town walls where the remains of a possible pool c. 100 m in diameter have been observed. This lake appears to have been fed by an aqueduct that entered through the city walls; this aqueduct, or possibly qanat, may be the subterranean channel that Volney (1787:150) described as leading water from the mountains to the north of the town. Steps descending into the pool were visible to the west and south. The south-western part of the pool was divided by a wall from the rest of the pool, which is reminiscent of Birketain at Jerash. A big courtyard (c. 60 m x 50 m) was found next to the pool. In the middle of this courtyard was a paved, square platform (c. 8 m x 8 m), in front of which was a well that had been scored by rope marks. No mention of a well is made by Lucian, but the Apology of Pseudo-Meliton tells the story of an unclean demon who lived in a well near Hierapolis (Drijvers 1980:934; Lightfoot 2003:336), though this could easily be another well. Every Atargateion is supposed to have had a fishpond (Stocks 1937; Goossens 1943:62; Drijvers 1980:79). In DDS 45, Lucian comments on the fact that the fish were large and that one was wearing gold jewellery, though this fact is not attested elsewhere. There is a strong association between fish and the cult of Atargatis. Much of the iconography connected with goddess is done so on the basis of fish, for example a cultic relief from northern Mesopotamia, the reliefs from Khirbet et-Tannur, a tessera and a relief on a beam from Palmyra, seals from Syria and Mesopotamia and a relief from Hierapolis (see Lightfoot 2003:67-70 for full details; also see Drijvers 1980:101-116). It would also seem that fish were strongly associated with Syria, for example, Xenophon (Anab. i.4, 9) comments on the presence of sacred fish in the River Chalos (Robertson Smith 1889:160; Lightfoot 2003:65). In a domestic context, Abadie-Reynal (2008:115) has also suggested that the link between triclinia and fishponds was a particularly eastern phenomenon. There are also several known parallels for temple sites with pools that have similar features to those attributed variously to Hierapolis. A sacred lake at Ascalon is supposed to have been the lake into which Derceto the mortal was thrown after angering Aphrodite and then metamorphosed into a fish (Diod. Sic. II.4; Erotas. Cataster. 38; Goossens 1943:62). Amrith (classical Marathus), to the south of Arados in northern Phoenicia, had a sacred lake, possibly

172 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST containing fish, in the centre of which was a monumental naos (Lightfoot 2003:49). This complex was begun in the 6th century BC and was apparently sacred to Melqart and possibly Eshmun; this suggests a long-standing association of sacred pools and lakes with temple complexes in the Near East. Other temples that had pools or ponds with steps leading into them have been found at Si’a and Mushannaf in the Hauran (Freyberger 1998: 55 and Taf 34c, 62 and Taf 41b; Lightfoot 2003:491). The nun Egeria/Etheria (Itinerarium Aetheriae 19,7), while staying at Edessa for 3 days, describes the carp-pond at the royal palace of King Abgar, north of the citadel, that was shown to her by the local bishop (Drijvers 1980:80). A carp-pond sacred to Ibrahim is still present in Edessa today and presents an interesting case of transferring elements of one religion into another (Drijvers 1980:79). As mentioned above, according to Lucian (DDS 47) the ‘descents to the lake’ involved taking the cult images down to the lake. This appears to be a practice with a long history and there are Hittite and Babylonian parallels (Haas 1994:745; 865; Caduff 1986: 256-7; West 1997:37, n. 147; Lightfoot 2003:491, n. 21). In the New Year festival at Edessa, the Epic of Creation was recited and on the eighth day the images of the deities would emerge from the temple and be offered water for washing (Drijvers 1980:59-60). Taking the cult images down to the lake may be related to a myth in Aelian (NA 12.30; also Pliny NH 31.37, 32.16) in which the goddess ‘Hera’ bathed in the River Aborrhas (Khabour) after her marriage to ‘Zeus’; balmy scents and tame fish were said to mark the spot, which makes an interesting parallel to the fish at Hierapolis. The other rituals involving water that are recounted by Lucian (DDS 13, 48) are those that involve carrying and pouring water. The first time this is mentioned, the water is said to be poured into a chasm that is associated with Deucalion and the flood myth. In the second instance, the practice is said to be part of a great festival by the sea and part of the ritual involves getting the seal of the vessels containing the water checked by a sacred cock that lives beside the lake before carrying their pots to the temple. In both cases, it is specified that the water is brought from the sea. Sea water is also stipulated in the story of the magicians who need to slay the unclean demon in the well near Hierapolis as recounted in the Apology of Pseudo-Meliton (§9; Lightfoot 2007:84, 99-105). Other water-

CHAPTER FIVE

173

pouring rituals such as that practiced at Tyre until the 18th century also made use of sea water, in this case to bring rains as well as make the water pure (Goossens 1943:71). In her commentary on the passages in Lucian, Lightfoot (2003:350) suggests that ‘a pilgrimage from Hierapolis to the real sea and back would have taken some time’ and so it is possible that the ‘sea’ is actually the Euphrates. While this is, of course, possible, there seem to be two problems with this interpretation. The first is that while it is a fairly long journey to the sea, we know archaeologically that Mediterranean fish were being brought inland to, for example, Andarin (Mango 2009:75), so the journey must have been possible in a relatively short period of time. Also, it seems to miss the significance of the insistence on sea water, which is more than likely due to its high salt content. Salt and salt water were well known for their antiseptic properties in ancient times. The Bible, for instance, is littered with references to the salt use for ritual cleansing starting with the dietary laws in Leviticus. In particular, in studies on the symbolic significance of salt it has been suggested that because salt can preserve bodies from decay, devils and demons were said to hate it (Jones 1923:116), which appears to corroborate the story of slaying the demon in the well with sea water.37 Indeed, in her commentary on the Apology of Pseudo-Meliton, Lightfoot (2007:99, 102-5) notes that the ritual 'stands a good chance of being genuine' because of the similarity of the Apology's Zoroastrian water-pouring ritual to that described in Lucian and that on the Derveni papyrus (col. vi.5-6: τοῖ δὲ ἱεροῖ[ς] ἐπισπένδουσιν ὕ[δω]ρ καὶ γάλα 'And on the offerings they pour water and milk'). It would seem then that it may be possible to take Lucian at his word here and accept that it was sea water that was used in the Hierapolis rituals. Maioumas and other festivals Other religious festivals involving water are also known from the Near East. Most (in)famous of these is perhaps the Maioumas festival. The festival may have included mixed bathing, nocturnal displays and ritual feasting (Coleman 1993:71 n. 109). There appear to 37

Thanks to Janice Kinory for pointing me in the direction of these studies.

174 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST have been three Maioumas festivals that took place at Antioch, Ostia and Constantinople and there is epigraphic evidence for Maioumas festivals at Nicea (in 3rd century), in Tyre and in Gerasa/Jerash in AD535; contrary to common opinion, there is no evidence in the literary sources for this festival at Gaza or Maiuma (Belayche 2004:14-16). At Jerash a small theatre overlooking Birketain, a large double reservoir near one of two springs that supplied the town’s water, is thought to be the location for the festival. The literary and epigraphic evidence appears to suggest that these festivals became increasingly popular in the late Roman period. Libanius (Ad Timocr. 16), however, felt that the festival was licentious (Liebeschuetz 1972:230-1). Given the increasing Christian emphasis at Jerash, the revival of this festival points to the complex nature of the co-existence between pagans and Christians. The bathing rituals at Ascalon on the spring equinox when young men and women bathed together in the sea may also have been linked to the Maioumas festival, or to Atargatis (Goossens 1943:71). Votive deposition, oracles and divination There are several examples of votive deposition, oracles and divination being associated with watery contexts. Sozomen (Ecclesiastical History ii.4) tells us that: ‘… even at Aphaca, where…the goddess of the spot was held to be the Urania or celestial Astarte, the pilgrims cast into the pool jewels of gold and silver, and webs of linen and byssus, and other precious stuffs… Similarly, at the annual fair and feast of the Terebinth, or tree and well of Abraham at Mamre, the heathen visitors, …, not only offered sacrifices beside the tree, but illuminated the well with lamps, and cast into it libations of wines, cakes, coins, myrrh and incense.’

Such practices are well-attested archaeologically in other parts of the Empire, for example Britain, where they have a long history (Bradley 1998; Fulford 2001; Kamash 2008). In the Near East less appears to have been done to identify these deposits in the archaeological record, but we are able to know about the organic materials, such as linen, wine and cakes, which are more difficult to trace archaeologically and so adds another layer of understanding of these practices.

CHAPTER FIVE

175

Another practice for which there is no evidence from eg Britain is the use of these votive depositions in water for oracles and divination. The Stygian waters in the desert beyond Bosra were reputed by Damascius (Vita Isidori, §199) to have the power to swallow up or cast back the votive offerings deposited in them, which demonstrated whether the god was propitious or not (Robertson Smith 1889:154, 162). These waters also had the power to kill people who swore falsely by them. Zosimus (I.58) tells a similar story about the springs at Aphaca. The fall of Palmyra was said to have been predicted at the sanctuary of Aphrodite at Aphaca when the offerings made by Palmyra the previous year floated back up to the surface of the pond (Robertson Smith 1889:162). Another suggested location for this sanctuary is Yammoune on the western side of the Lebanon mountain range as there are a temple, spring and pools here (Butcher 2003:349, 460). Hajjar (1990:2275) has also suggested, tentatively, that Yarhibol may have been associated with hydromancy at Palmyra. Finally, the oracle at Daphne could be accessed by dipping a laurel leaf into the water (Sozomen v.19.11; Robertson Smith 1889:163). Water, healing and purification The general idea that water is associated with healing and purification has been explored with reference to built structures such as bathhouses and miqva’ot in Chapter 4. There is also evidence that some ‘natural’ locations were associated with healing and purification. The sulphur springs in the Yarmouk Valley, for example, were associated with the cult of Asclepius, particularly amongst the military (Jalabert 1906:1 157-161; Sourdel 1952:47). We also hear, for example, of the Syrian general Naaman who was instructed by Elisha to bathe in the River Jordan seven times to cure his leprosy (2Kings 5:10-14; Robertson Smith 1889:156); interestingly, Naaman originally refused because he felt that the rivers of Damascus were better. Also, Elisha caused an iron axe head to float in the River Jordan (2Kings 6:1-7), which is an interesting inversion of the stories above about floating items being bad omens.

CONCLUSIONS Several themes become apparent when looking at the role of water in religion in the Near East. Firstly, there is a complex relationship

176 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST between rain/storm gods and the sun, which appears to be related to agricultural cycles. This is a long-standing relationship and we see that the nature of the gods is maintained over very long periods of time. Indeed, we even see some pagan practices involving water continuing into the Christian era and beyond. In addition to this link between rain/storms and the sun, there is also a long-term association of water with not only purification, but also death and destruction, which is particularly clear in some of the aetiological stories. In the Roman period the cult about which we can say the most is, perhaps inevitably, that of Atargatis. By looking at this cult we can get a sense of what some of the rituals and activities taking place with and around water were. There was a particular focus on fish, which seems to have become closely associated with a Syrian identity. We also see the use of ponds and lakes, not only for the keeping of fish, but also for other activities such as the washing of cult images. Finally, the carrying and pouring of water seems to have had a special significance that was heightened by the use of specific kinds of water, in this case sea water. The special attention given to the mineral properties of water is also seen with the sulphur springs in the Yarmouk valley. Finally, there are hints that some of the practices we see in the East may be part of broader, cross-cultural rituals involving water. The use of watery locations for the deposition of artefacts is wellattested in northern Europe in the prehistoric and Roman periods. The impression gained archaeologically of these activities is tantalisingly similar to those we hear about in the written records from the East. This is a topic that deserves more attention and close analysis of the archaeological record may be able to shed more light on this apparent link.

CONCLUSIONS This study has taken a thematic approach to the main methods for the transport, storage and use of water in the rural and urban areas of the Roman and late Roman East. In particular, it has focussed on issues of technological transfer (Chapter 1), economic uses of water (Chapter 2), attitudes towards water as a resource (Chapter 3), perceptions of hygiene and cleanliness (Chapter 4) and the uses of water in pagan religious practices in the East (Chapter 5). It has raised several issues that have direct relevance to ancient water supply studies in general and in some cases also contribute to wider debates: the relationship between cities and their hinterlands as shown by aqueducts and differences in water management between the Roman and late Roman periods. This final chapter, then, provides an opportunity to draw together some of these strands and to answer the following questions. How did water technology compare to other areas of the Empire? Why did changes in technology occur (or not occur)? Were there variations in the region itself? Who were the agents of these changes? How are these changes related to perceptions of how water should be used, and to what extent is this related to the climate of the region? Finally, what role did water play in the various religious lives of people living in the Near East and how did religious belief impact on the ways in which water was used? Overall, as the breadth of these questions shows, water has acted as an excellent diagnostic. Its study has enabled insights into urban organization and the nature of the community; large organizations such as the army and religious institutions; ethnicity and identity; religious beliefs and concepts of purity and pollution. In general, the degree of similarity with the rest of the Roman Empire depends on the realm of water supply and management under discussion. The areas with which the Near East had the most similarities were North Africa, Egypt and Spain, particularly in their 177

178 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST respective approaches to water-lifting, damming and irrigation. There is, for example, literary and archaeological evidence for the widespread use of similar water-lifting devices (Chapter 1) across the empire, especially in Egypt (Oleson 1984). Two possible exceptions are norias and saqiyas. Although there is some evidence for norias used elsewhere in the Empire, they seem to have been particularly associated with the eastern provinces. On present evidence, it would seem that saqiya pot use was restricted to Egypt, the East and possibly Spain. It seems then that it was mostly used in the more arid areas of the Empire, whereas bucket chains were used more commonly in provinces further north. Dams were an interesting example of a technology that derived from the East, but was modified during the Roman period when it also spread to other areas of the empire, primarily Spain, Asia Minor and North Africa. These dams exhibited some of the new design features noted in the East, such as steps on the Proserpina and Kasserine dams. The irrigation techniques used in the East were also found in other parts of the empire. Large-scale irrigation channels were used in Spain, where the recent find of a bronze inscription from Agon near Zaragoza detailing irrigation regulations demonstrates their use in the Roman period (Wilson 2003a:117-118, 139). Qanats were definitely used in Egypt from the 5th century BC at ‘Ayn Manawir. Recent work in the Sahara suggests that the qanat had reached this area before the 4th century AD, from where it spread to Tunisia and Algeria. The date when qanats reached Spain is more controversial (Ibid:133-138). Finally, floodwater farming and its associated field systems have parallels in North Africa as illustrated by the UNESCO Libyan Valleys Survey (Gilbertson and Hunt 1996). As in the East, the distribution of techniques seems to have been governed by the landscape setting, for example the North African landscape is more suited to floodwater farming and aquifer-fed qanats than river-fed irrigation channels. By and large the technologies associated with urban aqueducts in the East did not differ from those in the rest of the empire. The exceptions were stone pipelines, which seem to have been most common in the eastern Mediterranean and ceramic pipelines in inverted siphons, which have very few comparanda elsewhere, though a few examples are known from Italy, North Africa and Spain (Stenton and Coulton 1986; Hodge 1992:110; Wilson 2000e:599). The financing system for the construction and mainte-

CONCLUSIONS

179

nance of aqueducts also seems similar to that elsewhere, with the exception of higher levels of private funding in North Africa. The biggest distinguishing factor may have been the extent to which their water was used in rural areas as well as in urban centres. There is some evidence, albeit limited to Tiberias, to suggest that rural areas were more of a priority in the East than elsewhere. In urban contexts there was less reliance on (stored) aqueduct water than North Africa. The use of castella divisoria was limited, but this was not surprising as they were also rare in other parts of the Empire. It was unsurprising to find nymphaea in several cities as these were a common feature of Mediterranean cities (Glaser 2000:465), but they do not seem to have been as elaborate as nymphaea in Greece and Asia Minor. The relatively low numbers of street fountains were unusual as they are well-documented elsewhere; Pompeii alone probably had 50 (Eschebach 1979; Hodge 1992:304; Jansen 2000b:113). One of the main differences was in the actual distribution network, where ceramic pipes were the eastern Mediterranean preference by far, not only in the Near East, but also in Asia Minor and Greece. This is in contrast to lead pipes across the western Mediterranean. The other water sources, cisterns and wells, did not differ from other parts of the empire. Public latrines and bathhouses showed some interesting similarities and differences with the rest of the Empire. While the design of latrines was remarkably standardised and similar to other Roman latrines, they were a later phenomenon in the East than elsewhere, probably due to cultural inhibitions. Bathhouses also flourished later in the East than in other provinces, which may have been for similar reasons. In addition, the more widespread use of cisterns for public bathhouse supply in the East was in marked contrast to other areas where the practice is almost unknown. In the domestic realm, Antioch had the most similarities with higher class housing across the empire, particularly Italy. This was very marked in the use of water for display, for example fountains. Housing in other parts of the East did not have such a focus on display, which may be related to a seeming dislike for ostentatious uses of water. In addition, the use of lead pipes, as in urban contexts, was very limited in the East with a strong preference for ceramic. A common, though unsurprising, pattern for all housing was that only the highest class housing, both urban and rural, was connected to a piped supply.

