A History of Water: Series III, Volume 1: Water and Urbanization 9780755694310, 9781780764474

These three volumes present an original exploration of all aspects of water – social, cultural, political, religious, hi

223 23 88MB

English Pages [713] Year 2014

Report DMCA / Copyright


Polecaj historie

A History of Water: Series III, Volume 1: Water and Urbanization
 9780755694310, 9781780764474

Citation preview

We would like to thank the Norwegian Research Council, University of Bergen, Norway and the Nordic Africa Institute, Uppsala, Sweden.

Published in 2014 by I.B.Tauris & Co. Ltd 6 Salem Road, London W2 4BU 175 Fifth Avenue, New York NY 10010 www.ibtauris.com Distributed in the United States and Canada Exclusively by Palgrave Macmillan 175 Fifth Avenue, New York NY 10010 Copyright Editorial Selection and Introduction © 2014 Terje Tvedt and Terje Oestigaard Copyright Individual Chapters © 2014 Eduardo Araral, Sabine Barles, Martin Byrkjeland, Rong Chen, Leong Ching, James Crow, Francesca de Châtel, Niall Finneran, Jeffrey Fleisher, J. Aaron Frith, Sylvia Gierlinger, Daniel M. Gnatz, André Guillerme, Morten Hammerborg, Fransje L. Hooimeijer, Savitri Jalais, Michael Jansen, Petri S. Juuti, Tapio S. Katko, Marianne Kjellén, Irene J. Klaver, Demetris Koutsoyiannis, Fridolin Krausmann, Alphonce Kyessi, Peter Maw, Betsy McCully, Martin V. Melosi, Han Meyer, Michael Neundlinger, Lorenzo Nigro, Anna Patrikiou, Olof Pedersén, Gudrun Pollack, Michael J. Rawson, Katherine Rinne, Patricia Romero-Lankao, Federica Spagnoli, Federica Sulas, Xiaochang C. Wang, Leah J. Wilds, Kenneth R. Wright, Xun Wu, Stephanie Wynne-Jones The right of Terje Tvedt and Terje Oestigaard to be identified as the editors of this work has been asserted by them in accordance with the Copyright, Designs and Patents Act 1988. All rights reserved. Except for brief quotations in a review, this book, or any part thereof, may not be reproduced, stored in or introduced into a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior written permission of the publisher. Every attempt has been made to gain permission for the use of the images in this book. Any omissions will be rectified in future editions. The websites mentioned in the book were accurate at the time of publication. ISBN: 978 1 78076 447 4 A full CIP record for this book is available from the British Library A full CIP record is available from the Library of Congress Library of Congress Catalog Card Number: available Designed and typeset by 4word Ltd, Bristol Printed and bound in Great Britain by T. J. International, Padstow, Cornwall

i-viii WaterUrban Prelims.indd 4

16/05/2014 15:23

Urban Water Systems—A Conceptual Framework Terje Tvedt and Terje Oestigaard There can be no doubt that the dominant tradition in urban studies has given scant attention to the universal and structural importance of water in urbanization processes. Peter Hall, in his acclaimed Cities in Civilization (1998), does discuss the role of water in the development of Rome, Paris and London, but this volume on cities in civilization has a register with no general entries on either sewage, water supply system, rivers, canals or aqueducts. In the same author’s book on the future of cities from 2002, the water issue is of marginal interest (Hall, 2002). A summary of the content of all the volumes of the journal Urban Studies between 2006 and 2012 shows that out of 14,363 pages, only 86 pages were devoted to the water issue. These pages were not concerned with the physical or manmade environment impacting city development and affected by city development, or with its role in shaping patterns of social activities, power, or control. The few articles dealt with water as a case in studies of political– economic issues, mainly and not surprisingly the water-pricing issue. There were altogether four articles that dealt with such issues. None analyzed the interaction between water systems and cities, and how these impacted the social and economic life of the people in the cities. The book with the allincluding title Understanding the City (Eade and Mele, 2002) does not give the water issue any attention whatsoever. A textbook in sociology in an influential series on sociology in the twenty-first century, The World of Cities, are only dealing with social aspects of urbanization, although it claims to be broad and comprehensive in its outlook. The book promises to “take a journey across time and space, over the urban landscape and to be historical and comparative in perspective” (Orum and Xiangming Chen, 2003: xi). It has, however, no discussion on the relationship between cities and water whatsoever, and carries not one reference to either water, rivers, sewage or waterways and canals (Orum and Xiangming Chen, 2003). Theoretical books on urban politics are neither concerned with the urban/ water issue and how it frames and shapes both power relations in cities and makes footprints in the water landscape (see, for example, Parker, 2003; and Davies and Imbroscio, 2009). This volume and article takes as a starting point that modern urban studies have persistently tended to

001-022 WaterUrban Introduction.indd 1

12/05/2014 11:04


A History of Water

neglect the water issue and the interlinkages between city development and water. The new “cultural geography” or human geography, concerned with unmasking the meaning of cities, landscapes or buildings, unpacking it as a text, have in general not been interested in unpacking the meaning of urban water landscapes. Since one of the most important urban infrastructures, the water supply and sewage system, is a truly hidden, not visible structure in a strict material sense because they are underground and not a part of the “built landscape”, this aspect of the confluence between city and water has naturally been difficult to unpack as text. We will here show how a focus on water/urban relationships can further our knowledge of city developments, and how urban studies can deepen our understanding of the role of water in societies. The basic premises for this proposal are two facts of huge importance for understanding the history and development of the city; the universal needs of water and the actual, physical waterscape at a given place. All urban dwellers1—from the first few people who settled around a natural spring in a desert in Jericho and built a wall around it almost 10,000 years ago, to the Incas living in the royal city Machu Picchu on a mountain top in the Andes, to stock traders relaxing in spacious apartments in a skyscraper on Manhattan, and to the party officials assembled for one of their meetings in a grand conference hall in Beijing—share the need for one resource: water. The theoretical and empirical importance of this state of affairs can hardly be overestimated: these people all need water to survive, and as long as they live in cities, it has to be provided in one way or another. Water is the only universal urban resource that in this sense is a must and that can be controlled in this strict understanding of the word.2 For theoretical and empirical reasons, it is also very important for urban studies to acknowledge the natural fact that the hydraulic systems that envelop and underpin the city as climate and weather always vary, from place to place, and also from time to time at the same place. The character of the actual water system helps to define a sense of place in a fundamental way—for example, whether it is its relative humidity, the average number of rainy days, whether it receives snow and for how many months a year, or if it is situated in a desert. The urban dwellers’ interactions with and their patterns of activities in relation to their water will also reflect the local hydrological cycle’s particular characteristics, how their water has been modified locally in the past, and how they and their predecessors have conceived of their water and how it should be managed. All urban places are tied up in this continuous web of relationships with water’s simultaneous universalism and particularism. The geography and history of all cities in the world are therefore written in water, and in the most varied ways and manners. There are two very different, though interrelated questions to be answered: first, how can a focus on water/urban relationships help our understanding of cities’ developments; and second, how can urban studies

001-022 WaterUrban Introduction.indd 2

12/05/2014 11:04

Urban Water Systems—A Conceptual Framework3

broaden our knowledge of the role of water in societies? The possibilities of approaching these questions will be explored from a water systems perspective. The book argues that urban studies will benefit from new theoretical and conceptual approaches and a focus on water, since all cities at any time and in all places have been forced to accomodate this resource in order to grow. But how can the complex urban/water relationships be uncovered and mapped? In which manner can this broad multidimensional relationship which has impacted and changed urban history and landscapes be interpreted? How can the process and history by which city developments change and impact water landscapes be analyzed? And how can the flows of water in urban places be used to study flows of capital, power, labor, and ideas in cities? The recognition of this range of phenomena requires some unifying approach, but an approach that can offer at the same time open and nonreductionist frameworks within which the various elements and their relationships can be analyzed. Analyses of urban/water relationships should study the water system in its multidimensionality as an “open and multifunctional water system” (Tvedt, 2010a, 2010b). They should take into consideration the three layers of all cities’ water systems: the physical, natural waterscape; the humanly modified water systems; and the ideas and managerial assumptions about water. The way “system” is used here is different from the way systems theoreticians use it. It is a descriptive term, denoting three different aspects or layers of water in connection with its social importance that are best understood in relation to each other. When the words “first”, “second”, and “third” layer of the water system are used, these should not be regarded as signifying a static hierarchical ordering. Instead, they denote separate, distinctive, and related, layers that also may differ in explanatory importance according to what aspect of urban development one primarily focuses on. Although specific cities in this volume are used to exemplify these layers, it is important to stress that all of these layers are present at all times, in all cities in the world—and any city could have been used to exemplify the three layers. The first layer of water systems: the natural waterscape The first layer is water’s physical form and behavior. This covers precipitation and evaporation patterns, river discharges and velocity measurements, and aquifers and their behavioral characteristics (i.e., the natural waterscape or hydraulic system with relevance in the area where the city in question is located). This layer should be seen as an exogenous, physical factor, with certain particular characteristics, although these are always in a state of flux. This physical aspect of the water system should not be

001-022 WaterUrban Introduction.indd 3

12/05/2014 11:04


A History of Water

regarded as a separate “watery” ecosystem in nature, but as constituting a central, distinguishable aspect of all ecosystems reflecting and bringing with it traces or the meaning of the enveloping landscape. To understand how water runs through nature and urban societies, we need natural science data such as rainfall variations, rivers’ sediment loads, evaporation patterns, hydrological data series, aquifer developments, etc. All of these are important, although they are of different importance to different urban places. This focus on the physical water system and its importance for city development does not suggest a one-to-one relationship between a certain waterscape and city development. Moreover, the task of analyzing the role of the physical water context should not be seen as a simple descriptive exercise, because although water will always play a very important role in cities’ location process, there is no simple causal relationship between the physical character of the water system and city locations. It is true that some places do exhibit such clear-cut causal connections, but even when there are such correlations, levels two and three of the water systems approach are equally important (see discussion below). Everyone who has visited Jericho, the oldest urban settlement in the world, and already between 8500 and 7500 bc encircled by a defensive stone wall, knows this. In the beginning of the third millennium bc, Jericho became a flourishing city, located around the ‘Ain es-Sultan. This spring, which in the Bible is known as Prophet Elisha’s Spring, provided 4,000–5,000 liters of fresh water each minute. Importantly, this happened without any human action needed, and the water was easily distributed by gravity and canals. The water is still coming up from the ground, as if it is an ongoing miracle—but there are specific hydraulic reasons for it. From the earliest habitation up to the Ottoman period, this spring was the focal point for urban development and it decided the location of the city.3 In other cases, cities are located along large rivers. Babylon, situated along the Euphrates on the Mesopotamian flood plain in today’s Iraq, is often associated with the Hanging Gardens, although this has proven difficult to establish archaeologically. At the time of Nebuchadnezzar II (604–562 bc), Babylon was the leading metropolis in the world, measuring about 4.5 km2 at the beginning of his reign. Canals from the river were built for extensive irrigation, and some of the most impressive water canals were built within the city itself. The Inner City and the Western City were both surrounded by city walls and moats. By the end of Nebuchadnezzar’s reign, the eastern parts of the city were also protected by city walls and mounts, covering an area of about 9 km2.4 Babylon could not have developed where it did had not the Euphrates crossed the flood plain, but human modifications of the water landscape was necessary to develop and sustain it. Rain-harvesting offers yet other possibilities. Mohenjo-Daro in presentday Pakistan was built on artificial wells and developed a sophisticated water system both for supply and sewage. It was a part of the Indus

001-022 WaterUrban Introduction.indd 4

12/05/2014 11:04

Urban Water Systems—A Conceptual Framework5

civilization and one of the largest cities in the world of the third millennium. In the city, it is estimated that there were more than 700 wells and perhaps as many as 2,000—the highest density in the world. Each of these wells had an average catchment area radius of only 17 m, making the density unparalleled in the history of water supply. The most spectacular and well-known water structure in Mohenjo-Daro is the “Great Bath”, a tank 12 m × 7 m, and 2.4 m deep. However, the use and (ritual) function of this bath has been more difficult to establish. Another noteworthy development in the city is that it seems that almost every household had a separate “bathroom”.5 But the most intriguing aspect of the city is perhaps the system for sewage removal. The foundation for this water system was the fact that the water table on the Indus Plain was very close to the ground, so accessing the waters by wells was comparatively very easy. Often a city will be dependent upon utilization of different water resources, and it is these different resources in combination that form the particular water landscape developed within a city. Aksum was the capital of the Ethiopian civilization with the rise of the kingdom of Aksum (50 bc– ad 800). The centralized state emerged in the second century ad, reaching its maximum expansion by the mid-millennium. Rainfall, surface water, and groundwater were the basis for the rural population. In the city itself, there were several springs and cisterns receiving runoff water from the surrounding hills. The largest and the most celebrated one was originally about 65 m in diameter and 5 m deep. This is seen as a rain-harvesting cistern, and together with the other wells it would have supplied a sufficient amount of water to the estimated several thousand inhabitants in the city’s heyday. The name Aksum itself indicates the importance of the local water resources; it may be derived from “water”—“ak-” may derive from the Cushitic root for “water” and the Semitic term for “chief ” is “šum”. Another hypothesis is that the name comes from the western Agaw word “akuesem”, meaning “water reservoir”. 6 Athens is a case that undermines the notion that water’s impact on city location has a necessary, predictable pattern. Classical Athens was not located where it was because of an abundance of water. According to mythology, there was a competition between Athena and Poseidon regarding who could give the best gift to the city. Athena had wisdom and knowledge of arts and crafts, and Poseidon, as the god of waters, offered the Athenians a well at the Acropolis. The Athenians voted for wisdom instead of abundance of water.7 In the highlands of Ethiopia, one can study an example of yet another locational relationship between city and water. Here, rainfall is usually heavy but also strongly erratic and seasonal. The location of the country’s capital is related to water, but not as a dire necessity. Addis Ababa has the curing and healing aspects and capacity of bathing as the point of origin. The queen of Ethiopia, Taytu Betul (c.1851–1918), had spent much time at thermal springs and requested her husband, King Menelik I, to build a