180 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST Finally, the industrial uses of water in the East were broadly similar to the rest of the empire. The use of mills and waterpower in the East adds to a growing picture of the use of waterpower that must modify the now outmoded picture of technological stagnation in the Roman world. The use of arubah penstock mills was a particular feature of the East and must be related to their efficacy in areas with smaller stream flows. The East also provided some evidence for more specialised uses of waterpower, such as the sawmill at Jerash and the ore-crushing mill at Faynan. So, while the most prevalent development in the Roman and late Roman periods was one of intensification of previous practices, rather than transformation, for example in irrigation, there were some changes or additions to previous technology (Chapter 1). The use of the pot garland with a saqiya gear and the noria were probably the biggest additions to previous water-lifting practices. Arches and arcades, which were particularly important in aqueduct building, were also introduced during the Roman period. There was also progress in waterpower, which saw innovations such as the arubah penstock mill and the Jerash sawmill. One of the biggest advances in technology seems to have been in dam construction. Increased length and steps improved earlier dam forms. In addition, a pragmatic approach to dam construction was taken that maximised a dam’s utility. New designs, such as the arch dam, were also introduced, though with limited spread. Limited uptake of a new technology was not limited to the arch dam, but was also seen in the restricted use of lead in all contexts: inverted siphons, urban distribution pipelines and domestic pipelines. This suggests that while a technology might be viewed by some as ‘superior’, it would not necessarily be used to any great degree; a theme that deserves more exploration in analyses of the ancient world in general. In these cases, the established, older technology seems to have been viewed as adequate for its task and there seems to have been little propensity towards an ‘onwards and upwards’ attitude. Where such an attitude did exist, it seems to have been held either by the imperial authority or by one consciously trying to become or seem Roman, for example Herod (see below). New ways of using and managing water were also introduced during this period with varying degrees of spread. Most notable among these were the provision of castella divisoria, nymphaea and latrines. Facilities, such as public bathhouses became increasingly widespread during the Roman and late

CONCLUSIONS

181

Roman period. In addition, there was some use of water for display in high-class houses. These developments must be linked to increased provision of running water in aqueducts, but cannot be taken as representing responses of the region as a whole. There were some marked regional differences in water management in the East. In technological terms, these were most apparent in irrigation practices (Chapter 1), which were largely governed by landscape setting. This meant that irrigation channels were most prevalent in Syria (particularly in the Euphrates area) and qanats, which needed access to aquifers, were located in limestone geology. Access to water sources also played a key role in the distribution of aqueducts in the landscape. There were few aqueducts in the Syrian steppe, for example, because the water resources in that area were more suited to qanat technology. The other major regional bias was in the distribution of latrines (Chapter 4). Latrines showed a particularly restricted geographical spread, with most examples coming from Syria and from two sites (Caesarea and Scythopolis) in Judaea. This was a clear example of rejection of this new installation in Jewish areas due to religious propriety (see below). Even in non-Jewish areas the spread of latrines was late in comparison to other areas of the empire, which suggests that similar inhibitions and concerns over modesty affected other cultural groups in the East, whose opinions and cultural tastes are not so easily observable. The answer to the question of why these developments and changes happened (or even did not happen) seems to vary in emphasis according to the different areas of water management. The reasons for these patterns appear to be an amalgam of various factors, including who was the agent of change or conservatism, local environmental constraints, social rules and taboos and shifts in the wider social and economic milieux of the empire over time. In the early part of the period, one figure stands out as an agent of change and as a middleman in technology transfer: Herod. Herod and his palaces seem to have been primary locations for the introduction of new ideas and designs in certain areas of water management, notably arches and arcades on aqueducts, as well as bathhouses. The nature of the transfer from Herod to other areas of the East is not easy to elucidate, but in the case of Caesarea, Herod’s legacy and Roman influence appear to have been maintained throughout the life of the city, in the form, for example, of a

182 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST nymphaeum, latrines and lead pipes, until its downturn at the end of our period. The army in the Roman and late Roman periods also seems to have been a major agency of change. Military areas of involvement seem to have been in the construction of ‘necessary’ facilities for a Roman way of life: aqueducts, bathhouses and latrines. Although aqueducts would have had a tangible effect on urban and rural water supplies, the extent to which the army’s use of bathhouses and latrines had a deep and far-reaching impact on the wider population is debatable. In some cities, for example, Dura Europos and Palmyra, these facilities, arguably, were for the exclusive use of military personnel. The late Roman army appears to have made a stronger impression on rural water management, in particular dams and irrigation techniques. In this case the presence of the army seems to have made their construction both possible and necessary. As well as the late Roman army, the Christian Church may also have been responsible for some of the intensification of irrigation in the later period. Furthermore, the Church had a powerful effect on late Roman urban water management. The shift from public benefaction to spending on religious building or charity meant that the nature of the cities changed. This affected water management as installations (such as reservoirs and latrines) began to encroach on monumental public space or even contributed to its dismantling (eg the Jerash sawmill) and the towns and cities became increasingly industrial. As a corollary to this, the Church began to take on some of the responsibilities and roles of urban water management that might have been expected of civic authorities in an earlier period, for example the upkeep of the network and the building of storage installations and bathhouses. Judaism also had a strong impact on the use of water-related facilities. The dictates of Jewish law made latrine-use unacceptable among devout Jews as well as apparently slowing down the rate of spread of bathhouses. In this case then, religion was an agent that slowed, or even prevented, change. The final group of agents were the indigenous population. It was this group of people that visibly were responsible for the uptake, or not, of many of the trappings of Roman life: latrines, bathhouses, nymphaea and water for display in domestic contexts. The power of this group seemed to lie in their multiple identities: religious, public (civic) and private. Sometimes these hybrid identities

CONCLUSIONS

183

clashed as in the case of the Herod and his latrine in his Second Palace where a compromise had to be met between his Jewish identity and his desired Roman identity. Balancing these identities, in which willingness to change behaviour was a key component, meant that a selective approach was taken to new ideas. Where a change in behaviour would have meant the loss of or damage to identity, it seems to have been met with varying degrees of resistance, such as in latrine use or public display in private houses. In other areas of life where a pragmatic approach would not be detrimental to identity, change seems to have been more acceptable. Key factors, which can be manifested in different sets of concerns, and which seem to be at play in these identity wrangles, were a wish to appear Roman and behave in a Roman way and the socially-embedded attitudes towards water and water technology and what were deemed its appropriate uses. In some cases the desire to act as part of the wider Roman world appears to have overridden more typically Eastern views on water use. This seems particularly restricted to the inhabitants of certain urban centres, such as Antioch, Beirut and Caesarea. In these places, it seems to be that the conditions for this overriding wish to be Roman comprised a mixture of the history of the settlement and its ongoing role and prominence in the Roman world. In these situations, the clash of identities seems to result in a ‘win’ for Rome. For the large majority of the population, however, life involving water and water technology does not appear to have undergone major changes, and there seems to have been a conservative attitude towards change. In part, this appears to be related to the region’s harsh and unpredictable climate where one cannot be assured of a steady supply of water. As can be seen from the religious role of water (Chapter 5), water, and gaining access to it, had long played an important role in the lives of those in the Near East, with a particular focus on its capricious and unpredictable nature. Such climatological factors would seem to have led to technological and social conservatism in the Roman and late Roman periods. When technological choices had to be made concerning water storage and harvesting, in both urban and rural contexts, the security of knowing what works meant that only minor changes in practice were made to ensure that the balancing act of obtaining and preserving adequate resources was maintained. The use of dams in the East, notably an indigenous technology, indicated that the constant-

184 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST offtake principle (whereby water was not stored in significant quantities at any point along its course) might need revision for the East as well as for North Africa, as it was clear that water storage did form at least part of the water management strategy (Chapter 3). This did not take the form of large buffer reservoirs, however. Where such installations occurred, they often appear not to have been contemporary with aqueduct construction, but rather were a late Roman phenomenon. Instead, the response in the East to seasonal variations in water supply seems to have been to make greater use of complementary water sources, i.e. using aqueduct water in conjunction with rainwater cisterns and wells. Supply of the cities was not, of course, the only concern for inhabitants of the East. Water management in rural areas was also important to provide water for drinking, animals and agriculture. The evidence from rural branch lines in the East has contributed to the idea that there was a more balanced and symbiotic relationship between urban centres and rural hinterlands than has been suggested for some other parts of the Empire (Chapter 2). In some cases, it seems that the city may not have been the primary consumer of aqueduct water, for example at Tiberias, and that rural supply may have been an equal concern, possibly reflecting long-standing practices of water use and distribution. Socially, this concern over water as a resource was manifested in an apparent unease with lavish and wasteful uses of water that led to, for example, a general lack of fountains in both public and private spheres. Environment, however, was not the sole mover in determining the choices people made. As shown in Chapter 4, there were also deeply-rooted concerns and rules about cleanliness, hygiene and purity. In many cases these concerns were manifested in religious taboos, of which the easiest to identify are Jewish. The strictures over water and social behaviour in Judaism brought about one of the most striking rejections of technology in the Roman world: latrines. There is nothing especially complicated about the water management of latrines, rather these simple structures just did not, and in most of Judaea could not, fit into the accepted ways of doing things. While not all conservatism in the East can be seen in such strong terms as this example, the lack of use of, for example, lead pipes, at the very least points to a distinct lack of interest in Roman technologies and methods and so of engaging in the wider Roman world. Indeed, an enduring Eastern flavour to water

CONCLUSIONS

185

and practices associated with water can also be seen in some of the pagan rituals involving water, such as the close association between Atargatis, fish and the East. In this case, it is an association that is so strong that it appears to have survived several major episodes of religious change, so that sacred fish can still be seen in Edessa today. This is not to say, however, that these groups of people were insular or remained unaffected by occurrences in the rest of the Roman world. Rural water use appears to have undergone some significant changes in the late Roman period as illustrated by the study of dams and irrigation techniques (Chapter 1). Overall the change seems to have taken the form of increased irrigation, shown by higher numbers of dams, irrigation channels, water-lifting devices (especially pot garlands) and (probably) field systems in this period. While many of these techniques were known and utilised before the late Roman period, it was in that period that there seems to have been a wider diffusion of these technologies. The trend towards higher levels of irrigation fits in with the picture of general agricultural intensification in the late Roman East, particularly in Syria. The construction and use of more irrigation networks appears to have been related to the need to supply the permanentlystationed late Roman army, population and taxation pressure and, in the case of irrigation systems associated with monasteries, the rise of the power of the Church. In addition to increased irrigation, the late Roman period also saw some interesting changes in urban water management. The major shift here was related to the character of the late Roman city in the East (Chapter 2). Water installations often encroached on previously public, monumental space, for example the reservoir and latrine inserted into the agora at Apamea and the officina tinctoria in the macellum at Jerash. The evidence from Jerash also suggested that more industrial activities, some involving spoliation of the earlier pagan monumental architecture, were taking place in the city in this period. This change in urban character was associated with the decline of civic self-government, the rise in the power of the Church and a shift in the concept of community. The increased threat from further east, as well as natural disasters such as earthquakes and drought, may also have contributed to these changes in emphasis. Analysis of these adjustments in water supply and management methods in the late Roman period contributes an alterna-

186 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST tive outlook to the traditional view of ‘decline’ in this period. While it is clear that the nature of the city changed radically, becoming increasingly industrial, it is not immediately obvious that this represents a decline in urbanisation in the region; rather it would appear that the towns and the cities were bustling, active places whose inhabitants had shifted their focus and mindset away from communality and the preservation of memory through civic munificence towards increased individualism and support of the Church, alongside a greater concern for the productive capacity of urban spaces. In a version such as this, the wider, shifting socioeconomic, religious and military background to the period can still be seen to have an impact on late Roman life, but not necessarily in a wholly negative and catastrophic way. It appears, then, that the social, religious and military changes seen across the entire Empire in the late Roman period certainly had an effect on water management strategies and priorities in the East as well. These changes were not, however, an imposition from above, but rather adaptations to the current situation, which demanded alternative responses. This can be seen in the fact that these changes were often not heralded by significant changes or developments in technology or practice, but rather by intensification of practices or shifting emphases. So, this study of archaeologies of water in the Near East has demonstrated the value of researching water to reach some of the broad themes that are of interest to any Roman archaeologist. By looking at the technologies of water management and supply in the region, it has become clear that the various groups of people living in the Near East related in numerous complex ways to the rest of the Roman world. This work has highlighted some of the nuanced rhythms of continuity and change that existed in the Near East, where there is no single story. Instead, water and water technology approaches a 'total social fact' (Mauss 1966:76-7) whereby studying water provides a way into looking at questions of economics, religion, aesthetics, and identity, opening up almost all facets of society to interrogation and increased understanding.

BIBLIOGRAPHY ANCIENT AUTHORS Aelian De Natura Animalium (Valdés, M. G., Fuego, L. A. L. and Guillén, L. R.-N. (eds); Berolini: De Gruyter, 2009). Ammianus Marcellinus (Rolfe, J. C. (ed.) Ammianus Marcellinus.; London, Cambridge MA: Harvard University Press, 1925). Lucian On the Syrian Goddess (Lightfoot, J. (ed.); Oxford: OUP, 2003). Ausonius Mosella (Green R. P. H. (ed.); Oxford: Clarendon Press, 1999). Choricius of Gaza (Foerster, R. and Richsteig, E. (eds) Choricii Gazaei Opera; Stuttgart: Teubner, 1972). Codex Theodosianus (Pharr, C. (ed.); New York: Greenwood Press, 1969). Columella De Re Rustica (Ash, H. B and Forster, E. S. (eds); Cambridge MA: Harvard University Press, 1960). Cyril of Scythopolis Life of Sabas (Schwartz, E. (ed.); Leipzig: Hinrich, 1939). Damascius The Philosophical History (Athanassiadi, P. (ed.); Athens: Apamea Cultural Association, 1999). Diodorus Siculus (Oldfather, C. H., Sherman, C. L., Wells, C. B. and Geer, R. M.; London: Heinemann, 1963-71). Eratosthenes Catasterismorum reliquiae (Berolini: Weidmann, 1963). Eunapius Vitae Sophistarum (Giangrande, G. (ed.); Rome: Typis Publicae Officinae Polygraphicae, 1956). Eusebius Ecclesiastical History (Williamson, G. A. (tr.) and Louth, A. (ed.) The History of the Church; London: Penguin, 1989). Frontinus De Aquis (Herschel, C. (ed.); Cambridge MA: Harvard University Press, 1925). Gregory of Nyssa Gregorii Nysseni Opera (McDonough, J. and Alexander, P. (eds); Leiden: Brill, 1962). 187

188 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST IGR: Inscriptiones Graecae ad Res Romanas Pertinentes III (Cagnat, R., Toutain, J. and Jouguet, P. (eds); Paris: E. Leroux, 1906-27). ILS: Inscriptiones Latinae Selectae (Dessau, H. (ed.); Berolini: Weidmann, 1962). Itinerarium Egeriae (Peregrinatio Aetheriae) (Heidelberg: C Winter, 1960). John Chrysostom (Budge, E. (trans.); Oxford: John Henry Parker, 1842). Josephus Antiquities (Thackeray, H. St J. and Marcus, R. (eds); Cambridge MA: Harvard University Press, 1934). Josephus Jewish Wars (Thackeray, H. St J. (ed.); Cambridge MA: Harvard University Press, 1928). Julian of Ascalon (Saliou, C. (ed.); Paris: De Boccard, 1996). Libanius (Petit, P. (ed.) Libanius et la vie municipale à Antioche au ivème siècle après J-C; Paris: P Geuthner, 1955). Lucretius De Rerum Natura (Rouse, W. H. D.; London: Heinemann, 1966). Malalas (Jeffreys, E., Jeffreys, M. and Scott, R. The Chronicle of John Malalas; Melbourne: Australian Association for Byzantine Studies, 1986). Palestine Pilgrim Texts (Stewart, A.; London, 1893). Pausanias Description of Greece (Jones, W. H. S. (tr.); London: Heinemann, 1918-1935). Pliny Natural History (Rackham, H.; Cambridge MA: Harvard University Press, 1983). Procopius Buildings (Dewing, H. B.; Cambridge MA: Harvard University Press, 1996). Procopius of Gaza Epistolae et Declamationes (Garzya, A. and Loenertz, R. J. (eds); Eltal: Buch-Kunstverlag, 1963). Pseudo-Meliton Apology (Cureton, W. B. (tr.) Spicilegium Syriacum, containing remains of Bardesan, Meliton, Ambrose and Mara Bar Serapion; London: Francis and John Rivington, 1855). SHA: Scriptores Historiae Augustae (Magie, D.; Cambridge MA: Harvard University Press, 1921). Sozomen Ecclesiastical History (Walford, E. (tr.); London: Henry G Bohn, 1855). Strabo Geography (Jones, H. L.; London: Heinemann, 1930). Varro Res Rusticae (Best, L. S.; London: G Bell and Sons, 1912). Vitruvius De Architectura (Granger, F. S. (tr.); London: Heinemann, 1931).