001-022 WaterUrban Introduction.indd 5

12/05/2014 11:04


A History of Water

house by a specific spring in the highlands; he complied. Soon the area developed into a royal settlement to which Taytu Betul gave the name Addis Ababa, meaning “New Flower”. The king himself, aware of the curative powers of the waters, and troubled by rheumatism, established a royal enclosure, a palace, and an audience hall, paving the way for further city expansion.8 One can still see the remnants today on a hilltop outside the center of the city. Cities have been established not only where there is limited water availability or only one source of water, but also where there is too much water. One city with an overabundance of water is Rotterdam. Any visitor taking a boat ride around Rotterdam’s river and canals will realize that this is a city where water is everywhere. The fight against too much water made this city possible, as the latter part of the name indicates. Rotterdam is a polder city (enclosed by dikes), created by man in a bid to control water. Changes in sea level and river discharges have forced the city to employ different policies throughout its history. The city has undergone several phases with regards to managing the changing water: natural water management (until 1000); defensive water management (1000–1500); anticipative water management (1500–1800); offensive water management (1800–90); manipulative water management (1890–1990); and adaptive manipulative water management (1990 until today).9 The physical layer of this water system approach includes not only the absence and presence of water at a given place, but also the very form and nature it takes throughout the seasons. It makes it possible to integrate in the analysis how non-cultural and non-social facts affect how water must be controlled and has to be distributed horizontally. Such non-social variables influence the technology that can be chosen, the type of equipment that must be used or can be used, the size and complexity of the water distribution system, and how it is operated, etc. Northern countries with freezing waters may be a case in point. In Finland, winter may last for up to 200 days in northern Lapland and 100 days in the southern areas. The temperature varies from −45°C to −50°C at the coldest in the north to 35°C in the summer in the south. The water infrastructure therefore needs to handle variations of at least 70°C, and because the soil freezes in the city of Tampere (in southern Finland) to a depth of about 2 m during the winter, all water and sewage pipes have to be placed at a depth of 2.5 m in order to function in extreme conditions.10 It is obvious that a city like Tampere (subject to extreme seasonal fluctuations) will necessarily be organized and physically constructed in a different way than, for example, cities like Dhaka (flood plain) or Mecca (desert) because of differences in their respective water systems. But these natural facts are still overlooked, although they continue to produce and re-produce different possibilities and conditions for organized urban life. Although almost all capitals and big cities are located on riverbanks (though with important exceptions), there are, as shown above, no

001-022 WaterUrban Introduction.indd 6

12/05/2014 11:04

Urban Water Systems—A Conceptual Framework7

laws or fixed pattern governing the relationship between physical water systems and city locations. A focus on water/urban relationships will, however, make possible a better understanding of how fundamental physical structures of water systems impact different cities’ development. While most citizens have increasing mobility and the “flow” of people between countries and cities, and between rural areas and urban spaces, is increasing, cities remain quite fixed geographically. Indeed, one of the clear features of history is that people are stuck with the location of their cities—but at the same time they have gradually altered the original waterscape, due to improvements in technology and organizational skills. Cities can and must change; they must be retrofitted and repurposed in relation to their water resources. Indeed, they always have been—by urban governance and citizenry; by the hydraulic engineer; and by the architect and the urban designer. Cities’ relations to their water systems will become more and more challenging, because the uncertainty about future waterscapes due to global warming scenarios becomes gradually more important, and cities and their populations’ expectations of and demand for water will continue to increase and diversify. Another reason why understanding the relation between the first layer and cities’ location and development should not be seen as simply a useful descriptive exercise is that the physical character of the water system is changing over time (although very slowly in some cases, reflecting the water source in question). Changing water systems will therefore impact urban location and development in various ways over time. Xi’an, which became the greatest ancient city in Chinese history with a population of up to 1 million inhabitants and an urban area covering 83.1 km2, was surrounded by eight rivers. The name itself, Chang’an in Chinese, perhaps hints at this particular water situation, as it means “long-lasting safety and prosperity”. Still, due to both the nature of the rivers and human attempts at modifying them, the city experienced water problems. Throughout the different Chinese dynasties, the city was moved to more optimal places.11 But everyone seeing the city’s surroundings today will realize that the name is still a fitting one—rivers are the arteries of the city. the Second layer of water systems: human modifications of the waterscape A water system that is of relevance to a city’s development will always— with specific but complex consequences—be modified in one way or another by urban action. The basic reason for this is that the physical water system underpinning a certain city’s location in the first place will be “appropriated” for different demands and reasons at different points in its history. This is because the need for water will always be there and will also change over time, not only because of increases in urban population, but

001-022 WaterUrban Introduction.indd 7

12/05/2014 11:04


A History of Water

also because changing economic and social activities will put greater stress on the water resources. Other actors in the watershed and their efforts at using and controlling the waters must also be analyzed in order to understand the character of a particular city’s water system. The human modifications will also in themselves have varying scales and histories. Most cities today are therefore enveloped by both an engineered waterscape and a waterscape that is still mirroring, to different degrees, the local character of how the hydrological cycle manifests itself in the landscape. Specific types of modified waterscapes have been the very symbol of urbanized life, distinguishing it from the natural haphazard dominating rural life (i.e., the dependency of erratic rainfall patterns for rainfed agriculture). Fountains, in general placed at the very heart of cities, have had many functions, but one of them has been to symbolize man’s control over nature—cultures’ appropriation of the forces of nature. Here the unruly, treacherous water element is completely controlled, to serve human needs for aesthetic beauty. Classical Rome was known as the city of fountains and baths, because what made the imperial capital possible was not the Pantheon or Colosseum: it was the human modifications of the waterscape. The initial localization of Rome was both mythical and practical. According to mythology, Rome was founded in c.753 bc by the Tiber River, “the river closest to god”. The foundation myth starts with a flood, when Romulus and Remus washed ashore at the foot of the Palatine Hill. Springs, streams, and marshes were central in the early city building, and streams were channeled in order to dry saturated land. But as the city expanded, the numerous fresh water sources were insufficient to meet the city’s water demands. Rome’s first aqueduct, the Aqua Appia of 312 bc, ran mostly underground, and between 312 bc and ad 226, 11 aqueducts were constructed. Rome’s history is one of many cases where the control of water was a main strategy for attaining power and demonstrating power.12 A telling expression of this can still be seen: in the middle of the Four Rivers fountain by Bernini from the sixteenth century, standing in the center of Piazza Navona, there is an obelisk, and on top of that, the Pope who restored the water system in Rome had placed his personal symbol. Byzantine Constantinople developed and was dependent upon a network of long-distance channels and reservoirs feeding the city with water. By ad 373, an estimated 130 new bridges and the first line of channels from major springs had been completed, measuring 268 km in total. Within the city, the Aqueduct of Valens had 87 arches and was 971 m long, one of the longest in the Roman world. The channel system was continuously expanded and developed, and the single length of one of the channels was 227 km. If a supplemented line from the late fourth century is included, the total length was 268 km. But this was still not enough for the city’s water supply and, around ad 400, the system was extended to new springs almost 130 km from the city. When the second phase of the