BIBLIOGRAPHY

189

Xenophon Anabasis (Brownson, C. L. and Todd, O. J. (tr.); London: Heinemann, 1922). Zosimus Historia Nova (Buchanan, J. J. And Davies, H. T. (tr.); San Antonio: Trinity University Press, 1967).

MODERN AUTHORS Abadie-Reynal, C. "Zeugma, rapport préliminaire des campagnes de fouilles de 2000," Anatolia Antiqua 9 (2001), 243-305. _____. "Eaux décoratives, eaux symboliques à Zeugma," Syria 85 (2008), 99-118. Abadie-Reynal, C., Desreumaux, A., Hesse, A., Leriche, P. and Laroche, M.-C. "Mission archéologique de Zeugma, rapport sur la campagne de prospection 1995," Anatolia Antiqua 4 (1996), 311-324. Abadie-Reynal, C. and Ergeç, R. "Mission de Zeugma-Moyenne vallée de l'Euphrate," Anatolia Antiqua 5 (1997), 349-370. _____. "Zeugma-Moyenne vallée de l'Euphrate; rapport préliminaire de la campagne de fouilles de 1997," Anatolia Antiqua 6 (1998), 379-406. Abel, F. M. Géographie de la Palestine, Tome 1: Géographie Physique et Historique (Paris: J. Gabalda, 1933). Adan, D. "The 'Fountain of Siloam' and 'Solomon's Pool' in 1stcentury CE Jerusalem," IEJ 29 (1979), 92-100. Aharoni, Y. Excavations at Ramat Rahel 2: Seasons 1959 and 1960 (Rome: Centro di Studi Semitici, 1962). Aisenstein, B. "The karez, an ancient system of artificial springs," Journal of the Association of Engineers and Architects 8 (1947), 2-3. Al Muheisen, Z. and Tarrier, D. "Water in the Nabataean period," Aram 13-14 (2001-2002), 515-524. Alcock, S. E. The Early Roman Empire in the East (Oxford: Oxbow, 1997). Aliquot, J. and Piraud-Fournet, P. "Le sanctuaire d'Ain el-Fijé et le culte du Barada," Syria 85 (2008), 87-98. Almagro, A. "The survey of the Roman monuments of Amman by the Italian mission in 1930," ADAJ 27 (1983), 607-639. Amery, H. A. and Wolf, A. T. (eds) Water in the Middle East: a Geography of Peace (Austin, Texas: University of Texas Press, 2000). Amit, D. "What was the source of Herodion's water?" Liber Annus 44 (1994), 561-578.

190 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST _____. "The aqueducts of Eleutheropolis (Beth Govrin)," in Amit et al. 2002(a), 168-176. _____. "New data for dating the High-level aqueduct and the Wadi el-Biyar aqueduct and the Herodion aqueduct," in Amit et al. 2002(b), 253-266. _____. "The aqueducts to the fortress of Alexandrion," in Amit et al. 2002(c), 306-312. Amit, D., Patrich, J. and Hirschfeld, Y. (eds) The Aqueducts of Israel, (Portsmouth, RI: JRA, 2002). Andersen, F. G., Nielsen, I. and Holm-Nielsen, S. Gadara: Umm Qes (Wiesbaden: Harroswitz, 1993). Anderson, E. W. The Middle East: Geography and Geopolitics (London and New York: Routledge, 2000). Anderson, J. Roman Architecture and Society (Baltimore: John Hopkins University Press, 1997). André, J. L’Alimentation et La Cuisine à Rome (Paris: Belles Lettres, 1981). Angert, A. "Caesarea coast," ESI 14 (1994), 138f. Ashenberg, K. Clean: an Unsanitised History of Washing (London: Profile, 2008). Avigad, N. "A building inscription of the Emperor Justinian and the Nea in Jerusalem," IEJ 27 (1977), 145-151. _____. Discovering Jerusalem (Nashville: T. Nelson, 1983). Avi-Yonah, M. Quarterly of the Department of Antiquities of Palestine 2 (1932). _____. "The Bath of the Lepers at Scythopolis," IEJ 13 (1963), 325-326. Ayalon, E. "The jar installation of Khirbet Sabiya," IEJ 29 (1979), 175-181. _____. "Yavne-Yam, 'Persian wheel' (saqiya) well," HA 109 (1999), 72*-73*. Ayalon, E., Milo, U. and Ziona, N. "Typology and chronology of water-wheel (saqiya) pottery pots from Israel," IEJ 50/3-4 (2000), 216-226. Bachmann, W., Watzinger, C., Wiegand, T. and Wulzinger, K. Petra (Berlin: Vereinigung Wissenschaftlicher Verleger, 1921). Bagg, A. "Dealing with water rights in the ancient Near East," in Ohlig et al. 2002, 223-232. Baker, S. W. Cyprus as I saw it in 1879 (London: Macmillan and Co., 1879).

BIBLIOGRAPHY

191

Ball, W. Rome in the East: the Transformation of an Empire (London and New York: Routledge, 2000). Balty, J.-C. Guide d'Apamée (Brussels: Centre belge de recherches archéologieques à Apamée de Syrie, 1981). _____. "La maison à Consoles' d'Apamée," AAAS 32 (1982), 9-44. _____. "Apamée (Qal'at el Mudiq, Hama): fouilles belges 19651983," AAAS 33/2 (1983), 259-272. _____. "Problèmes de l'eau à Apamée de Syrie," in Louis, P. and Métral, J. (eds) L'Homme et l'Eau en Mediterranée et au Proche Orient (Lyon: Ed. Du CTHS, 1987), 9-24. _____. "Apamea in Syria in the Second and Third Centuries AD," JRS 78 (1988), 91-104. Barag, D. "An inscription from the High Level aqueduct of Caesarea-reconsidered," IEJ 14 (1964), 250-252. Barker, G. and Jones, B. The UNESCO Libyan valleys survey 19791981: Palaeoeconomy and Environmental Archaeology in the Pre-desert (London: Society for Libyan Studies, 1982). Barker, G., Gilbertson, D. D., Jones B. and Mattingly, D. J. Farming the Desert: the UNESCO Libyan Valleys Archaeological Survey (Paris, Tripoli, London: Society for Libyan Studies, 1996). Barker, G. W., Adams, R., Creighton, O. H., Crook, D., Gilbertson D D, Grattan, J. P., Hunt, C. O., Mattingly, D. J., McLaren, S. J., Mohammed, H. A., Newson, P. A., Palmer, C., Pyatt, F. B., Reynolds, T. E. G. and Tomber, R. "Environment and land use in the Wadi Faynan, southern Jordan: the third season of geoarchaeology and landscape archaeology (1998)," Levant 31 (1999), 255-292. Barker, G. W., Adams, R., Creighton, O. H., Daly, P., Gilbertson D. D., Grattan, J. P., Hunt, C. O., Mattingly, D. J., McLaren, S. J., Newson, P. A., Palmer, C., Pyatt, F. B., Reynolds, T. E. G., Smith, H., Tomber, R. and Truscott, A. J. "Archaeology and desertification in the Wadi Faynan: the fourth (1999) season of the Wadi Faynan landscape survey," Levant 32 (2000), 27-52. Barker, G. W., Adams, R., Creighton, O. H., Gilbertson D D, Grattan, J. P., Hunt, C. O., Mattingly, D. J., McLaren, S. J., Mohammed, H. A., Newson, P. A., Reynolds, T. E. G. and Thomas, D. C. "Environment and land use in the Wadi Faynan, southern Jordan: the second season of geoarchaeology and landscape archaeology (1997)," Levant 30 (1998), 5-26.

192 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST Barker, G. W., Creighton, O. H., Gilbertson D D, Hunt, C. O., Mattingly, D. J., McLaren, S. J. and Thomas, D. C. "The Wadi Faynan Project, southern Jordan: a preliminary report on geomorphology and landscape archaeology," Levant 29 (1997), 19-40. Barker, G., Gilbertson, D. and Mattingly, D. (eds) Archaeology and Desertification: The Wadi Faynan Landscape Survey, Southern Jordan (Oxford: Oxbow Books, 2007). Baroja, J. C. "Norias, azucas, acenas," Revista de Dialectologia y Tradiciones Populares 10 (1954), 29-160. Baruch, Y. "Jerusalem, the Old City," HA 114 (2002), 75*-76*. Bashkin, N. "Khirbet Ni'ana," ESI 13 (1993), 59-61. Baumgarten, Y. "The use of the water-pulley in Iron Age II water systems," in Ohlig et al. 2002, 233-238. Baur, P. V. C., Rostovtzeff, M. I. and Bellinger, A. R. (eds) The Excavations at Dura-Europos: conducted by Yale University and the French Academy of Inscriptions and Letters: Preliminary Report of 1st9th Season of Work (New Haven and London: Yale University Press and OUP, 1929. Beaumont, P. "The qanat: a means of water provision from groundwater sources," in Beaumont et al. 1989, 13-31. Beaumont, P., Blake, H. G. and Wagstaff, J. M. The Middle East: A Geographical Study (New York: Halsted Press, 1998). Beaumont, P., Bonine, M. and McLachlan, K. (eds) Qanat, Kariz and Khattara: Traditional Water Systems in the Middle East and North Africa (London: SOAS and Middle East and North African Studies Press, 1989). Belayche, N. "Pagan festivals in fourth-century Gaza," in BittonAshkelony, B. and Kofsky, A. (eds) Christian Gaza in Late antiquity (Leiden and Boston: Brill, 2004), 5-22. Ben David, C. "The aqueducts of Hippos-Susita," in Amit et al. 2002, 199-206. Benoît, J., Fiches, J.-L., and Gazenbeek, M. "Recherches le long de l’aqueduc de Nîmes dans la basse vallée de l’Alzon (Gard)," in Favory, F. and Fiches, J.-L. (eds) Les Campagnes de la France Méditerranéenne dans l’Antiquité et le Haut Moyen Age: études microrégionales (Paris: Editions de la Maison des sciences de l'homme, 1994), 151-62. Berda, L. "Rosh Ha'ayyin survey," HA 111 (2000), 99*.

BIBLIOGRAPHY

193

Berg, J. E., Sharon, I. and Zilberstein, B. "Water supply, distribution and disposal in Roman Dor," in Amit et al. 2002, 155167. Berger, A. Das Bad in der Byzantinischen Zeit (Munich: Institut für Byzantinistik und neugreichische Philologie der Universität, 1982). Bildirici, M. "Historical dams of Anatolia (Turkey)," in Ohlig et al. 2002, 137-143. Birman, G. and Wienberger, D. "Yaham," HA 113 (2001), 40*-42*. Blackman, D. R. and Peleg, Y. "Two sets of technical comments on P. Kessener, 'The aqueduct at Aspendos'," JRA 14 (2001), 411-414. Blair, I. "Roman London’s waterworks: the Gresham Street discoveries," Current Archaeology 180 (July, 2002), 509-16. Blair, I. and Hall, J. Working Water: Roman Technology in Action (London: Museum of London, 2003). Bodon, G., Riera, I. and Zanovello, P. (eds) Utilitas Necessaria: Sistemi Idraulichi nell'Italia Romana (Milan: Progetto quarto dimensione, 1994). Bonine, M. "Qanats, field systems and morphology: rectangularity on the Iranian plateau," in Beaumont et al. 1989, 35-57. Boraas, R. S. "A preliminary sounding at Rujm el-Malfuf, 1969," ADAJ 16 (1971), 31-45. Bounni, A. and As'ad, K. Palmyre: Histoire, Monuments et Musée (Damascus: s.n., 1989). Bradley, R. The Passage of Arms: An Archaeological Analysis of Prehistoric Hoard and Votive Deposits (Oxford: Oxbow, 1998). Braemer, F. "Prospéctions archéologiques dans le Hawran (Syrie)," Syria 61 (1984), 219-250. Braemer, F., Genequand, D., Dumond Maridat, C., Blanc, P.-M., Dentzer, J.-M., Gazagne, D. And Wech P. "Long-term management of water in the Central Levant: the Hawran case (Syria)," World Archaeology 41/1 (2009), 36-57. Briant, P. Irrigation et Drainage dans l'Antiquité: Qanats et Canalisations Souterraines en Iran, en Egypte et en Grèce (Paris: Thotm, 2001(a)). _____. "Polybe X.28 et les qanats. Le témoignage et ses limites," in Briant 2001(b), 15-40. Bridgeman, J. "Purple dye in late Antiquity and Byzantium," in Herzog et al. 1987, 159-165.

194 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST Brinker, W. "Zur Wasserversorgung von Resafa-Sergiupolis," Damaszener Mitteilungen 5 (1991), 119-146. Brossé, L. "La digue du lac de Homs," Syria 4 (1923), 234-240. Brown, F. E. "The Roman baths," in Rostovtzeff et al. 1936(a), 84105. _____. "Block F3," in Rostovtzeff et al. 1936(b), 49-83. _____. "Houses and shops of the final period," in Rostovtzeff, M. I., Bellinger, A. R., Brown, F. E. and Welles, C. B. (eds) The Excavations at Dura-Europos: Preliminary Report of the Ninth Season of Work 1935-1936, (New Haven: Yale University Press, 1944), 69-158. Browning, I. Jerash and the Decapolis (London: Chatto and Windus, 1982). Brunnow, R. E. and Domaszewski, A. v. Die Provincia Arabia: auf Grund zweier in den Jahren 1897 und 1898 unternommenen Reisen und der Berichte früherer Reisender (Strassburg, 1904). Bruun, C. "Water legislation in the ancient world c. 2200BC – c. AD 500," in Wikander 2000(a), 539-604. _____. "Water shortage and surplus in the ancient world," in Jansen 2000(b), 215-224. Bruun, C. and Saastamoinen, A. (eds) Technology, Ideology, Water: From Frontinus to the Renaissance and Beyond (Rome, 2003). Butcher, K. Roman Syria and the Near East (London: British Museum, 2003). Butler, H. C. and Prentice, W. Révue Archéologique 39 (1901), 62-76. Butler, H. C., Littmann, E., Magie, D. and Stuart, D. R. Syria: Publications of the Princeton University Archaeological Expeditions to Syria in 1904-1905 and 1909 (Leiden: Brill, 1907). Caduff, G. A. Antike Sintflugsagen (Göttingen: Vandenhock and Ruprecht, 1986). Calvet, Y. "Un barrage antique à Ras Shamra," in Geyer 1990, 487499. Calvet, Y. and Geyer, B. Barrages Antiques de Syrie (Paris: Maison de l'Orient, 1992(a)). Campbell, W. A. "The third season of excavation at Antioch on the Orontes," AJA 40 (1936), 1-9. Chehab, M. H. "Mosaiques du Liban," BMB 14 (1957), 8ff. Chesney, F. R. The Expedition for the Survey of the Rivers Euphrates and Tigris Carried on by Order of the British Government on the Years 1835, 1836 and 1837 (London, 1850).