001-022 WaterUrban Introduction.indd 8

12/05/2014 11:04

Urban Water Systems—A Conceptual Framework9

water supply was completed, about ad 450, the aqueduct channels had a total length of 494 km.13 Machu Picchu was the royal city in the Inca civilization, located in a mountaintop location. The city was established in ad 1450 and abandoned in ad 1572, but most likely ceased operation by ad 1540. Despite the importance of the city and its amazing architecture, the size and population was rather modest. Machu Picchu could support a resident population of about 300 people and up to 1,000 when the royal entourage visited the city. The annual rainfall is nearly 2,000 mm and the location of the city would not have been possible if a reliable source of groundwater had not been available. Water from the main spring was transported in a 749 m canal to the city center, which was distributed by 16 fountains still operating today.14 Building canals and aqueducts was not only important for cities in the past; it has also been intrinsic to many cities’ development today. Los Angeles, although initially a small city with sufficient natural water resources, soon became a “desert city” because of its rapid population increase. From having around 1,500 inhabitants in 1850, the population rose to more than 100,000 in 1900, and only four years later it was 200,000. The solution to the water crisis this caused was to build an almost 360 km aqueduct from the Owens Valley in the early twentieth century, turning Los Angeles into a “hydraulic society”. Yet already in the 1920s this was not enough, and the aqueduct system had to expand due to an increased population of 1.2 million in 1930, causing displacement and environmental degradation in the areas from where the water was withdrawn.15 One of the central issues that the water system approach aims to handle is that water is both an external, physical factor and creates one of the most interconnecting structures in cities in the form of socially appropriated, manmade water systems. In modern cities, all citizens and units—from the smallest apartments to the largest malls and factories—are physically, institutionally, and politically connected through water supply and sewage systems. In many cities, one reservoir is the main source of all the water for the entire population, and one major sewage treatment plant treats the waste water—connecting each and every person in the city through the whole process as the water is used. Thus, the physical character of water and the structures of the manmade water infrastructure are linked, and as such they create both social cohesion and social hierarchies, and these again reflect place specific forms and are continuously changing throughout history. Dar es Salaam is one of the many cities in Africa where the well-being and health of its citizens have been hampered by incomplete and unreliable water supply and sanitation services. Today, only 10 percent have flushing toilets, while the majority is dependent upon pit latrines, which causes increased pollution, jeopardizing health conditions. Rapid urban growth, lack of investments, and unsatisfactory maintenance of the water

001-022 WaterUrban Introduction.indd 9

12/05/2014 11:04


A History of Water

systems, in addition to a large informal sector (without a centralized and governmental planning), have been a hindrance to poverty reduction, and the unsatisfactory water structures are inevitably and directly linked to the prevailing poverty.16 In this case, water has a direct impact on the formation of social structures. Alteration in urban water thus has huge social implications for social relations and formations in general. Manchester was the world’s first industrial city and it was crucial in the Industrial Revolution. Manufacturing of cotton from 1770 led to exceptional economic growth. In the first half of the nineteenth century, the city had the world’s highest concentration of cotton factories. Along with the industry, the population grew as well; in 1841, the city had 300,000 inhabitants and was the second largest city in Britain. A crucial aspect in the Industrial Revolution was the development of the water transport system. In addition to the natural watercourses, 250 km of year-round and ice-free canals, rivers, and streams were constructed within 15 km of Manchester city center. The transport system was manmade but reflected at the same time the specific character of the natural water landscape; the fact that it was raining throughout the year made supply of water for the canals much more easy here than in China or in southern France (Canal du Midi), and it was comparatively easy in the Manchester area to link canals to rivers because the rivers were running very close to each other, etc. The canals provided navigable routes to the major trading regions not only nationally but also globally. Importantly, the new manmade transport system also facilitated transportation of coal for industry on water instead of on land.17 New York is another example of how important it is to study this second layer of the water system on a grand scale, all the time interconnected with the first and third layers. New York started as a modest settlement in 1626; it was named New Amsterdam, and located on an island at the mouth of the Hudson River. It was thus a classical river town, established at the interface of running fresh water and salt water, of river and ocean. There can be no doubt that the Erie Canal, which connected the Hudson to the Great Lakes, opened in 1825, is one major reason why New York and not Boston became the economic capital of the USA. It married the two seas, as Governor Clinton said when the canal was opened. The city became a densely populated metropolis by the mid-nineteenth century, and local water sources were insufficient to meet the rising water demands. The history of New York’s waterworks is therefore among other things a history of how far into the hinterland, away from its original core, a city may go to secure a reliable water supply. New York is dependent upon water from other watersheds, thus challenging the water interests of other states. The Hudson, although it may have the look of a natural river, is not a natural river any more, but largely controlled by man.18 Mexico City, considered by many to be the most populous city in the world, is not located on a riverbank, but has managed to grow due to (among other things) massive investment in water transfer projects.

001-022 WaterUrban Introduction.indd 10

12/05/2014 11:04

Urban Water Systems—A Conceptual Framework11

The city is located in a basin, which is a naturally enclosed depression, at around 2,200 m above sea level. If one flies over this megacity at night, this is one of its clearest markers: there are lights everywhere and no dark thread reflecting a river that runs through the city. The Mexico Valley basin has a large number of springs in the mountains and the foothills, providing water to the lakes and the aquifers, which still supply almost 70 percent of the water requirements for the city’s now 20.5 million inhabitants. However, the city has been overexploiting its water resources for years, and the lake at which it was originally located is not there anymore, except as a bleak, small shadow of itself compared to how it was during the times of the Aztecs. This has caused the groundwater table to lower considerably, as much as 1–1.5 m annually. “The city is sinking while people are drinking,” goes the local saying. This is undermining the foundations of buildings and structures, and the area is becoming more prone to earthquakes and flooding. But since the city is located where it is, catering for a dramatic inflow of people, this means politicians and engineers have to look for water further and further away from the city.19 Las Vegas—one of the fastest-growing cities in the USA—is another example. It has never had enough water locally, since it is located in the middle of a desert. The city relies primarily on water pumped from the Colorado River, and this is stored in Lake Mead. But the amount of water the city is allowed to withdraw from the reservoir is not sufficient. One strategy has been to pump groundwater from eastern Nevada, but this has caused conflict with the interests of Los Angeles. Conservation and recycling processes are in place, but still there are huge challenges in meeting the increasing water requirements. When driving down the main street in Las Vegas, the city’s water infrastructure “lies”. It is difficult to maintain the impression that this is a desert city, because there are cascading waters everywhere—up from gigantic fountains and down from huge artificial waterfalls. The city architecture defies its location, and is— from one point of view—a celebration of the ability to modify the physical waterscape. The gamble on water is, however, still the biggest gamble in the city’s history, and the authorities have to fetch it from water resources far away from where this city was established, originally intended as a resting place for treks through the desert.20 The presence of water—in some cases water scarcity and in other cases water overabundance, like Rotterdam—always has to be modified. Boston is surrounded by lakes and receives a substantial amount of rain each year. The city itself owes its topography to the way water runs in the landscape. Boston was founded as an English colony in 1630. There were 25,000 people in 1800, growing to 140,000 by the mid-century and then to over 500,000 by 1900. As the city expanded, the water world had to be successively altered. The very nature of the water topography shaped the development of the city. In order to facilitate this urbanization, in the nineteenth century, Boston became a hydrological machine, controlling and