BIBLIOGRAPHY

195

Coleman, K. M. "Launching into history: aquatic displays in the early empire," JRS 83 (1993), 48-74. Conder, C. R. Heth and Moab: Explorations in Syria in 1881 and 1882 (London: Committee of the Palestine Exploration Fund, 1883). Conder, C. R. and Kitchener, H. H. The Survey of Western Palastine: Memoirs of the Topography, Orography, Hydrography and Archaeology. Volume I. Sheets I-VI. Galilee (London: Committee of the Palestine Exploration Fund, 1881). _____. The Survey of Western Palestine: Memoirs of the Topography, Orography, Hydrography and Archaeology. Volume II. Sheets VII-XVI. Samaria, (London: Committee of the Palestine Exploration Fund, 1882). _____. The Survey of Western Palestine: Memoirs of the Topography, Orography, Hydrography and Archaeology. Volume III. Sheets XVIIXXVI. Judaea, (London: Committee of the Palestine Exploration Fund, 1883). Corbier, M. "City, territory and taxation," in Rich, J. and WallaceHadrill, A. (eds) City and Country in the Ancient World (London: Routledge, 1991), 211-39. Corbo, V. "L’Herodium di Giabal Fureidis," Liber Annus (1967), 17: 5-121. Cotterell, B. and Kamminga, J. Mechanics of Pre-Industrial Technology (Cambridge: CUP, 1990). Croke, B. and Crow, J. "Procopius and Dara," JRS 103 (1983), 143159. Crouch, D. P. "The water system of Palmyra," Studia Palmyrenskie 7 (1975), 151-186. Crowfoot, J. W., Kenyon, K. M. and Sukenik, E. L. The Buildings at Samaria (London: Palestine Exploration Fund, 1942). Cumont, F. Etudes Syriennes (Paris: A Picard, 1917). Curtis, V. and Biran, A. "Dirt, Disgust and Disease: Is Hygiene in our Genes?," Perspectives in Biology and Medicine 44/1 (2001), 1731. Curvers, H. H. and Stuart, B. "Bey 008: the 1994 results," BAAL 1 (1996), 228-234. Dalley, S. "Water management in Assyria from the ninth to seventh centuries BC," Aram 13-14 (2001-2), 443-460.

196 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST Dalley, S. and Oleson, J.-P. "Sennacherib, Archimedes, and the water screw: the context of invention in the Ancient World," Technology and Culture 44 (2003), 1-26. Damati, E. "Khirbet el-Mureq," IEJ 22 (1972), 173. _____. "The irrigation systems in the gardens of the Monastery of St Martyrius (Ma'ale Adummim)," in Amit et al. 2002, 438-443. Dar, S. o. "Khirbet Summaqa -1985," ESI 4 (1985), 104-107. _____. Sumaqa: a Roman and Byzantine Jewish Village on Mount Carmel, Israel (Oxford: BAR, 1999). Dauphin, C. "Golan Survey 1983," ESI 3 (1984a), 35-6. _____. "Brothels, baths and babes: prostitution in the Byzantine Holy Land," Classics Ireland 3 (1996), 47-72. Dauphin, C. and Gibson, S. "Golan Survey, 1988," IEJ 41 (1991), 176-179. Dauphin, C. and Schonfield, J. J. "Settlements of the Roman and Byzantine periods on the Golan Heights: preliminary report on three seasons of survey (1979-1981)," IEJ 33 (1983), 189206. Davie, M. F., Makaroun, Y. and Nordiguian, L. "Les qanater Zubayde et l'alimentation en eau de Beyrouthe et de ses environs à l'époque romaine," BAAL 2 (1997), 262-289. De Haan, N. "Si aquae copia patiatur: Pompeian private baths and the use of water," in Koloski-Ostrow 2001, 41-49. De Haan, N. and Jansen, G. C. M. (eds) Cura Aquarum in Campania (Leiden: BABESCH, 1996). De Kleijn, G. The Water Supply of Ancient Rome: City Area, Water and Population (Amsterdam: J. C. Gieben, 2001). De Vaux, R. "Fouilles de Feshka: rapport préliminaire," Révue Biblique 66 (1959), 225-255. _____. "Excavations at Ain Feshka," ADAJ 4-5 (1960), 7-11. _____. Archaeology and the Dead Sea Scrolls (London: OUP, 1973). De Voguë, M. Syrie Centrale, Architecture Civile (Paris, 1865). Decker, M. Agricultural production and trade in Oriens 4th - 7th centuries AD, (University of Oxford, unpublished D Phil. thesis: 2001). Decker, M. "Water into wine: trade and technology in late antiquity," in Lavan, L., Zanini, E. and Sarantis, A. (eds) Technolology in Transition AD300-650 (Leiden: Brill, 2007), 65-92. DeLaine, J. and Johnston, D. E. Roman Baths and Bathing (Portsmouth, R.I: JRA, 1999).

BIBLIOGRAPHY

197

Delpech, A. Les Norias de l'Oronte: Analyse Technologique d'un Élément du patrimoine syrien (Damas: Institut français de Damas, 1997). Dentzer, J. M., Blanc, P. M. and Fournet, T. "Le développement urbain de Bosra de l'époque nabatéenne à l'époque Byzantine: bilan des recherches françaises 1981-2002," Syria 79 (2002), 75-154. Detweiler, A. H. "The palace of the Dux Ripae: the site and the building," in Rostovtzeff, M. I., Bellinger, A. R., Brown, F. E. and Welles, C. B. (eds) The Excavations at Dura Europos: Preliminary Report of the Ninth Season of Work 1935-1936 (New Haven: Yale University Press, 1952), 1-26. Di Segni, L. "The water supply of Roman and Byzantine Palestine in literary and epigraphical sources," in Amit et al. 2002, 37-67. Downey, G. "Greek and Latin inscriptions," in Stillwell, R. (ed.) Antioch on the Orontes II: the Excavations, 1933-1936 (Princeton: Princeton University Press, 1938). _____. "The water supply of Antioch on the Orontes in antiquity," AAAS 1 (1951), 171-187. _____. A History of Antioch in Syria: from Seleucus to the Arab Conquest (Princeton: Princeton University Press, 1961). Downey, S. The Stone and Plaster Sculpture: Excavations at Dura Europos Final Report III, Part I, Fasc. 2 (University of California: University of California Press, 1977). Drijvers, H. J. W. The Book of the Laws of the Countries: Dialogue on Fate of Bardaisan of Edessa (Assen: Van Gorcum, 1965). Dulière, C. Mosaiques des Portiques de la Grande Colonnade: Section VII, 16-17, (Bruxelles: Centre belge de recherches archéologiques à Apamée, 1974). Dumond-Maridat, C. "Les systèmes publics d'alimentation en eau dans le Hauran aux époques classique et byzantine," Syria 85 (2008), 71-86. Dunand, M. "Rapport préliminaire sur les fouilles de Sidon en 1964-1965," BMB 20 (1967), 27-44. Dunand, M. and Saliby, N. Le Temple d'Amrith dans la Pérée d'Aradus (Paris: P Geuthner, 1995). Dunbabin, K. "Baiarum grata voluptas: pleasures and dangers of the baths," PBSR 57 (1989), 1-46. Durand, J-M. "Problèmes d’eau et d’irrigation au royaume de Mari: l’apport des texts anciens," in Geyer 1990, 101-142.

198 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST Dussaud, R. "La digue du lac de Homs et le 'mur egyptien' de Strabon," Monuments et Mémoires Piot 25 (1921-1922), 133-141. Dvorjetski, E. "Social and cultural aspects of medicinal baths in Israel according to Rabbinic sources," in DeLaine and Johnston 1999, 117-129. _____. "Thermo-mineral waters in the eastern Mediterranean basin: historical, archaeological and medicinal aspects," ARAM 13-14 (2001-2), 487-514. Eadie, J. W. "Humayma 1983: regional survey," ADAJ 28 (1984), 212-224. Eadie, J. W. and Oleson, J.-P. "The water supply of Nabataean and Roman Humayma," BASOR 262 (1986), 49-76. Early, R. and Humphrey, J. H. Zeugma: Interim Reports, Rescue Excavations (Packard Humanities Institute), Inscription of Antiochus I, Bronze Statue of Mars, House and Mosaic of the Synãosai, and Recent Work on the Roman Army at Zeugma (Portsmouth, R.I.: JRA, 2003). Eck, W. "Die Wasserversorgung im römischen Reich: Soziopolitische Bedingungen, Recht und Adminstration," Die Wasserversorgung Antiker Städte 2 (1987), 51-88. Edelstein, G. "What's a Roman villa doing outside Jerusalem?" Biblical Archaeology Review 16/6 (1990), 32-44. Ein Gedy, M. "Ashqelon, El-Jura," HA 114 (2002), 90*. Eitram, S. Opferritus und Voropfer der Griechen und Römer Kristiana: J Durybad, 1915). Elderkin, G. W. and Stillwell, R. Antioch-on-the-Orontes (Princeton: Princeton University Press, 1934). Ellis, S. P. "Pooling resources - the use of water for social control in the Roman Empire," in Meadows K. L. and Heron, J. (eds) TRAC 96 (Oxford: Oxbow, 1997). _____. "The seedier side of Antioch," in Sandwell, I. and Huskinson, J. (eds) Culture and Society in Later Roman Antioch (Oxford: Oxbow, 2004), 126-133. Erdkamp, P. "Beyond the limits of the 'consumer city'. A model of the urban and rural economy in the Roman world," Historia 50/3 (2001), 332.356. Ergec, R. "Rescue excavations by the Gaziantep Museum (19921994)," in Kennedy, D. L. (ed.) The Twin Towns of Zeugma on the Euphrates: Rescue Work and Historical Studies (Portsmouth, RI: JRA, 1998), 80-91.

BIBLIOGRAPHY

199

Eschebach, H. "Die Gebrauchswasserversorgung des antiken Pompeji," Antike Welt 13/3 (1979), 21-6. Evenari, M., Shanan, L. and Tadmor, N. The Negev: The Challenge of a Desert (Cambridge, MA: Harvard University Press, 1982). Everman, D. "Survey of the area north of Caesarea and of the aqueducts: preliminary report," in Vann R. L. (ed.) Caesarea Papers: Straton's Tower, Herod's Harbour, and Roman and Byzantine Caesarea (Ann Arbor, MI: JRA, 1992), 181-193. Fagan, G. G. Bathing in Public in the Roman World (Ann Arbor: University of Michigan Press, 1999). _____. "Hygienic conditions in Roman public baths," in Jansen 2000, 281-288. Fahlbusch, H. "Die Wasserversorgung des antiken Beth She'an," in Ohlig et al. 2002, 55-63. Feissel, D. "Deux listes de quartiers d'Antioche astreints au creusement d'un canal: AD 73-74," Syria 62 (1985), 77-103. Fentress, E. Numidia and the Roman Army: Social, Military and Economic Aspects of the Frontier Zone (Oxford: BAR, 1979). Février, P-A. "Urbanisation et urbanisme de l’Afrique romaine," Aufstieg und Niedergang der Römischen Welt 2/10/2 (1982), Principat: 321-96. Février, P-A. "Armée et aqueducs," in Boucher, J.-P. (ed.) Journées d'Etudes sur les Aqueducs Romains (Paris, 1983), 133-40. Finkelstein, I. Living on the Fringe: the Archaeology of the Negev, Sinai and Neighbouring Regions in the Bronze and Iron Ages (Sheffield: Sheffield Academic Press, 1995). Finley, M. I. "Technical innovation and economic progress in the ancient world," Economic History Review 18 (1965), 29-45. _____. The Ancient Economy (California: University of California Press, 1999). Fischer, M. L. and Shacham, T. "The water systems of the En Boqeq oasis," in Amit et al. 2002, 402-408. Fisher, C. S. "Yale University - Jerusalem School expedition at Jerash: first campaign," BASOR 40 (Dec. 1930), 2-11. _____. "The baths," in Elderkin and Stillwell 1934, 4-31. _____. "Description of the site," in Kraeling 1938(a), 21-22. _____. "The Baths of Placcus," in Kraeling 1938(b), 265-269. Fisher, W. B. The Middle East: A Physical, Social and Regional Geography (London: Methuen, 1971).

200 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST Fitzpatrick, A. P. "Roman Britain in 2001: Greater London," Britannia 33 (2002), 326-40. Flinder, A. "A piscina at Caesarea: a preliminary survey," IEJ 26 (1976), 77-80. Foerster, G. "Herodium," Revue Biblique 77 (1970), 400-1. Foerster, G. and Tsafrir, Y. "Nysa-Scythopolis - a new inscription and the titles of the city on its coins," Israel Numismatic Journal 9 (1986-7), 53. _____. "Beat Shean project - 1988: Hebrew University Expedition," ESI 7-8 (1988-1989), 15-22. _____. "City Centre (North): excavations of the Hebrew University Expedition," ESI 11 (1992), 3-32. Forbes, R. J. Studies in Ancient Technology vol. I, (Leiden: Brill, 1964). Freyberger, K. S. Frühkaiserzeitlichen Heiligtümer de Karawanenstationem im Hellenisierten Osten: Zeugnisse eines Kulturellen Konflikts im Spannungsfeld Zweier Politischer Formationem (Mainz: P. von Zabern, 1998). Frösen, J., Fiema, Z. T., Haggren, H., Koistinen, K., Lavento, M. and Peterman, G. L. "The Finnish Jabal Harun project: report on the 1997 season," ADAJ 42 (1998), 483-502. Frösen, J., Fiema, Z. T., Lavento, M., Koistinen, K. and Holmgren, R. "The 1998 Finnish Jabal Harun project: a preliminary report," ADAJ 43 (1999), 369-410. Frumkin, A. "The Holocene history of the Dead Sea levels," in Niemi, T. M., Ben-Avraham, Z., and Gat, J. R. (eds) The Dead Sea: The Lake and its Setting (New York and Oxford: Oxford University Press, 1997), 237-48. _____. "The hydrogeology of Israel and the problem of water supply in antiquity," in Amit et al. 2002(a), 21-24. _____. "The water-supply network of Samaria-Sebaste," in Amit et al. 2002(b), 267-277. Frumkin, A., Greenbaum, N. and Schick, A. P. "Palaeohydrology of the northern Negev: comparative evaluation of two catchments," in Issar, A. S. and Brown, N. (eds) Water, Environment and Society in Times of Climatic Change (Dordrecht: Kluwer Academic Publishers, 1998), 97-111. Frymer-Kensky, T. "Pollution, purification and purgation in Biblical Israel," in Meyers, C. L. and O'Connor, M. (eds) The Word of the Lord Shall Go Forth: Essays in Honour of David Noel Freed-

BIBLIOGRAPHY

201

man in Celebration of his Sixtieth Birthday (Winona Lake, Indiana: Eisenbrauns, 1983), 399-414. Fulford, M. "Links with the past: pervasive “ritual” behaviour in Roman Britain," Britannia 32 (2001), 199–218. Furlan, I. "Cisterne a Dara," in Iacobini, A. and Zanini, E. (eds) Arte Profana e Arte Scara a Bisanzio (Roma: Argos, 1995), 51-65. Galor, K. "Qumran’s plastered installations: cisterns or immersion pools?" in Ohlig et al. 2002, 33-46. Garbrecht, G. (ed.) Historische Talsperren (Stuttgart: Wittwer, 1987). _____. Historische Talsperren 2 (Stuttgart: Wittwer, 1991(a)). _____. "Talsperre und Tunnel am Hafen Seleukeia," in Garbrecht 1991(b), 83-89. _____. "Der Staudamm von Resafa-Sergiopolis," in Garbrecht 1991(c), 238-248. Garbrecht, G. and Vogel, A. "Die Staumauern von Dara," in Garbrecht 1991, 90-95. Garbrecht, G. and Peleg, Y. "The water supply of the desert fortresses in the Jordan Valley," Biblical Archaeologist 57/3 (1994), 160-170. Gardiner, M. and McQuitty, A. "A watermill in Wadi el-Arab, north Jordan and watermill development," PEQ 119 (1987), 24-32. Gawlikowski, M. "A residential area by the south decumanus," in Zayadine 1986, 107-121. Gazenbeek, M. "Interaction entre aqueduc et habitat rural: deux cas d'étude en France méditerranéenne: Nîmes et Arles" in Jansen 2000, 225-230. Getzov, N. "Horvat Uza," ESI 13/100 (1993), 19-21. Geyer, B. Techniques et Pratiques Hydro-agricoles Traditionelles en Domaine Irrigue. Approche Pluridisciplinaire des Modes de Culture avant la Motorisation en Syrie (Paris: P. Geuthner, 1990). _____. "Des fermes byzantines et palais ommeyades," in Nordiguian, L. and Salles, J.-F. (eds) Aux Origines d'Archéologie Aerienne (Beirut, 2000), 109-22. Gibson, S. "The Sataf project of landscape archaeology and the Judean hills: a preliminary report on four seasons of survey and excavation (1987-1989)," Levant 23 (1991), 23-54. Gibson, S. and Edelstein, G. "Investigating Jerusalem's rural landscape," Levant 17 (1985), 139-155.