001-022 WaterUrban Introduction.indd 11

12/05/2014 11:04


A History of Water

developing its water world. Canals were built, water and sewer systems were created, gigantic water reservoirs were constructed, rivers were dammed, lakes and harbors were dredged, and marshlands were filled. Thus, the very development of Boston was a series of interdependent technologies to collect, channel, deepen, preserve, reroute, and discard— and control and supervise—all of the water.21 Overabundance of water might be a challenge, but on the other hand, even absence of water can and has been managed, although the changes to the natural waterscape may be marginal. The stone towns of the East African coast are part of the Swahili culture that flourished during the fourteenth, fifteenth, and sixteenth centuries. Two of these sites, Songo Mnara in the Kilwa archipelago of Tanzania, and Gede on the southern coast of Kenya, are located in an area where the waterscape is characterized by dryness. Songo Mnara is located on an island, and has no source of fresh water. The inhabitants had to use brackish water from wells dug into the coral. In other coastal islands, all the fresh water had to be transported from the mainland. City development has therefore been intrinsic to overcoming water scarcity by maximizing the resources available and minimizing water usage.22 All cities have naturally developed artificial water storage in some way or another, and in some locations these storage systems need to be bigger and more costly than in other areas. This is not only because the cities vary, but because the physical water systems enveloping particular cities vary. Extreme fluctuations and inequalities in the natural discharge curves of rivers or rains are therefore important aspects of urban development and different cities’ efforts at controlling their waters. For this reason, it is fruitful to think in terms of a water system with different interrelated layers. In order to understand the manmade water system one should also study how it is reflecting the character of the natural water system, and the water system that envelops a city at any point in its development is a product of both natural and social factors. Moreover, the physical water system, which was once more than sufficient for a particular city’s water needs and development, can over time become insufficient because of growing populations or changing economies, in spite of (or because of) physical or manmade changes in the water system. It is this dynamic that ahistorical theories of urbanization processes that only focus on social variables will ignore. This notion of a second “modified” layer as part of a much wider water system can be made clearer by being contrasted with notions about the “built environment”, usually defined as being only a reflection of culture or being a socially constructed environment. The notion may also be clarified by being compared to talk about a “factitiously separated and imagined [italics added] ‘natural’ environment”.23 The second layer of the water system will encourage investigations of a city’s “water machine” or water technological setup as an effort to solve the water question in a context

001-022 WaterUrban Introduction.indd 12

12/05/2014 11:04

Urban Water Systems—A Conceptual Framework13

where both the first layer (the physical and hydrological) and the third layer (ideas and practices about water management) must be studied. Analytically it will include changes in the way the water runs or flows through the urban landscape: as river embankments and canals, dams and reservoirs, pipes beneath the city and into the houses, bottles of water and transportation of them, etc. The water system approach is open and not reductionist: it includes in the analytical picture everything the human being has done to bring natural water to the city—in all sorts of sectors and for all sorts of purposes. It thus enables a description and understanding of the water system that is an integral part of any city’s planning environment at any point in time. The notion underlines that any existing water system impacts possibilities, limitations and patterns of action, as it also reflects the cities’ economy and technological competence, among other things. These two layers guarantee that the water system approach is neither nature-centric nor anthropocentric, but allow analyses that can capture and combine the dynamic and dialectic development between the two as factors of relevance and importance in urban development. This concept underlines that urban places and nature should not be perceived as geographic opposites, where cities are manufactured social creations, and nature is outside of human construction. Water and the city are in this perspective fully intertwined, since cities integrate and use water at every level of development and activity. The flow of water through urban space that has played a pivotal role in freeing cities from disease, squalor, and human misery is water that flows both as a natural and a social element at the same time in the same form. Water provides a link between the material experience of physical and social space and the abstract dynamics of urbanization. As shown, this hydrological dynamic is not restricted to the modern city: archaeological and historical evidence of past complex water engineering projects abounds. Urban configurations have impacted on resource flows across a range of scales. Physical and human-modified water systems, always in flux and always particular, also exhibit a diversity of response to urbanization, so much so that the same urbanization processes may have a different impact on the water system. The waste output of even a small city has overtaxed the absorptive capacity of local aquatic systems in some cases. Furthermore, this aspect will be intrinsically related to the first layer; its effect is contingent upon the physical character of the water system. What should be studied and understood is therefore the relationship between the built environment or the man-modified water system (which is never the sole result of human activity, but also of the water system’s physical properties), and the response of water ecosystems in a diversity of climate, landscape, economic and cultural settings. This human-modified water system successively changes the physical water system in an everlasting process of mutual interaction. These modifications have fundamentally defined the structural, but changeable, context in which cities develop. The

001-022 WaterUrban Introduction.indd 13

12/05/2014 11:04


A History of Water

impact of the urbanization process on the water system is an aspect of this layer. The concept of “open and multifunctional water systems” encourages analyses of the urban place and its confluences with water. This is caused by the form and level of any city’s alteration of the water landscape it interacts with, which mirror technological traditions and managerial ideas. In some cases, this also echoes broader technological and cultural patterns in the particular society in question. Moreover, no urban water landscape is completely natural or controlled, because urban development presupposes modification of the natural waterscape. However, in the long term, all hydraulic structures are vulnerable to climatic changes—whether human or natural. The water world is always changing. Urbanization processes have a direct and indirect impact on the enveloping water system due to human modifications of the environment. Cities themselves impact on precipitation patterns as they influence local climate. Existing water systems thus reflect not only natural and geographical conditions, but also societies’ ability or determination to manipulate their water in the form of damming, draining, canalizing, embanking, storing, piping, or recycling. Human modifications have changed the way water flows through and under the city; where it flows, to whom, and how. And in changing the waterscape and controlling water, the urban population and the urban settlements have also changed themselves. The second layer of an open and multifunctional water system covers those changes that human beings have made to their natural water landscape. These will undeniably vary over time, and hence have to be studied accordingly. The modifications will cover everything from stream alterations and river embankments when cities were first established, tunneling water across great distances through mountains and deserts, to weather modifications by chemical bombardment. Examples of this is what the Israelis do to enhance regional rainfall and what the Chinese did to create nice weather during the 2008 Olympics in Beijing. Two questions arise: within this approach to what extent the modifications are passive adaptations to the natural water landscape; or to what extent they represent an effort to overcome the limitations experienced by the existing water system. This leads naturally to the third level: all modifications are carried out by human agents, and their acts are by necessity influenced and structured by the ideas that it must be possible to reconstruct and identify. The third layer of water systems: ideas and managerial concepts of water The third layer of the water system constitutes the cultural, institutional and conceptual dimensions. This includes the management practices and “habits of thought” or ideas about water and water control, its religious

001-022 WaterUrban Introduction.indd 14

12/05/2014 11:04

Urban Water Systems—A Conceptual Framework15

or spiritual significance, including notions about purity, and other conceptualizations of water that have developed over time in different urban contexts. This concept also encompasses the importance and permanence of water management practices in all urban places, and how these practices and habits of thought have been influenced over the centuries to different degrees by the physical and hydrological context (layer one), and the historical water control context (layer two) in which the actors operate. This concept then does not ignore the ways in which nature or water is socially constructed. The way water is conceived in different cities shows the endurance and the instability of meaning, and the coherence and fragmentation of habits of thoughts when it comes to water and water control. It also places this production of cultural metaphors in a water system context. The water system that, at any given time, envelops a city or is part of the infrastructure and linkages of the city is not necessary or deterministic. A city could have developed in other directions, given the limitations and possibilities of the actual waterscape (level one) and the means available for modification (level two). The human modifications vary according to technological level, economic strengths and organizational capabilities. In addition, they definitely reflect the ideas of individual community leaders from the past who, for religious or other reasons, wanted to do something with their water. They also reflect the visions of individual entrepreneurs of modern times who might champion huge water control projects, which have been opposed by the majority for a long time. The human element— the agent of change—must not be left out, and one way of securing this is by not talking about “material structures”, but about physical and manmade structures creating what is, at any point in time, the existing material context. Without a doubt, a focus on water/urban relationships does not imply a view on human beings as agents through whom an active nature constantly works.24 This focus does not imply a history without subjects or a history opposed to narration and chronology, because water is also constructed discursively and materially, and it is undoubtedly implicated in the exercise of social power. Not only that; it is also part of religion and cosmology as a whole. Take the case of Motya—a little island of 45 ha located 1 km away from the coast of Sicily. The urban settlement was structured around a sacred compound named the Temple of the Kothon. The temple was erected close to a fresh water spring. This was an underground water source from where a small lake naturally emerged. The location of the temple and the city in relation to the water suggests that the water was put under control of a divine authority, and most likely that it was holy. A water deity was apparently the focal core in the urbanization process.25 The first Phoenician settlement occurred in the first half of the eighth century bc, and it was a flourishing city from the end of that century to the beginning of the fourth century bc.

001-022 WaterUrban Introduction.indd 15

12/05/2014 11:04


A History of Water

Varanasi in India, located along the Ganges, is the most sacred Hindu pilgrimage site and is believed to be the oldest city in the world that is still inhabited. The continuous arrival of pilgrims originates from the religious ideas of the deity Ganga, and these religious ideas of water have more than anything influenced the history of the city. The constructions of ghats (stairs leading down to a river) where people can burn their relatives, before they throw the ashes to the river in order to escape the eternal cycle, have fundamentally shaped the urban topography and the cultural construction of the city itself. Today, the ghats form a monumental and continuous riverfront more than 6 km long. They create continuity between the city and the river, materializing a particular space and cosmology.26 Religion and cosmology are in some cities the most important structuring mechanisms. If one then moves from the pure to the impure—from holiness to pollution—changing ideas of sanitation and sewage have been one of the key drivers in urban water developments. When in England in 1829, Paul Pry’s cartoon showed the organisms that a microscope could reveal—an image revealing how little beasties populated the water—it was part of a changing understanding of water, which had great implications. His caption reads: “Monster Soup commonly called Thames Water.” When London later was hit by cholera epidemics, it was understood that pathogenic organisms—bacteria, worms, viruses and fungi—were responsible for outbreaks of waterborne diseases, and the water system underpinning London’s growth was changed. The management of water now became the management of the city’s health, and this idea had far-reaching consequences for urban development all over the globe. In fact, developments in many cities’ infrastructure were due to sanitation rather than a need for increased water supply. By the end of the eighteenth century, Paris, with more than 500,000 inhabitants, was three times more densely populated than London. Around 1850, the city had passed 1 million people. Especially after the cholera epidemic of 1832, the sewer network in Paris was expanded, and in that year there were 78 bathing facilities in the city. Paris was the most important river port in France and the river became severely polluted. Due to increased urbanization and industrialization, it was only in the 1960s that the first plans and policies to rehabilitate the river and its banks were developed, and the actual projects were implemented in the 1980s. Deindustrialization along the river front, together with the “polluter pays” principle and large investments in the sanitation sector (especially from the 1980s), have since then significantly improved the water quality.27 In Vienna, the development of the sanitation system can be described in three different phases, but only the latter made a watershed. In the first phase, called Amphibious Sanitation, which lasted from Roman to early modern times, sanitation practices were local in scale where the inhabitants adopted and made use of natural water resources. In the second

001-022 WaterUrban Introduction.indd 16

12/05/2014 11:04

Urban Water Systems—A Conceptual Framework17

phase, called Patchworked Sanitation, which lasted from the 1700s to the 1860s, the Habsburg authorities drew attention to sanitary issues, providing punctual solutions at specific sites throughout the city. In the third phase, called Integrated Sanitation, approximately from the 1860s to 1910, sanitation was actively included in urban planning. Thus, Vienna’s waterscape was subject to major undertakings as the city grew, providing for this change in its metabolism.28 In other cases, there are examples that it was the need for neither safe water supply nor improved sanitation that triggered urban water developments. Bergen is the city in the Western Hemisphere that receives the most rain, with an average of 2,250 mm a year. However, the city has also on numerous occasions experienced water scarcity, and the general abundance of water has required specific developments to be made regarding water supply and sewage systems. In relation to Norway’s capital Oslo in the nineteenth century, Bergen as a former capital lagged behind, and it was internal national rivalry between the cities which prompted development in the water sector in Bergen, rather than combating diseases and desire for public good and health. Moreover, development in water supplies was not followed by sewage systems to the same extent. The new health regulations from 1865 banned homemade WCs and forced citizens to erect outhouses with buckets, which ideally should result in waste product being able to be used in agriculture. However, due to the rainy weather in Bergen, the manure could not be dried and the plan failed. It was only in the beginning of the twentieth century that the idea of the WC won through and was implemented on a larger scale.29 Yet another reason for development in urban water systems was the problem of a different kind that cities faced: fire. Houston was a city with a population of only 1,500 people in 1837, but grew rapidly to become the fourth largest city in the USA and today it contains more than 2 million people. The city is the regional and commercial center commonly known as the “energy capital of the world”. All this is possible because of the water infrastructure, but this was built rather late. The first public water supply was built in the 1870s, when the city had emerged as a city center. One of the major reasons for constructing a public water system was the need for fire protection. Thus, it was not the public health with regards to waterborne diseases and a safe water supply that was the first and main driver,30 but all of these developments were eventually made possible by political decisions. Policies and managerial practices are thus crucial for a city’s development, for better or worse. Water deficit may also be manmade due to a number of reasons. Damascus was renowned for its abundant water resources and variously described as “bride of cities”, “the grain of beauty on the world’s cheek”, and “a lover’s torment”, and indeed was seen as the biblical Garden of Eden. The city’s strategic location by the Barada River and as a crossroads from India, Persia, the Arabian Peninsula, East Africa,