202 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST Gilbertson, D. D. and Hunt, C. O. "Romano-Libyan agriculture: walls and floodwater farming," in Barker 1996, 191-227. Gilbertson, D. D. and Kennedy, D. L. "An archaeological reconnaissance of water harvesting structures and wadi walls in the Jordanian desert, north of Azraq oasis," ADAJ 28 (1984), 151-162. Ginzburg, A. "Yafo, 3022 Street," HA 111(2000), 42*. Girard, F., Roumi, M. and Robine, G. "Les norias hydrauliques du Moyen-Oronte, partimoine syrien: étude d'une technologie en voie de disparition," in Geyer 1990, 367-382. Glaser, F. "Fountains and nymphaea," in Wikander 2000, 413-451. Glueck, N. "Jerash in the spring of 1934: preliminary report of the Joint Expedition of the School of Jerusalem and Yale University," BASOR 53 (Feb. 1934), 2-13. _____. Explorations in Eastern Palestine II (New Haven: ASOR, 1935). _____. Explorations in Eastern Palestine III (New Haven: ASOR, 1937-1939). _____. The River Jordan: Being an Illustrated Account of Earth's Most Storied River (London: Lutterworth, 1946). _____. Explorations in Eastern Palestine IV (New Haven: ASOR, 1951). _____. Rivers in the Desert: The Exploration of the Negev, an Adventure in Archaeology (London: Weidenfeld and Nicholson, 1959). _____. The Other Side of the Jordan (Cambridge, MA: ASOR, 1970). Goblot, H. Les Qanats: Une Technique d'Aquisition de l'Eau (Paris: Mouton, 1979). Goossens, G. Hiérapolis de Syrie: Essai de Monographie Historique (Louvain: Bibliothèque de l'Université, 1943). Green, A. R. W. The Storm-God in the Ancient Near East (Winona Lake, Ind.: Eisenbrauns, 2003). Green, J. and Tsafrir, Y. "Greek inscriptions from Hammat Gader: a poem by the Empress Eudocia and two building inscriptions," IEJ 32 (1982), 77-96. Gsell, S. Enquête Administrative sur les Travaux Hydraliques Anciens en Algérie (Paris: Imprimérie Nationale, 1902). Guz-Zilberstein, B. "The typology of the Hellenistic coarseware and selected loci of the Hellenistic and Roman periods," in Stern, E. (ed.) Excavations at Dor, Final Report. Areas A and C: the Finds (Jerusalem: IES, 1995), 289-433.

BIBLIOGRAPHY

203

Haas, V. Geschichte der Hethitischen Religion (Leiden: Brill, 1994). Haddad, E. and Velednizki, N. "Gelilot," HA 115 (2003), 37*-38*. Hajjar, Y. "Divinités oraculaires et rites divinatoires en Syrie et en Phénicie à l'époque gréco-romaine," ANRW II.18.4 (1990), 2237-2783. Haidt, J., Rozin, P., McCauley, C. and Imada, S. "Body, Psyche and Culture: The Relationship between Disgust and Morality," Psychology and Developing Societies 9 (1997), 107-131. Hammond, P. C. The Nabataeans - their History, Culture and Archaeology (Gothenburg: P Astrom, 1973). Harding, G. L. The Antiquities of Jordan (London: LutterworthPress, 1967). Harper, R. P. "Neocaesareia - Qasrin - Dibsi Faraj," in Margueron, J. C. (ed.) Le Moyen Euphrate: Actes du Colloque de Strasbourg 1012 Mars 1977 (Leiden:Brill, 1980), 327-348. Harper, R. P. and Wilkinson, T. "Excavations at Dibsi Faraj, northern Syria 1972-1974: a preliminary note on the site and its monuments," Dumbarton Oaks Papers 29 (1975), 319-338. Hartal, M. "Banias aqueduct," ESI 13 (1993), 1f. _____. "Banias, the aqueduct," ESI 16 (1996), 5-7. _____. "The aqueduct to Banias," in Amit et al. (2002), 88-103. Healy, J. F. Mining and Metallurgy in the Greek and Roman World (London: Thames and Hudson, 1978). Hedges, R. E. M., Hously, R. A., Pettitt, P. B., Bronk Ramsey, C. and Van Klinken, G. J. "Radiocarbon dates from the Oxford AMS system: archaeometry datelist 21," Archaeometry 38/1 (1996), 181-207. Heim, C., Nowacyzk, N. R. and Negendank, J. F. W. "Near East desertification: evidence from the Dead Sea," Naturwissensschaften 84, 398-401. Heimann, M. "Qadesh Barne'a survey," ESI 3 (1984), 89-90. Henig, M. and Whiting, M. Engraved Gems from Gadara in Jordan: The Sa’d Collection of Intaglios and Cameos (Oxford: Oxford University Committee for Archaeology, 1987). Henry, D. O. "The impact of radiocarbon dating on Near Eastern Prehistory," in Taylor, R. E., Long, A. and Kra, R. S. (eds) Radiocarbon After Four Decades: An Interdisciplinary Perspective (New York: Springer-Verlag, 1992), 324-334. Herzog, I., Spanier, E. and Ron, M. The Royal Purple and the Biblical Blue: Argaman and Tekhelet: The Study of Chief Rabbi Dr. Isaac

204 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST Herzog on the Dye Industries in Ancient Israel and Recent Scientific Contributions (Jerusalem: Keter, 1987). Hidiroglu, P. "L’eau et les bains à Qoumran," Revue des Études Juives 159 (2000), 19-47. Hild, F. and Hellenkemper, H. Kilikien und Isaurien (Wien: Verlag der Osterreischischen Akademie der Wissenschaften, 1990). Hirschfeld, Y. The Judaean Desert Monasteries in the Byzantine Period (New Haven: Yale University Press, 1992). _____. The Roman Baths of Hammat Gader: Final Report (Jerusalem: IES, 1997). _____. Ramat Hanadiv Excavations: Final Report of the 1984-1998 Seasons (Jerusalem: IES, 2000). _____. "The aqueducts of Emmaus-Nicopolis," in Amit et al. (2002a), 187-198. _____. "The aqueducts at Ramat Hanadiv and Shuni," in Amit et al. (2002b), 387-400. _____. "A climatic change in the early Byzantine period? Some archaeological evidence," PEQ 136/2 (2004), 133-149. Hirschfeld, Y. and Birger, R. "Early Roman and Byzantine estates near Caesarea," IEJ 41 (1991), 81-111. Hitchner, R. "Irrigation, terraces, dams and aqueducts in the region of Cillium (mod. Kasserine). The role of water works in the development of a Roman-African town and its countryside," in Trousset, P. (ed.) Productions et Exportations Africaines: Actualitées Archéologiques en Afrique du Nord Antique et Médievale (Paris: Ed du CTHS, 1995), 143-157. Hodge, A. T. (ed.) Future Currents in Aqueduct Studies (Leeds: Francis Cairns, 1991). _____. Roman Aqueducts and Water Supply (London: Thames and Hudson, 1992). _____. "Anomalies in the flow at the Pompeii castellum," in De Haan and Jansen (1996), 13-18. _____. "Wells," in Wikander (2000a), 29-33. _____. "Aqueducts," in Wikander (2000b), 39-65. _____. "Reservoirs and dams," in Wikander (2000c), 331-339. Holm-Nielsen, S., Nielsen, I. and Andersen, F. G. "The excavation of Byzantine baths in Umm Qeis," ADAJ 30 (1986), 219-232. Honari, M. "Qanats and human ecosystems in Iran," in Beaumont et al. (1989), 61-85.

BIBLIOGRAPHY

205

Hopkins, C. "Private houses," in Rostovtzeff, M. I. (ed.) Excavations at Dura Europos: Preliminary Report of the Fifth Season of Work (New Haven: Yale University Press, 1934), 31-72. _____. "House in Block E4," in Rostovtzeff 1936, 4-48. Horsfield, G. Jerash - Annual Report of Works of Conservation 19251926 (Government of Trans-Jordan Antiquities Bulletin 2, 1926). Horton, F. L. "A sixth-century bath in Caesarea's suburbs and the transformation of bathing culture in late antiquity," in Raban, A. and Holum, K. G. (eds) Caesarea Maritima, A Retrospective After Two Millennia (Leiden: Brill, 1996), 177-189. Hoss, S. Baths and Bathing: The Culture of Bathing and the Baths and Thermae in Palestine from the Hasmoneans to the Moslem Conquest; with an Appendix on Jewish Ritual Baths (miqva’ot) (Oxford: BAR, 2005). Hunt, C. O., Mattingly, D. J., Gilbertson, D. D., Dore, J. N., Barker, G. W., Burns, J. R. and Van der Veen, M. "ULVS XIII: interdisciplinary approaches to ancient farming in the Wadi Mansur, Tripolitania," Libyan Studies 17 (1986), 7-47. Ilan, Z. and Amit, D. "The aqueduct of Qumran," in Amit et al. 2002, 380-386. Iliffe, J. H. "A building inscription from the Syrian limes," QDAP 10 (1944), 62-4. Ionides, M. G. and Blake, G. S. Report on the Water Resources of TransJordan and their Development (London: E & F N Spon, 1939). Israel, Y. "Ashqelon," ESI 13 (1993), 100-105. Issar, A.S. Climate Changes in the Holocene and their Impact on Hydrological Systems (Cambridge: Cambridge University Press, 2003). Jackson, R. "Spas, water and hydrotherapy in the Roman world," in DeLaine and Johnston 1999, 107-116. Jacobsen, T. and Lloyd, S. Sennacherib's Aqueduct at Jerwan (Chicago: University of Chicago Press, 1935). Jalabert, L. "Inscriptions grecques et latines de Syrie (deuxième série)," Mélanges de la Faculté Orientale de Beyrouth 2 (1907), 173421. Jansen, G. C. M. (ed.) Cura Aquarum in Sicilia (Leiden: BABESCH, 2000(a)). _____. "Urban water transport and distribution," in Wikander 2000 (b), 103-125.

206 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST _____. "Studying Roman hygiene: the battle between the optimists and the pessimists," in Jansen 2000(c), 275-280. _____. "Water pipe systems in the houses of Pompeii," in KoloskiOstrow 2001, 27-40. Jashemski, W. F. "The use of water in Pompeian gardens," in De Haan and Jansen 1996, 51-58. Johnson, B. L. "The pottery," in Weinberg, G. D. (ed.) Excavations at Jalame (Columbia: University of Missouri Press, 1988), 137226. Jones, A. H. M. The Later Roman Empire 284-602 AD: A Social, Economic and Administrative Survey (Oxford: Blackwell, 1964). Jones, E. "The symbolic significance of salt in folklore and superstition," in Jones, E. (ed.) Essays in Applied Psycho-Analysis (London: The International Psycho-Analytical Press, 1923), 112-203. Jones, G. D. B. "Capena and the Ager Capenas," PBSR 30 (1962), 116-207. Jones, R. and Robinson, D. "Water, wealth and social status at Pompeii: the House of the Vestals in the first century," AJA 109 (2005), 695-710. Kaiser, T. The Religious Life of Palmyra (Stuttgart: Franz Steiner Verlag, 2002). Kamash, Z. Water Supply and Management in the Near East 63 BC – AD 636 (Oxford University: Unpublished DPhil thesis, 2006a). http://ads.ahds.ac.uk/catalogue/library/theses/kamash_2006/ _____. "Recent research on Roman dams in the Near East," in Wiplinger, G. (ed.) Cura Aquarum in Ephesus (Dudley, Mass.: Peeters, 2006b), 233-238. _____. "What lies beneath? Perceptions of the ontological paradox of water," World Archaeology 40/2 (2008), 224-237. _____. "Perceptions of latrines: socio-cultural, moral and religious issues," in Jansen, G., Koloski-Ostrow, A. and Moorman, E. (eds) Roman Toilets: Their Archaeology and Cultural History (Leiden: BABESCH, forthcoming 2010a). _____. "Which way to look? Exploring latrines in the Roman world," in Molotch, H. and Norén, L. (eds) Toilet: The Public Restroom and the Politics of Sharing (New York: NYU Press, forthcoming 2010b).

BIBLIOGRAPHY

207

Karmon, N. and Spanier, E. "Archaeological evidence of the purple dye industry from Israel," in Herzog et al. 1987, 147-158. Kedar, Y. "Ancient agriculture at Shivtah in the Negev," IEJ 7 (1957), 178-189. Kennedy, D. L. The Twin Towns of Zeugma on the Euphrates: Rescue Work and Historical Studies (Portsmouth, RI: JRA, 1998). Kennedy, D. L. and Bewley, R. Ancient Jordan from the Air (London: CBRL, 2004). Kerner, S. "Umm Qays - Gadara: a preliminary report 1993-1995," ADAJ 41 (1997), 283-302. Kerner, S. and Hoffmann, A. "Gadara - Umm Qeis: preliminary report on the 1991 and 1992 seasons," ADAJ 37 (1993), 359384. King, G. "Preliminary report on a survey of Byzantine and Islamic sites in Jordan 1980," ADAJ 26 (1982), 85-95. King, G., Lenzen, C. J. and Rollefson, G. O. "Survey of Byzantine and Islamic sites in Jordan, second season report, 1981," ADAJ 27 (1983), 385-436. Kirk, G. E. "Archaeological exploration in the southern desert," PEQ 70 (1938), 211-235. Klein, C. Fluctuations of the level of the Dead sea and climatic fluctuations in Erez-Israel during historical times (The Hebrew University of Jerusalem: unpublished PhD thesis, 1986). Kloner, A. "Ancient agriculture at Mamshit and the dating of the water diversion systems in the Negev," Eretz-Israel 12 (1975), 124*; 167-170. Kobori, I. "Les qanats en Syrie," in Geyer 1990, 321-328. Kobori, I. and Endo, K. Qanawat Romani of Taibe Oasis (Tokyo: Tokyo University Scientific Mission, 1980). Köhler, J. "Research on Roman thermo-mineral baths - from Italy to Israel," in Ohlig et al. 2002, 295-306. Koloski-Ostrow, A. O. "Finding social meaning in the public latrines of Pompeii," in De Haan and Jansen 1996, 79-86. _____. "Cacator cave malum: the subject and object of Roman public latrines in Italy during the first centuries BC and AD," in Jansen 2000(a), 289-296. _____. Water Use and Hydraulics in the Roman City (Dubuque, Iowa: Kendall/Hunt Publising Company, 2001).

208 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST _____. "Water, water (literally) everywhere: Turkey, Italy, Greece, Near East and Mesopotamia, North Africa, Gallia, Germania, Spain," JRA 21/2 (2008), 549-556. Koloski-Ostrow, A. O., Crouch, D. P., De Haan, N., Bruun, C., Leigh, S. and Wilson, A. I. "An epilogue: where to go from here? Future courses for water studies," in Koloski-Ostrow 2001, 97-104. Kopp, C. and Stève, A.-M. "Le desert de Saint Jean près d’Hébron," Révue Biblique 53 (1946), 559-75. Kraeling, C. H. Ptolemais: City of the Libyan Pentapolis (Chicago: ASOR, 1962). Kraemer, C. Excavations at Nessana (Princeton: Princeton University Press, 1958). Krencker, D. M. and Zschietzscmann, W. Römische Tempel in Syrien (Berlin and Leipzig: W de Gruyter and Co., 1938). Krug, A. Heilkunst und Heilkult. Medizin in der Antike (Munich: CH Beck, 1985). Lacoste, H. "La VIIème campagne de fouilles à Apamée," L'Antiquité Classique 10 (1941), 115-121. Lambton, A. K. S. "The origin, diffusion and functioning of the qanat, " in Beaumont et al. 1989, 5-12. Lassøe, J. "The irrigation system at ULHU 8th century BC," Journal of Cuneiform Studies 5 (1951), 21-32. Lassus, J. "Une villa de plaisance à Yakto," in Stillwell, R. (ed.) Antioch on the Orontes: The Excavations 1933-36 (Princeton: Princeton University Press, 1938), 95-147. _____. "L'eau courante à Antioche," in Boucher, J.-P. (ed.) Journées d'Études sur les Aqueducs Romains (Paris 1983), 207-229. Lassus, J. and Stillwell, R. Les Portiques d'Antioche (Princeton: Princeton University Press, 1972). Lauffray, J. "Forums et monuments de Béryte," BMB 7 (1944-5), 13-80. _____. Halabiyya-Zenobia: Place Forte du Limes Oriental et la HauteMesopotamie au VIe Siècle (Paris: P Geuthner, 1983). _____. Halabiyya-Zenobia: Place Forte du Limes Oriental et la HauteMesopotamie au VIe Siècle (Paris: P Geuthner, 1991). Lavan, L. "What killed the ancient city? Chronology, causation and traces of continuity," JRA 22/2 (2009), 803-812. Lavan, L. and Bowden, W. Recent Research in Late-Antique Urbanism (Portsmouth, RI: JRA, 2001).

BIBLIOGRAPHY

209

Lavento, M. and Huotari, M. "A water management system around Jabal Harun, Petra - its design and significance," in Ohlig et al. 2002, 93-106. Le Bas, P. and Waddington, W. H. Inscriptions grecques et Latines Recueillies en Grèce et en Asie Mineure (Paris: F Didot, 1853). Le Strange, G. Palestine Under the Moslems: a Description of Syria and the Holy Land from AD 650 to 1500, tr. from the Works of the Medieval Arab Geographers by G le Strange (London: PEF, 1890). Leblanc, J. and Poccardi, G. "Étude de la permanence antiques de traces urbains et ruraux antiques à Antioche-sur-l'Oronte," Syria 76 (1999), 91-110. Lehmann, C. H. and Holum, K. G. The Greek and Latin Inscriptions of Caesarea Maritima (Boston: ASOR, 2000). Lepaon, T. "Les édifices balnéaires de Gerasa de la Décapole: premières observations," Syria 85 (2008), 51-70. Leveau, P. "À quoi servaient les aqueducs romains?," L'Histoire 105 (1987), 96-104. _____. "Research on Roman aqueducts in the past ten years," in Hodge 1991, 149-162. _____. "L'aqueduc de Nîmes et les aqueducs antiques," in Fabre, G., Fiches, J-L, Paillet, J-L. (eds) L'Aqueduc de Nîmes et le Pont du Gard: Archéologie, Geosystème et Histoire (Nîmes: Conseil Général de Nîmes, 1992), 223-250. Leveau, P. and Paillet, J.-L. L’Alimentation en Eau de Caesarée de Mauretanie et l’Aqueduc de Cherchel (Paris: Editions l'Harmattan, 1976). Levi, D. Antioch Mosaic Pavements (Princeton: Princeton University Press, 1947). Levine, L. I. and Netzer, E. "Excavations at Caesarea Maritima 1975, 1976, 1979 – Final Report," Qedem 21 (1986). Lewis, M. Millstone and Hammer: The Origins of Water Power (Hull: University of Hull, 1997). _____. "Vitruvius and Greek aqueducts," PBSR 67 (1999), 145172. _____. "Theoretical hydraulics, automata and water clocks," in Wikander 2000, 343-369. Liebeschuetz, J. H. W. G. Antioch: City and Imperial Administration in the Later Roman Empire (Oxford: Clarendon Press, 1972). _____. "The end of the ancient city," in Rich, J. (ed.) The City in Late Antiquity (London: Routledge, 1992), 1-49.

210 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST Lightfoot, D. R. "Syrian qanat romani: history, ecology, abandonment," Journal of Arid Environments 33 (1996), 321-336. _____. "Qanats in the Levant: hydraulic technology at the periphery of early empires," Technology and Culture 38/2 (1997), 432451. _____. "The origin and diffusion of qanats in Arabia: new evidence from the northern and southern peninsula," The Geographical Journal 166 (2000), 1-20. Lightfoot, J. (ed.) Lucian, on the Syrian Goddess (Oxford: OUP, 2003). _____. "The Apology of Pseudo-Meliton," Studi Epigrafici e Linguistici 24 (2007), 59-110. Loffreda, S. "Preliminary report on the second season of excavations at Qal'at el Mishnaqa, Machaerus," ADAJ 25 (1981), 8594. Lohmann, H. Beobachtungen zum Stadtplan von Timgad, Wohnungsbau im Altertum. Bericht über ein Kolloquium veranstaltet vom ArchitekturReferat des Deutschen Archäologischen Instituts in Berlin vom 21.11 bis 23.11 1978 (Berlin, 1979). Lux, U. "Der Mosaikfussboden eines spätantiken Bades in Umm Qes," ZDPV 82 (1966), 64-73. Macadam, H. I. "Epigraphy and village life in southern Syria during the Roman and Byzantine periods," Berytus 31 (1983), 103-115. Macalister, R. A. S. Gezer I-III. The Excavations of Gezer, 1902-1909 (Jerusalem; Hebrew Union College, 1912). Madella, M., Osborne, R. and Shaw, J. (eds) The Archaeology of Water. World Archaeology 41/1 (2009). Mango, C. Architettura Bizantina (Milan: Electa, 1974). Mango, M. Artistic patronage in the Roman diocese of Oriens 313641 AD (University of Oxford: Unpublished DPhil. Thesis, 1984). _____. "Fishing in the desert," Palaeoslavica 10/1 (2002), 323-330. _____. "Baths, reservoirs and water use at Androna in late antiquity and the early Islamic period," in Bartl, K. and al-Razzaq Moaz, A. (eds) Residences, Castles, Settlements. Transformation Processes from Late Antiquity to Early Islam in Bilad al-Sham (Rahden: Deutsches Archäologisches Institut, 2009), 73-88. Masterman, E. W. G. "The water supply of Damascus," QDAP 52/4 (1920), 181-187. Matheson, S. B. "The tenth season at Dura Europos 1936-1937," Syria 69 (1992), 121-140.

BIBLIOGRAPHY

211

Matsuoka, M., Izuma, S., Budiawan, T., Nakata, N. and Saeki, K. "Mycobacterium leprae DNA in daily using water as a source of leprosy infection," Indian Journal of Leprosy 71/1 (1999), 617. Matthews, J. F. "The tax law of Palmyra: evidence for economic history in a city of the Roman East," JRS 74 (1984), 157-180. Mattingly, D. J. and Salmon, J. Economies Beyond Agriculture in the Classical World (London: Routledge, 2000). Mattingly, G. L., Pace, J. H., Stephenson, R. A. and Wagnon, E. P. "The water catchment system of Nakhl, Jordan," ADAJ 42 (1998), 331-339. Maundrell, M. A Journey from Aleppo to Jerusalem in 1697 (Oxford, 1749). Mauss, M. (tr. Cunnison, I.) The Gift (New York: Norton and Co., 1966). Mayence, F. "La troisième campagne de fouilles à Apamée," Bulletin des Musées Royaux d'Art et d'Histoire 5 (1933), 1-6. Mayerson, P. "Ancient agricultural remains in the central Negeb: the teleilat el-anab," BASOR 153 (1959), 19-31. _____. The Ancient Agricultural Regime of Nessana and the Central Negeb (London: British School of Archaeology in Jerusalem, 1960(a)). _____. "The ancient agricultural remains of the central Negeb: methodology and dating criteria," BASOR 160 (1960(b)), 2737. _____."The city of Elusa in the literary sources of the 4th-6th centuries," IEJ 33 (1983), 247-253. _____. "Choricius of Gaza on the water supply system of Caesarea," IEJ 36 (1986), 269-272. Mazar, A. "A survey of the aqueducts to Jerusalem," in Amit et al. 2002(a), 210-243. _____. "Appendix: The aqueduct to Herodion," in Amit et al. 2002(b), 243-244. Mazar, B. The Mountain of the Lord (New York: Doubelday, 1975). Mazar, B. and Dunayevsky, I. "En Gedi: third season of excavations: preliminary report," IEJ 14 (1964), 121-130. Mazor, G. "Beat Shean project - 1988: Department of Antiquities expedition," ESI 7-8 (1988-1989), 22-32.

212 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST Mazor, G. and Bar-Nathan, R. "The Bet Shean excavation project 1992-1994: Antiquity Authority expedition," ESI 17 (1998), 536. McQuitty, A. "Watermills in Jordan: technology, typology, dating and development," in Amr, K., Zayadine, F. and Zaghloul, M. (eds) Studies in the History and Archaeology of Jordan 5 (Amman: Dept of Antiquities, 1995), 745-751. Meshel, A. and Amit, D. "The water-supply systems of Cypros fortress," in Amit et al. 2002, 313-329. Métral, F. "Périmètres irrigués d'état sur l'Euphrate syrien: modes de gestion et politique agricole," in Louis, P. and Métral, F. (eds) L'Homme et l'Eau en Mediterranée et au Proche Orient IV: l'Eau dans l'Agriculture (Lyons: Ed. du CTHS, 1987), 111-145. Meyers, E. M. "Aspects of everyday life in Roman Palestine with special reference to private domiciles and ritual baths," in Bartlett, J. R. (ed.) Jews in the Hellenistic and Roman Cities (London: Routledge, 2002), 211-220. Millar, F. "The Roman coloniae of the Near East: a study of cultural relations," in Kajava, M. and Solin, H. (eds) Roman Eastern Policy and Other Studies in Roman History (Helsinki: Finnish Society of Sciences and Letters, 1990), 7-58. _____. The Roman Near East 31 BC - AD 337 (Cambridge, MA; London: Harvard University Press, 1993). _____. "Narrative and identity in mosaics from the late Roman Near East: pagan, Jewish and Christian," in Eliav, Y. Z., Friedland, E. A. and Herbert, S. (eds) The Sculptural Environment of the Roman Near East: Reflections on Culture, Ideology and Power (Leuven: Peeters, 2008), 225-256). Miller, R. "Water use in Syria and Palestine from the Neolithic to the Bronze Age," World Archaeology 11/3 (1980), 331-341. Miller, S. S. "Stepped pools and the non-existent monolithic 'Miqveh'," in Edwards, D. and McCullough, T. (eds) The Archaeology of Difference: Gender, Ethnicity, Class and the 'Other' in Antiquity (Boston: ASOR, 2007), 215-234. _____. "Stepped pools, stone vessels, and other identity markers of 'complex common Judaism'," Journal for the Study of Judaism 41 (2010), 214-243. Mougdad, S. and Makowski, C. "Nymphée de Bosra et ses abords: recherches archéologiques en 1979. Rapport préliminaire," AAAS 33/1 (1983), 35-46.

BIBLIOGRAPHY

213

Mouterde, R. and Poidebard, A. (eds) Le Limes de Chalcis: Organisation de la Steppe en Haute Syrie Romaine (Paris: P Geuthner, 1945). Nadelman, Y. "Close-up of a 22 km long aqueduct: the Low Level Aqueduct at Ramot Rahel in Jerusalem," in Jansen 2000, 159162. Nau, F. (ed. and trans.) "Notice historique sur le monastère de Qartamain," Actes du XIV Congres International des Orientalistes, 2 (1907), 1-75. Negev, A. "The High Level Aqueduct at Caesarea," IEJ 14 (1964), 237-249. _____. The Nabatean Potter's Workshop at Oboda (Bonn: Habelt, 1974). Negev, N. "Nahal Hevron - dams," ESI 16 (1996), 128-131. _____. "Horbat Kohal," ESI 19 (1999), 88*. Netzer, E. "The Winter Palaces of the Judaean kings at Jericho at the end of the Second Temple Period," BASOR 228 (1977), 113. _____. Masada III: the Yigael Yadin Excavations 1963-1965: Final Reports. The buildings, Stratigraphy and Architecture (Jerusalem: IES, 1991). _____. Hasmonean and Herodian Palaces at Jericho: Final Reports of the 1973-1987 Excavations, Volume 1 (Jerusalem: IES, 2001). Netzer, E. and Garbrecht, G. "Water channels and a royal estate of the late Hellenistic period in Jericho's western plains," in Amit et al. 2002, 366-379. Neudecker, R. Die Pracht der Latrine: Zum Wandel öffentlicher Bedürfnisanstalten in der Kaiserzeitlichen Stadt (München: F Pfeil, 1994). Newson, P. A. Settlement, land use and water management systems in Roman Arabia: an integrated archaeological approach (University of Leicester: unpublished PhD thesis, 2002). Nielsen, I. Thermae et Balnea: The Architecture and Cultural History of Roman Public Baths (Aarhus: Aarhus University Press, 1993). _____. "Early provincial baths and their relations to early Italic baths," in DeLaine and Johnston 1999, 35-44. Northedge, A., Bennett, C. M. and Bowsher, J. Studies on Roman and Islamic Amman: The Excavations of Mrs C-M Bennett and Other Investigations (Oxford: OUP, 1992). Ofer, A. and Porath, Y. "En Gedi," ESI 5 (1986), 27-29.

214 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST Ohlig, C. "Ammerkungen zum Funktionsmodell des Castellum Aquae im antiken Pompeji" in De Haan and Jansen 1996, 1928. _____. De Aquis Pompeiorum. Das Castellum Aquae in Pompeji: Herkunft, Zuleitung und Verteilung des Wassers (Nijmegen, 2001). _____. "Vitruvs technisches Konzept zur Wasserversorgung einer Stadt und der Verteileranlage im antiken Pompeji. Ein Betrag zur Entwicklungsgeschichte eines Faktoids," in Ohlig et al. 2002, 201-212. Ohlig, C., Peleg, Y. and Tsuk, T. (eds) Cura Aquarum in Israel. (Siegburg: DW hG, 2002). Oleson, J.-P. "A Roman water-mill on the Crocodilion River near Caesarea," ZDPV 100 (1983), 137-152. _____. Greek and Roman Mechanical Water-Lifting Devices: The History of a Technology (Toronto: University of Toronto Press, 1984). _____. "The Humayma hydraulic survey: preliminary report of the 1986 season," ADAJ 30 (1986), 253-261. _____. "The Humayma Hydraulic Survey: preliminary report of the 1987 season," ADAJ 32 (1988), 157-170. _____. "Nabataean and Roman water use in Edom: the Humayma hydraulic survey, 1987," Classical Views 7 (1988), 117-129. _____. "The Humeima Hydraulic Survey: preliminary report of the 1989 season," ADAJ 34 (1990), 285-311. _____. "Aqueducts, cisterns and the strategy of water supply at Nabataean and Roman Auara," in Hodge 1991, 45-63. _____. "The origins and deisgn of Nabataean water supply systems," in Amr, K., Zayadine, F. and Zaghloul, M. (eds) Studies in the History and Archaeology of Jordan 5, (Amman: Dept of Antiquities, 1995), 707-719. _____. "Water-lifting," in Wikander 2000, 217-302. _____. "A miscellany on technology in late antiquity," JRA 22/2 (2009), 813-819. Oleson, J.-P., Amr, K., Foote, R., Logan, J., Reeves, M. B. and Schick, R. "Preliminary report of the al-Humayma excavation project, 1995, 1996, 1998," ADAJ 43 (1999), 411-450. Oren, E. D. "Jerusalem, the French Hill," IEJ 22 (1972), 172-3. Orssaud, D. "Notice sur la céramique," in Lauffray 1991, 260-275. Ovadieh, A. "Excavations in the area of the ancient synagogue at Gaza: preliminary report," IEJ 19 (1969), 193-198.

BIBLIOGRAPHY

215

Patrich, J. "The aqueducts of Hyrcania-Kastellion," in Amit et al. 2002, 336-352. Patrich, J. and Amit, D. "The aqueducts of Israel: an introduction," in Amit et al. 2002, 9-20. Payen, A. "Notice adressé à M. le Géneral Desvaux sur les travaux hydrauliques anciens, dont il existe encore de nombreux vestiges dans la partie du Hodna," RecConstantine 8 (1864), 1-14. Peleg, Y. "The characteristics of water distribution in Roman towns," in Jansen 2000, 241-246. Pena, J. T., Rossiter, J. J. and Wilson, A. Carthage Papers: The Early Colony's Economy, Water Supply, A Public Bath, and the Mobilization of State Olive Oil (Portsmouth, R.I: JRA, 1998). Perkins, A. L. The Art of Dura-Europos (Oxford: Clarendon Press, 1973). Perring, D. "Excavations in the souks of Beirut: an introduction to the work of the Anglo-Lebanese team and summary report," Berytus 43 (1997-1998), 9-34. _____. "Excavating the souks of Beirut," Current World Archaeology 16 (2006), 23-33. Perring, D., Seeden, H., Sheehan, P. and Williams, T. "Bey 006, 1994-1995: the Souks Area interim report of the AUB project," BAAL 1 (1996), 176-206. Petit, P. Libanius et la Vie Municipale à Antioche au ive. Siècle aprés J.-C (Paris: P Geuthner, 1955). Philip, G., Jabour, F., Beck, A., Bshesh, M., Grove, J., Kirk, A. and Millard, A. "Settlement and landscape development in the Homs region, Syria: research questions, preliminary results 1999-2000 and future potential," Levant 34 (2002), 1-23. Philip, G, Abdulkarim, M., Newson, P., Beck, A., Bridgland, D., Bshesh. M., Shaw, A., Westaway, R., and Wilkinson, K. "Settlement and landscape development in the Homs region, Syria: report on work undertaken during 2001-2003," Levant 37 (2005), 21-42. Phillips, M., Senior, T. F. and David, A. S. "Disgust: the forgotten emotion of psychiatry," British Journal of Psychiatry 172 (1998): 373-375. Piras, S. "Water luxury in Roman times. Simply a matter of excellent planning and engineering or of politics and philanthropy?," in Jansen 2000, 247-253.