001-022 WaterUrban Introduction.indd 17

12/05/2014 11:04


A History of Water

and Anatolia made it an economic, cultural, and religious center. Today, it is a city characterized by decline, uncontrolled urbanism, and acute water scarcity. All of these factors are to a large extent the consequences of neglect and bad policies for decades. High population growth, pollution, over-utilization of water resources, unsustainable groundwater extraction, and lack of technical capacity and political will, in addition to limited capacity to enforce water management strategy, have all led to a deteriorating situation.31 On the other hand, Singapore is a success story when it comes to development and water. Four decades ago, the country and city suffered from slums, poverty, and waterborne diseases. From this position, Singapore has become a rising star. Given the minimal available water resources present, advanced rain-harvesting techniques collect all the rainwater. Treated waste water is used for drinking and all water is recycled and reused. Thus, Singapore has been able to overcome water shortages and insecurity by being at the forefront of developing sustainable practices. However, Singapore still has to import water from Malaysia. The water import started as early as 1927 and has caused tensions between the two countries. An agreement was signed in 1962, providing Singapore with water until 2061, but Malaysia has threatened to turn off the tap if Singapore’s policies were to be prejudicial to Malaysia’s interests.32 Thus, the availability of water resources is often a source of contestation within a city as well as in a country and between nations. This is also evident in the conflicts and challenges rising from privatization of water. The water supply and sanitation services in Manila metropolis in the Philippines were privatized in 1997. Concession contracts involved in such processes present a fine line between flexibility and accountability. In practice, this is a difficult balance and the regulator’s role often becomes extended to an unintended degree. This includes confusion about responsibilities, inconsistency in regulatory decisions, and opportunistic choices and implementations.33 Thus, the third layer of a water system often represents the decisive factor in the actual physical layout of a city’s water structure and the water use by whom and at what time. It impacts on how the natural waterscape (level one) is understood and modified for different purposes (level two), like building a canal to supply a city with water (whether it is pipes from Malaysia to Singapore or aqueducts in Rome, Constantinople, or Los Angeles). It is the waterscape (absence of water) which sets the premises. Canals and aqueducts modify the waterscape, and this is only possible because of ideas, managerial plans, laws, and regulations constituting the practices for developing these water structures. It is precisely, therefore, that these three levels should be seen most fruitfully as analytical categories which are best understood in relation to each other. All cities are outcomes of how these three levels influence each other and none of these levels can work in isolation from each other.

001-022 WaterUrban Introduction.indd 18

12/05/2014 11:04

Urban Water Systems—A Conceptual Framework19

Conclusion Urbanization is one of the most dominant forces in the world of today. Nature, peoples, and cities are woven together in an inseparable dialectic of destruction and creation, changing the face of the earth. Water and the concept of an open and multifunctional water system as an analytical approach is a vehicle for understanding this development. The city— an attraction that has fascinated writers and researchers from a variety of disciplines—is in practice incomprehensible to individuals, both researchers and planners, due to its totality and complexity. Analytical approaches should therefore be seen as complementary, since the strength of one might be the weakness of the other and vice versa. A focus on water should also be seen as helpful to understanding other flows and interlinked developments and processes in history; as that of technology, labor, and people, and how cities produce urban realities and landscapes. From this perspective, a water system approach should be able to encourage a much broader type of research on urban development— multidimensional and multidisciplinary in approach—because such an approach will be helpful in order to gain a better understanding of the urban world. Theoretically, it is important to highlight the role of water, because it is important to learn to look at and to think about urban space not just as culture, as a reservoir of collective memory and imagination, or as an engineering relation, but also as nature, and on how these factors are interconnected. The aim of this volume is to summarize some of the case studies that have been carried out, and to present some historicgeographical illustrations of how the role of water in urban history can be studied and understood. By proposing an open inclusive, and nonreductionist approach to urban studies, it will hopefully also stimulate, and be part of, a new and broader research agenda on cities in general, and on water and urbanism in particular. Notes  1 There are of course a number of definitions of towns, cities, and urban centers, and no general agreement on one definition can be reached. Max Weber argued in The City that a full urban settlement must display the following features: fortification; market; a court of its own and at least partially autonomous law; a related form of association and partial autonomy and voting rights (Weber, 1958). This definition looks very outdated today and Western-biased. Ira S. Lowry in “World Urbanizationin Perspective” argued that “Rules of classification vary between nations, but it will do for now to define an ‘urban place’ as any permanent settlement containing at least 2,000 people who are not engaged in agriculture and who live within easy walking distance of one another, and a city as a similar dense settlement with at least 100,000 inhabitants” (1990: 148). This definition is too present-oriented

001-022 WaterUrban Introduction.indd 19

12/05/2014 11:04


A History of Water

and non-Western biased, since a great number of the cities in Europe (and elsewhere) fall outside the definition of a city. In this volume, urban places, cities, and towns are used interchangeably.   2 As the ancients were well aware, water was not the only essential element. Without air, death strikes in about five minutes. Without water, it takes a thousand times longer to die: three to four days. But from a historical and social perspective, air cannot be tamed, piped, controlled, or diverted (except in a number of specific purposes such as scuba diving, submarines, firefighting, space travel, etc.). But in a city, there is never overabundance of air, and it is not delivered in constant flux the same as water. From the point of view of urban development, the parallel to water is, therefore, not very interesting, except when it carries “alien” matter and becomes a major pollutant as it happened in “smoggy” London in the early twentieth century, and in Beijing in the early twenty-first century.   3 See Nigro, Chapter 1 of this volume.   4 See Pedersén, Chapter 5 of this volume.   5 See Jansen, Chapter 2 of this volume.   6 See Sulas, Chapter 8 of this volume.   7 See Koutsoyiannis and Patrikiou, Chapter 6 of this volume.   8 See Finneran, Chapter 12 of this volume.   9 See Hooimeijer and Meyer, Chapter 25 of this volume. 10 See Katko and Juuti, Chapter 22 of this volume. 11 See Wang and Chen, Chapter 3 of this volume. 12 See Rinne, Chapter 7 of this volume. 13 See Crow, Chapter 10 of this volume. 14 See Wright, Chapter 9 of this volume. 15 See Klaver and Frith, Chapter 23 of this volume. 16 See Kjellén and Kyessi, Chapter 24 of this volume. 17 See Maw, Chapter 19 of this volume. 18 See McCully, Chapter 16 of this volume. 19 See Romero-Lankao and Gnatz, Chapter 27 of this volume. 20 See Wilds, Chapter 29 of this volume. 21 See Rawson, Chapter 18 of this volume. 22 See Wynne-Jones and Fleisher, Chapter 11 of this volume. 23 See Harvey, D. (1996). 24 For this viewpoint, see Sauers, C. O. (1970). 25 See Spagnoli, Chapter 4 of this volume. 26 See Jalais, Chapter 13 of this volume. 27 See Barles and Guillerme, Chapter 17 of this volume. 28 See Neundlinger, Gierlinger, Pollack, and Krausmann, Chapter 15 of this volume. 29 See Hammerborg and Byrkjeland, Chapter 21 of this volume. 30 See Melosi, Chapter 20 of this volume. 31 See de Châtel, Chapter 14 of this volume. 32 See Araral and Ching, Chapter 28 of this volume. 33 See Ching and Wu, Chapter 26 of this volume.