216 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST Poccardi, G. "L'ile d'Antioche à la fin de l'antiquité: histoire et problème de topographie urbaine," in Lavan 2001, 155-172. Pococke, R. A. Description of the East, and Some Other Countries, ii: Observations on Palestine or the Holy Land, Syria, Mesopotamia, Cyprus and Candia (London, 1745). Poidebard, A. "Recherches sur le limes romain (campagne d'automne 1930)," Syria 12 (1931), 274-280. _____. La Trace de Rome Dans le Désert de Syrie: le Limes de Trajan à la Conquête Arabe, Recherches Aeriennes 1925-1932 (Paris: P Geuthner, 1934). Politis, K. D. "The stepped dam at Wadi el-Jilat," PEQ 125 (1993), 43-49. Pollard, N. Soldiers, Cities, and Civilians in Roman Syria (Ann Arbor: JRA, 2000). _____. "Roman material culture across imperial frontiers? Three case studies from Parthian Dura Europos," in Colvin, S. (ed.) The Greco-Roman East: Politics, Culture, Society (Cambridge: Cambridge University Press, 2004), 119-144. Poole, J. B. and Reed, R. "The 'tannery' of Ain Feshka," PEQ 93 (1961), 114-123. Porath, Y. "The evolution of the urban plan of Caesarea's southwest zone," in Raban, A. and Holum, K. G. (eds) Caesarea Maritima, a Retrospective after Two Millennia (Leiden: Brill, 1996), 105-120. _____. "Hydraulic plaster in aqueducts as a chronological indicator," in Amit et al. 2002(a), 25-36. _____. "The water-supply to Caesarea: a re-assessment," in Amit et al. 2002(b), 104-129. Porath, Y., Raban, A. and Patrich, J. "The Caesarea Excavation Project - March 1992-June 1994," ESI 17 (1998), 37-79. Prentice, W. Publications of the Princeton University Archaeological Expedition to Syria in 1904-5 and 1909. III, Greek and Latin Inscriptions B, North Syria (Leiden, 1922). Preusser, C. Nordmesopotamische Baudenkmäler Altchristlicher und Islamicher Zeit (Leipzig: Otto Zeller Verlag, 1911 (repr. 1984)). Pritchard, J. B. "The Roman port at Sarafand (Sarepta): preliminary report on the seasons of 1969 and 1970," BMB 24 (1971), 3956.

BIBLIOGRAPHY

217

Raban, A. "Dor-Yam: maritime and coastal installations at Dor in their geomorphological and stratigraphic context," in Stern and Berg 1995, 285-354. Rapuano, Y. "Jerusalem, Ya-ar Ramot," HA 109 (1999), 74*-75*. Reich, R. "Ritual baths," in Meyers, E. M., 1997 The Oxford Encyclopedia of Archaeology in the Near East. Vol. 4 (Oxford: OUP, 1997), 430-1. _____. "Two possible miqva’ot on the Temple Mount," IEJ 39 (1998), 63-65. Reich, R. and Billig, Y. "Jerusalem, Robinson's Arch, the Jerusalem Archaeological Park of the Second Temple Period," ESI 18 (1998), 88-90. Reich, R. and Shukron, E. "Jerusalem, Mamilla," ESI 14 (1994), 9296. Reifenberg, A. The Struggle Between the Desert and the Sown: The Rise and Fall of Agriculture in the Levant. With an Introduction by W.C. Lowdermilk (Jerusalem: Publishing Dept of the Jewish Agency, 1955). Rey-Coquais, J.-P. "Inscriptions grecques d'Apamée," AAAS 23 (1973), 39-84. Ritti, T., Grewe, K. and Kessener, P. "A relief of a water-powered sawmill on a sarcophagus at Hierapolis and its implications," JRA 20 (2007), 138-163. Robinson, D. M. and Hoyningen-Huene, G. Baalbek, Palmyra (New York: JJ Augustin, 1946). Robinson, O. F. Ancient Rome: City Planning and Administration (London: Routledge, 1992).s Ron, Z. "Agricultural terraces in the Judean mountains," IEJ 16 (1966), 33-49, 111-122. _____. "Development and management of irrigation systems in the mountain regions of the Holy Land," Transactions of the Institute of British Geographers 10 (1985), 149-169. _____. "Qanats and spring flow tunnels in the Holy Land," in Beaumont et al. 1989, 210-236. Rosen, B. and Finklestein, I. "Subsistence patterns, carrying capacity and settlement oscillations in the Negev Highlands," PEQ 124 (1992), 42-58. Rostovtzeff, M. I., Bellinger, A. R., Hopkins, C. and Welles, C. B., (eds) The Excavations at Dura Europos: Preliminary Report of the

218 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST Sixth Season of Work October 1932 - March 1933 (New Haven: Yale University Press, 1936). Rosumek, P. Technischer Fortschritt und Rationalisierung im Antiken Bergbau (Bonn: Habelt, 1982). Roussin, L. "East meets west...," in Johnson, P., Ling, R. and Smith, D. (eds) Fifth International Colloquium on Ancient Mosaics held at Bath (Ann Arbor: JRA, 1994), 30-42. Rubin, R. "Water conservation methods in Israel's Negev Desert in Late Antiquity," Journal of Historical Geography 14/3 (1988), 229244. Saarisolo, A. The Boundary Between Issachar and Naphtali: An Archaeological and Literary Study of Israel's Settlement in Canaan (Helsinki: Suomalaisen Tiedeakatemian Toimitksia, 1927). Sadler, S. "Le terroir agricole de Diyateh. L'irrigation comme condition d'existence de ce terroir," in Geyer 1990, 421-452. Safadi, C. "La foggara, système hydraulique antique, serait-elle toujours concevable dans la mise en valeur des eaux souterraines en Syrie?," in Geyer 1990, 285-293. Saghieh-Beydoun, M., Allam, M., Ala'eddine, A. and Abulhosn, S. "Bey 004: the monumental street 'cardo maximus' and the replanning of Roman Berytus," BAAL 3 (1998-1999), 95-126. Sagiv, N., Zissu, B. and Amit, D. "The northern system of Eleutheropolis (Beth Govrin)," in Amit et al. 2002, 177-186. Satlow, M. L. "Jewish constructions of nakedness in late Antiquity," Journal of Biblical Literature 116/3 (1997), 429-454. Schattner, I. The Lower Jordan Valley: A Study in the Fluviomorphology of an Arid Region (Jerusalem: Magnes Press, 1962). Schiøler, T. Roman and Islamic Water-lifting Wheels (Odense: Odense universitetsforlag, 1973). _____. "The watermills at the Crocodile River: a turbine mill dated to AD 345-380," PEQ 121 (1989), 133-143. _____. "How to saw marble," International Molinology 70 (2005), 345. Schmid, S. "L'eau à Pétra: l'exemple du Wadi Farasa Est," Syria 85 (2008), 19-32. Schmidt-Colinet, A. "Die sogennante Maison à Atrium," in Balty 1984, 141-149. Schnitter, R. N. "A short history of dam engineering," Water Power (1967), 142-8.

BIBLIOGRAPHY

219

_____. "Les barrages romains," Dossiers d’Archéologie 38 (1979), 205. Schomberg, A. "Ancient water technology: between Hellenistic innovation and Arabic tradition," Syria 85 (2008), 119-128. Schumacher, G. and Le Strange, G. Northern Ajlun, within the Decapolis (London: A P Watt, 1890). Sear, F. "Cisterns, drainage and lavatories in Pompeian houses, Casa del Granduca (VII.4.56)," PBSR 72 (2004), 125-166. Segal, A. From Function to Monument: Urban Landscapes of Roman Palestine, Syria and Provincia Arabia (Oxford: Oxbow, 1997). _____. "Horbat Susita," HA 111 (2000), 12*-13*. _____. "Horbat Hanut," HA 115 (2003),36*-37*. Seigne, J. "A sixth-century water-powered sawmill at Jarash," ADAJ 46 (2002), 205-213. _____. "Fontaines et adduction d'eau à Gerasa (Jérash, Jordanie)," Syria 85 (2008), 33-50. Seyrig, H. "Antiquités syriennes 76: caractères de l'histoire d'Emèse," Syria 36 (1959), 184-192. Shahada, K. "The Salamia-Afamia (or Lover's) Canal in Syria (near Hama) circa 2nd century AD," AAAS 7 (1957), 155-166 (Arabic section). Shanks, H. "Report from Jerusalem," Biblical Archaeology Review 3/4 (1977), 16-21. Shaw, B, "Water and society in the ancient Maghrib: technology, property and development," Antiquités Africaines 20 (1984), 121-73. _____. "The noblest monuments and the smallest things: wells, walls and aqueducts in the making of Roman Africa," in Hodge 1991, 63-91. Siegelmann, A. "Appendix: The tunnels of Taninim and Snunit," in Amit et al. 2002, 130-140. Simms, D. L. "Water-driven saws, Ausonius and the authenticity of the Mosella," Technology and Culture 24 (1983), 635-643. _____. "Water-driven saws in late Antiquity," Technology and Culture 26 (1985), 275-276. Sinclair, T. A. Eastern Turkey: An Architectural and Archaeological Survey II (London: Pindar Press, 1989). Small, D. B. "Late Hellenistic baths in Palestine," BASOR 266 (1987), 59-74. Smith, N. A. F. A History of Dams (London: P Davies, 1971).

220 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST _____. "Attitudes to Roman engineering and the question of the inverted siphon," History of Technology 1 (1976), 45-71. Smith, R. "The 1967 excavations at Pella of the Decapolis," ADAJ 14 (1969), 5-10. Sourdel, D. Les Cultes du Hauran à l'Époque Romain (Paris: Impr. Nationale, 1952). Spijkerman, A. The Coins of Decapolis and Provincia Arabia (Jerusalem: Franciscan Printing Press, 1978). Sprague de Camp, L. The Ancient Engineers (London: Souvenir Press, 1963). Stager, L. E. "Eroticism and infanticide at Ashkelon," Biblical Archaeology Review 17/4 (1991). Stenton, E. C. and Coulton, J. J. "Oinoanda: the water supply and aqueduct," Anatolian Studies 36 (1986), 15-60. Stern, E. "Tel Dor - 1984," ESI 4 (1985), 21-24. _____. Dor: Ruler of the Seas: Twelve Years of Excavations at the IsraelitePhoenician Harbor Town on the Carmel Coast (Jerusalem: IES, 1994). Stern, E. and Berg, J. Excavations at Dor: Final Report (Jerusalem: IES, 1995). Stern, E., Berg, J. E. and Sharon, I. "Tel Dor 1988-9 preliminary report," IEJ 41 (1991), 46-61. Stern, E. and Gilboa, A. "Tel Dor - 1989," ESI 9/2 (1989-1990), 114-117. Stern, E. and Sharon, I. "Tel Dor 1986 preliminary report," IEJ 37 (1987), 201-211. Stieglitz, R. R. "A late Byzantine reservoir and piscina at Tel Tanninim," IEJ 48/1-2 (1998), 54-65. Stillwell, R. Antioch-on-the-Orontes III (Princeton: Princeton University Press, 1941). Stocks, H. "Studien zu Lukians 'De Dea Syria'," Berytus 4 (1937), 140. Strang, V. The Meaning of Water (Oxford and New York: Berg, 2004). ____. "Fluidscapes: water, identity and the senses," Worldviews 10/2 (2006), 147-154. Syon, D. "Nahal Zippori," ESI 14 (1994), 46-49. Taylor, R. "Torrent or trickle? The Aqua Alisetina, the Naumachia Augusti, and the Transtiberim," AJA 101 (1997), 465-92.

BIBLIOGRAPHY

221

Teixidor, J. The Pagan God: Popular Religion in the Greco-Roman Near East (Princeton: Princeton University Press, 1977). Thorpe, R. "Bey 007: the souks area: preliminary report of the AUB/ACRE project," BAAL 3 (1998-1999a), 31-55. _____. "Bey 045: preliminary report on the excavations," BAAL 3 (1998-1999b), 57-83. Trakadas, A. "The archaeological evidence for fish processing in the western Mediterranean," in Bekker-Nielsen, T. (ed.) Ancient Fishing and Fish Processing in the Black Sea Region (Aarhus: Aarhus University Press, 2005), 47-82. Trombley, F. R. and Watt, J. W. The Chronicle of Pseudo-Joshua the Stylite (Liverpool: Liverpool University Press, 2000). Tsaferis, V. "A Roman bath at Rama," Atiquot 14 (1980), 66-75. Tsaferis, V. and Avner, R. "Banias - 1989," ESI 9/1 (1989-1990), 3f. Tsaferis, V. and Israeli, S. "Banias - 1993," ESI 15 (1995), 5-7. Tsafrir, Y. and Foerster, G. "Beat Shean excavation project 1988/1989," ESI 9/2 (1989-1990), 120-128. Tsuk, T. "The aqueducts to Sepphoris," in Amit et al. 2002(a), 278294. _____. "An aqueduct to Legio," in Amit et al. 2002(b), 409-411. _____. "The water supply system of Shivta in the Byzantine period," in Ohlig et al. 2002(c), 65-80. Uscatescu, A. and Martin-Bueno, M. "The macellum of Gerasa (Jerash, Jordan): from a market place to an industrial area," BASOR 307 (1997 ), 67-88. Van Ess, M. and Weber, T. Baalbek (Mainz am Rhein: Philipp von Zabern, 1999). Vetrali, L. "Le inscrizioni dell'acquedotto romano presso Betlemme," Liber Annus 17 (1967), 149-161. Vetters, H. "Ephesos - Vorlaufer Grabungsbericht 1983," Anz. Wien. 121 (1984), 225. Volney, C.-F. Travels through Syria and Egypt, in the Years 1783, 1784, 1785 (London: G G J and J Robinson, 1787). Waheeb, M. "Archaeological excavation at Ras an-Naqab - Aqaba road alignment: preliminary report (1995)," ADAJ 40 (1996), 339-348. Waheeb, M. and Zubi, Z. "Recent excavations at the Amman Nymphaeum: preliminary report," ADAJ 39 (1995), 229-240.

222 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST Ward-Perkins, B. From Classical Antiquity to the Middle Ages: urban public building in northern and central Italy AD 300-500 (Oxford: OUP, 1984). Weber, T. "Gadara of the Decapolis: a summary of the 1988 season at Umm Qeis," ADAJ 32 (1988), 349-352. _____. "Thermalquellen und Heilgötter des Ostjordanlandes in römischer und byzantinischer zeit," Damaszener Mitteilungen 11 (1999), 433-443. Weiss, Z. and Netzer, E. "Zippori - 1992/1993," ESI 14 (1994), 43. Welles, C. B. "Dura Papyrus 101: purchase of vineyard on the Chabur by a Roman veteran in AD 227," Archives d'Histoire du Droit Oriental 1 (1937), 1-28. _____. "Inscriptions," in Kraeling 1938, 355-494. West, M. L. The East Face of Helicon: West Asiatic Elements in Greek Poetry and Myth (Oxford: Clarendon Press, 1997). Weulersse, J. L'Oronte: Étude d'une Fleuve (Tours: Arrault, 1940(a)). _____. Le Pays des Alaouites (Tours: Arrault, 1940(b)). White, L. Medieval Technology and Social Change (Oxford: Clarendon Press, 1962). Wiemken, R. C. and Holum, K. G. "The Joint Expedition to Caesarea Maritima: eighth season 1979," BASOR 244 (1981), 27-52. Wigley, T. M. L. and Framer, G. "Climate of the eastern Mediterranean and Near East," in Bintliff and Van Zeist 1982, 3-37. Wikander, O. "Water-mills in ancient Rome," Opuscula Romana 12 (1979), 13-36. _____. "The use of water-power in Classical Antiquity," Opuscula Romana 13 (1981), 91-104. _____. Exploitation of Water Power or Technological Stagnation? A Reappraisal of the Productive Forces in the Roman Empire (Lund: CWK Gleerup, 1984). _____. "Archaeological evidence for early water-mills - an interim report," History of Technology 10 (1985), 151-179. _____. "Ausonius' saw-mills - once more," Opuscula Romana 17 (1989), 185-190. _____. "Water-power and technological progress in Classical Antiquity," Opuscula Romana 17 (1990), 185-190. _____. Handbook of Ancient Water Technology (Leiden: Brill, 2000(a)) _____. "The water mill," in Wikander 2000(b), 371-400.

BIBLIOGRAPHY

223

_____. "Industrial applications of water power," in Wikander 2000(c), 332-341. _____. "Canals," in Wikander 2000(d), 321-331. _____. "The Iron Age and the Archaic and Classical periods," in Wikander 2000(e), 617-630. Wilde, W. R. Narrative of a Voyage to Madeira, Tenerife and Along the Shore of the Mediterranean II (Dublin: Dublin and Co., 1840). Wilkinson, T. J. The Archaeological Landscape of the Balikh Valley, Syria, 1994-95 Annual Report (1995) http://oi.uchicago.edu/OI/AR/94-95/94-95_Jazira.html

_____. "Water and human settlement in the Balikh Valley, Syria: investigations from 1992-1995," Journal of Field Archaeology 25/1 (1998(a)): 63-87. _____. "Settlement and irrigation in the Balikh Valley, Syria from the 3rd to 1st millennium BC," in Lebeau, M. (ed.) About Subartu: Studies devoted to Upper Mesopotamia (Subartu IV) (Turnhout: BREPOLS, 1998(b)), 151-70. _____. "Archaeological survey of the Tell Beydar region, Syria 1997: a preliminary report," in Van Leberghe, K and Voet, G. (eds) Tell Beydar Environmental and Technical Studies (Subartu VI) (Turnhout: BREPOLS, 2000(a)), 1-37. _____. "Settlement and land use in the zone of uncertainty in Upper Mesopotamia," in Jas, R. (ed.) Rainfall and Agriculture in Northern Mesopotamia (Amsterdam: nederlands Institut voor het Nabije Osten, 2000b), 3-35. Wilkinson, T. J. and Barbanes, E. "Settlement patterns in the Syrian Jazira during the Iron Age," in Bunnens, G. (ed.) Essays on Syria in the Iron Age (Leuven: Peeters, 2000), 397-422. Wilkinson, T.J. and Tucker, D. J. Settlement and Development in the North Jazira, Iraq: A Study of the Archaeological Landscape (Warminster, 1995). Will, E. "Le developpement urbain de Palmyre: témoignage épigraphiques anciens et nouveaux," Syria 60 (1983), 69-81. _____. "Antioche sur l'Oronte, métropole d'Asie," Syria 74 (1997), 99-113. Wilson, A. I. "Running water and social status in North Africa," in Horton, M. and Widermann, T. (eds) North Africa from Antiquity to Islam (Bristol: Centre for Mediterranean Studies, 1995), 52-6.

224 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST _____. "Water-power in North Africa and the development of the horizontal water-wheel," JRA 8 (1995), 499-510. _____. Water management and usage in Roman North Africa: a social and technological study, (Unpublished DPhil. Thesis: Oxford University, 1997). _____. "Deliveries extra urbem: aqueducts and the countryside," JRA 12 (1999), 312-332. _____. "Land drainage," in Wikander 2000(a), 303-317. _____. "Drainage and sanitation," in Wikander 2000(b), 151-179. _____. "Incurring the wrath of Mars: sanitation and hygiene in Roman North Africa," in Jansen 2000(c), 307-312. _____. "Industrial uses of water," in Wikander 2000(d), 127-150. _____. "The aqueducts of Italy and Gaul," JRA 13 (2000(e)), 597604. _____. "Urban water storage, distribution and usage in Roman North Africa," in Koloski-Ostrow 2001(a), 83-96. _____. "Water-mills at Amida: Ammianus Marcellinus 18.8.11," Classical Quarterly 51/1 (2001(b)), 231-236. _____. "Detritus, disease and death in the city," JRA 15 (2002(a)), 479-484. _____. "Machines, power and the ancient economy," JRS 92 (2002(b)), 1-32. _____. "Urban production in the Roman world: the view from North Africa," PBSR 70 (2002(c)), 231-73. _____. "Classical water technology in the early Islamic world," in Bruun and Saastamoinen 2003(a), 115-141. _____. "Opus reticulatum panels in the Severan Basilica at Lepcis Magna," Quaderni di Archeologia della Libya 18 (2003(b)), 369379. _____. "Water for the Pompeians," JRA 19 (2006a), 501-508. _____. "Fishy business: Roman exploitation of marine resources," JRA 19 (2006b), 525-537. Wilson, R. J. A. "Tot aquarum tam multis necessariis molibus ... Recent studies on aqueducts and water supply," JRA 9 (1996), 5-29. Wilson, N. G. Scholars of Byzantium (Baltimore, Maryland: John Hopkins University Press, 1983). Winogradov, Z. S. "The aqueduct of Tiberias," in Amit et al. 2002, 295-304. Wiplinger, G. "Stand der Erforschung der Wasserversorgung in Ephesos/Türkei," Schriftenreihe der Frontinus-Gesellschaft:

BIBLIOGRAPHY

225

Frontinus-Tagungen 2004/2006 in Wien und Berlin 27 (20042006), 15-48. Wood, B. G. "To dip or to sprinkle? The Qumran cisterns in perspective," BASOR 256 (1984), 45-60. Yadin, Y. "The excavation of Masada – 1963/64, preliminary report," IEJ 15 (1965), 1-120. Yegül, F. K. Baths and Bathing in Classical Antiquity (New York, London: MIT Perss, 1992). Yeivin, Z. and Finkielsztejn, G. "Horbat Castra - 1993-1997," HA 109 (1999), 23*-27*. Yogev, O. "Tel Yizre'el - 1987/1988," ESI 7-8 (1988-1989), 189195. Zayadine, F. and Farajat, S. "Excavation and clearance at Petra and Beida," ADAJ 35 (1991), 275-311. Zissu, B., Tepper, Y. and Amit, D. "Miqwa'ot at Kefar Όthnai near Legio," IEJ 56 (2006), 57-66.

INDEX Balikh, 11, 12, 103 Barada, 12, 13 bathhouse, 22, 51, 69, 75, 83, 88, 120, 121, 128, 129, 146, 150, 151, 153, 155, 181 Beirut, 7, 49, 57, 58, 59, 80, 81, 82, 84, 85, 86, 87, 96, 108, 119, 185 Bosra, 71, 74, 94, 107, 115, 119, 120, 121, 176 branch lines, 85, 86, 87, 88, 89, 90, 91, 92, 98, 186 bridge, 46, 48, 49 Bronze Age, 102 Caesarea, 49, 52, 54, 57, 60, 66, 68, 70, 71, 80, 83, 86, 87, 88, 90, 108, 110, 111, 114, 119, 120, 127, 134, 135, 142, 145, 149, 151, 155, 183, 185 Caligula, 66 Capitolias, 55, 56, 57, 104, 107, 110 Caracalla, 114, 150 castellum divisorium, 51, 53, 54, 122 Christianity, 15, 97, 99, 111, 151, 168 Church, 54, 55, 73, 74, 89, 93, 98, 111, 113, 154, 184, 187

Aleppo, 10, 34, 107 Amman, 13, 115 Anastasius, 26 Andarin, 35, 83, 120, 151, 175 Antioch, 8, 12, 58, 59, 64, 66, 70, 72, 73, 77, 79, 80, 82, 85, 91, 92, 93, 94, 107, 112, 119, 120, 121, 123, 125, 126, 128, 133, 142, 151, 153, 155, 175, 181, 185 Apamea, 20, 50, 52, 54, 58, 59, 69, 94, 96, 107, 110, 119, 120, 134, 141, 142, 187 aqueduct, 6, 46, 48, 49, 50, 51, 52, 53, 54, 57, 65, 66, 67, 69, 70, 71, 72, 74, 79, 83, 84, 86, 87, 88, 90, 91, 92, 96, 98, 102, 103, 105, 107, 108, 109, 111, 113, 115, 121, 122, 125, 128, 133, 136, 138, 140, 153, 172, 181, 182, 186 Arabia, 73, 112, 168 arcades, 48, 49, 69, 74, 182, 183 Arcadius, 81 army, 8, 9, 65, 66, 68, 69, 74, 94, 95, 96, 128, 179, 184, 187 Atargatis, 35, 159, 161, 168, 169, 171, 176, 178, 187 Baal, 159, 160 Baalbek, 12, 160

227

228 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST cistern, 43, 46, 55, 58, 78, 81, 83, 105, 108, 120, 121, 124, 138 cleanliness, 147, 149, 155, 156, 186 climate, 7, 8, 12, 14, 30, 85, 93, 103, 112, 118, 129, 179, 185 Constantine, 69 constant-offtake principle, 14, 51, 56, 101, 102, 103, 105, 109, 110, 113, 186 dam, 24, 25, 27, 28, 29, 43, 68, 102, 103, 180, 182 Damascus, 7, 10, 13, 32, 34, 107, 177 Dead Cities, 93, 121 Decapolis, 119 Diocletian, 2, 81 domestic, 50, 58, 122, 123, 124, 125, 137, 173, 181, 182, 184 drain, 133, 135, 140 Dura Europos, 58, 59, 108, 120, 121, 123, 124, 133, 134, 160, 169, 172, 184 dyeing, 77, 80, 81, 88 earthquake, 66, 70, 78, 112 Edessa, 94, 172, 174, 187 Elagabalus, 21, 114, 150 Emperor, 8, 65, 67, 69, 70, 71, 74, 89, 93, 112, 114, 115, 153 Euphrates, 12, 31, 82, 94, 108, 135, 175, 183 fish, 77, 82, 83, 84, 85, 125, 168, 172, 173, 175, 178, 187 fishponds, 82, 83, 173 floodwater farming, 31, 38, 41, 46, 180 fountain, 47, 72, 73, 81, 114, 115, 116, 123, 125 fulling, 64, 77, 79, 80, 81, 85

gardens, 42, 46, 89, 91, 93, 122 Gordianus, 150 Hadad, 159, 160 Hadrian, 66, 72, 115, 172 Hammat Gader, 150, 152 Hasmonean, 47, 49, 92, 93, 137, 138 Hauran, 10, 73, 86, 89, 160, 174 Herod, 47, 49, 65, 69, 72, 73, 74, 93, 115, 120, 145, 146, 148, 149, 152, 182, 183, 185 Herodian, 47, 48, 49, 68, 92, 93, 138 Hierapolis, 35, 78, 160, 169, 172, 173, 174 Homs, 7, 10, 21, 28, 29, 32, 37, 44, 91, 107, 109 House, 58, 122, 125, 126 hygiene, 131, 134, 136, 149, 150, 156, 186 identity, 1, 59, 98, 131, 146, 148, 156, 178, 179, 185, 188 irrigation, 4, 6, 8, 9, 13, 24, 30, 31, 33, 37, 42, 43, 44, 46, 49, 77, 83, 86, 87, 90, 91, 92, 95, 96, 161, 180, 182, 183, 184, 187 Jerash, 65, 72, 73, 77, 78, 80, 81, 85, 96, 98, 108, 110, 115, 119, 121, 135, 142, 150, 173, 175, 182, 184, 187 Jericho, 87, 91, 92, 120, 137, 138, 139 Jerusalem, 8, 9, 39, 49, 50, 55, 58, 67, 68, 108, 110, 112, 115, 128, 135, 137, 139, 151, 153 Jewish, 14, 67, 131, 137, 138, 145, 146, 148, 183, 184, 186 Josephus, 67, 73, 107, 115

INDEX Judaea, 11, 55, 142, 145, 148, 149, 168, 183, 186 Judaism, 15, 139, 146, 147, 184, 186 Julius Caesar, 66 Jupiter Dolichenus, 160 Justinian, 24, 25, 26, 66, 71, 112 kalybe, 115, 118, 129 Kanata, 69, 119 Khabour, 11, 12, 32, 174 latrines, 14, 51, 55, 59, 69, 74, 110, 123, 125, 127, 131, 133, 134, 140, 142, 145, 146, 147, 148, 157, 159, 181, 182, 184, 186 Libanius, 9, 64, 66, 70, 73, 89, 91, 94, 123, 155, 176 Litani, 12 Lucian, 36, 172, 174 Maioumas, 175 Maximilianus, 69 milling, 24, 62, 77 mills, 61, 64, 77, 78, 79, 85, 87, 90, 182 miqveh, 137, 138, 140 Monastery, 43, 147 Nabataean, 26, 27, 28, 30, 56, 92, 169 Negev, 10, 37, 39, 58, 67, 87, 108, 109, 121 noria, 18, 20, 22, 182 nymphaeum, 47, 51, 72, 97, 114, 115, 118, 119, 125, 184 Orontes, 12, 21, 22, 29, 32, 108, 169 pagan, 97, 159, 160, 168, 177, 187 palace, 47, 59, 69, 133, 148, 174

229 Palmyra, 34, 36, 37, 72, 102, 108, 119, 160, 169, 172, 173, 177, 184 Pella, 114, 119 Petra, 30, 52, 53, 108, 119, 120, 136 Philippopolis, 120 piped water, 58, 59, 82, 83, 88, 123 pipeline, 46, 49, 50, 57, 83, 88, 90, 97, 113, 182 ceramic, 48, 50, 52, 58, 70, 180 lead, 50, 57, 59, 181 stone, 49, 180 pollution, 131, 136, 149, 155, 179 Procopius, 24, 25, 109, 118 purity, 136, 147, 149, 154, 157, 179, 186 qanat, 10, 33, 34, 35, 46, 83, 172, 180, 183 religion, 159, 160, 174, 177, 184, 188 Resafe, 55, 57, 103, 104, 110, 111 reservoir, 29, 32, 35, 37, 42, 43, 55, 57, 72, 73, 83, 87, 88, 102, 103, 104, 108, 109, 112, 138, 176, 187 reservoir-cistern, 55, 56, 57, 104, 110 river, 10, 11, 12, 13, 21, 32, 34, 46, 80, 84, 103, 109, 136, 168, 169, 180 River Jordan, 10, 13, 31, 136, 152, 168, 177 saqiya, 18, 21, 22, 43, 120, 180, 182 sawmill, 78, 96, 182, 184

230 ARCHAEOLOGIES OF WATER IN THE ROMAN NEAR EAST Scythopolis, 9, 52, 53, 70, 73, 74, 80, 82, 85, 97, 108, 119, 120, 134, 142, 145, 149, 150, 154, 183 Seleucid, 2, 66 Serjilla, 120, 121, 155 sewer, 73, 78 Sidon, 89 siphon, 49, 50, 67, 68, 182 Sozomen, 176, 177 springs, 9, 11, 13, 72, 111, 135, 152, 159, 161, 167, 176, 177, 178 Strata Diocletiana, 95 Syria, 2, 8, 10, 11, 31, 32, 34, 35, 37, 69, 70, 93, 103, 141, 142, 146, 159, 168, 169, 173, 183, 187 tanning, 77, 80, 82 Teleilat al-anab, 44 temple, 67, 97, 151, 160, 169, 172, 173, 174, 177 terrace farming, 38 Theodosian Code, 69, 88 Theodosius, 81, 153 Tiberias, 13, 49, 86, 87, 88, 90, 108, 110, 150, 181, 186

Tigris, 12 Titus, 67, 115 Trajan, 66, 69, 72, 73, 114, 150 tunnel, 35, 44, 46, 56, 68, 83, 87, 110 Tyre, 13, 80, 81, 89, 107, 174, 175 Umm al-Jimal, 107 Umm Qes, 119, 136, 150, 151 vivaria, 83 votive deposition, 15, 159, 161, 168, 172, 176 wadi, 27, 39, 41, 43, 204 wadi farming, 38 Wadi Faynan, 37, 38, 39, 41, 42, 64, 77 water-lifting, 4, 14, 18, 21, 22, 23, 87, 180, 182, 187 watermill, 62 water-pouring, 172, 174 Yarhibol, 160, 161, 177 Yarmouk, 2, 11, 12, 13, 177, 178 Zenobia, 120 Zerqa, 11, 12, 13 Zeugma, 59, 80, 82, 123, 125, 127, 134, 135, 142 Zosimus, 177