001-022 WaterUrban Introduction.indd 20

12/05/2014 11:04

Urban Water Systems—A Conceptual Framework21

references Davies, Jonathan and David L. Imbroscio (eds) (2009). Theories of Urban Politics, London: Sage. Eade, J. and C. Mele (eds) (2002). Understanding the City: Contemporary and Future Perspectives, Oxford: Blackwell. Hall, P. G. (1998). Cities in Civilization: Culture, Innovation and Urban Order, London: Weidenfeld & Nicolson. ——— (2002). Cities of Tomorrow, An Intellectual History of Urban Planning and Design in The Twentieth Century, Oxford: Basil Blackwell. Harvey, D. (1996). “Cities or urbanization”, City, 1(2), pp. 38–63: 60. Lowry, I. S. (1990). “World Urbanization in Perspective”, Population and Development Review, Vol. 16, Supplement: Resources, Environment, and Population: Present Knowledge, Future Options, pp. 148–76. Orum, A. M. and X. Chen (2003). The World of Cities. Places in Comparative and Historical Perspective, London: Blackwell. Parker, S. (2003). Urban Theory and the Urban Experience.: Encountering the City, London: Routledge. Sauers, C. O. (1970). Man’s Role in Changing the Face of the Earth, Chicago, IL: University of Chicago Press. Tvedt, T. (2010a). “Water systems, environmental history and the deconstruction of nature”, Environment and History, 16(2), pp. 143–66. ——— (2010b). “Why England and not China and India? Water systems and the history of the industrial revolution”, Journal of Global History, 5, pp. 29–50. Weber, M. (1958). The City (transl. and ed. by R. Martindale and G. Neuwirth), New York: The Free Press.

001-022 WaterUrban Introduction.indd 21

12/05/2014 11:04

001-022 WaterUrban Introduction.indd 22

12/05/2014 11:04


Aside the Spring: Tell Es-Sultan/Ancient Jericho: The Tale of an Early City and Water Control in Ancient Palestine

Lorenzo Nigro Introduction Tell es-Sultan/ancient Jericho—nowadays an archaeological park in the Palestinian territories under the aegis of the Ministry of Tourism and Antiquities, Department of Antiquities and Cultural Heritage—was one of the earliest towns of the Near East, exemplifying the extraordinary phenomenon that is the formation and growth of an integrated human society, a sociocultural achievement that gained the site the universally renowned title of “the oldest city in the world”. Jericho was one of the earliest sites in the Near East (and, indeed, in the world: as early as 10,500 years before present (BP)) to attain such a status (illustrated first of all by impressive architectural works—basically settlement defenses—defining the town space; Kenyon, 1957: 65–9). It was also one of the earliest to develop into a flourishing city (at the beginning of the third millennium bc), thereby illustrating the rise of an urban center as manifest by impressive public architecture, social organization, economy, trade, craftsmanship, and international relations (illustrated by numerous finds from the site itself and from the nearby necropolis; Nigro, 2010). This chapter discusses the urban development of Jericho from the perspective of water use and its contribution to the establishment and growth of the city, from the earliest times to the Ottoman Period. The clue to the early success of this human community settled at the foot of the Jebel Quruntul (the Mount of Temptation), on the limestone plateau flanking the alluvial deposits of the Jordan River, most probably lies in the ‘Ain es-Sultan (also known in the Bible as Prophet Elisha’s Spring). This water source provided 4,000–5,000 liters of fresh water each minute (Figure 1.1). Such a generous flow of water (Figure 1.2) made possible the early development of animal breeding and agriculture, the domestication of sheep, goats, and also cattle, as well as the growing of cereals and horticultural products. Located at the westernmost tip of the Fertile Crescent, on the rift valley connecting Asia and Africa,

023-051 WaterUrban Pt1 Ch01 Nigro.indd 25

12/05/2014 11:10

023-051 WaterUrban Pt1 Ch01 Nigro.indd 26

12/05/2014 11:10

Figure 1.1.  View of the site of Tell es-Sultan, cut off to the east by the modern road, and of the nearby spring of ‘Ain es-Sultan.

The Tale of an Early City and Water Control in Ancient Palestine27

Figure 1.2.  The spring of ‘Ain es-Sultan after the rehabilitation works carried out in 2009–10 by the Ariha Municipality.

Jericho thus witnessed epochal steps in the story of humanity (Helms, 1988; Taha and Qleibo, 2010). Early steps in human culture (10500–4500 bc) In the Mesolithic or Natufian Period (10500–8500 bc), groups of gatherers and hunters started to camp on the limestone plateau overlooking the spring of ‘Ain es-Sultan. Their diet was based upon the hunting of the gazelle, which was widely spread in the Jordan Valley. Gradually this became a stable occupation, giving rise to a settlement that already in Pre-Pottery Neolithic times (8500–6000 bc) marked a definite step in human cultural development. The earliest settlement (Pre-Pottery Neolithic A: 8500–7500 bc) consisted of round single-roomed houses with walls made of loaf-shaped mud bricks rising up to a domed, tapering

023-051 WaterUrban Pt1 Ch01 Nigro.indd 27

12/05/2014 11:10


A History of Water

roof (similar to the still-existing “beehive” houses of northern Syria for both storage and housing) (Kenyon, 1957: 70–2). This kind of architecture was made possible by a plentiful and continuous supply of water. The village was then encircled by a defensive wall made of stones, and had at least one huge well-known building: a round tower with a base diameter of 8.5 m, a preserved height of 8 m (Figure 1.3), and an inner staircase with 22 stone steps leading to the top (Kenyon, 1957: 67–9; 1981: 18–43, pls. 5–11, 203–12). According to the town’s excavator, Dame Kathleen Kenyon, this building was for defensive purposes, along with the attached “Town Wall”. Inside the tower, a layer of destruction was found containing the skeletons of 12 people who had apparently been killed during a fight or a riot (Kenyon, 1981: 32–3). Such finds shed light on the early history of one of the first communities in the world to control their water supply. This necessitated the development of protective devices and an embryonic management apparatus—these arose within the social groups living in the southern Jordan Valley. Moreover, thanks to the spring waters, Pre-Pottery Neolithic A inhabitants of Jericho were able to domesticate plants (especially cereals: wild barley, emmer wheat, and einkorn) and animals (caprovines), thus laying the foundation for agriculture and livestock farming. Canals were dug from the spring, transforming the wild dry country into a cultivated oasis. In the following Pre-Pottery Neolithic B period (7500–6000 bc), the introduction of cigar-shaped, thumb-impressed mud bricks allowed the shift to rectilinear architecture, characterized by finely plastered walls and floors (Figure 1.4; Kenyon, 1957: 52–6). A typical feature of this period is the first appearance of plastic works associated with the rise of ancestor cults. A series of plastered human skulls with eye seashell inlays and painted decorations, also known in other Pre-Pottery Neolithic B sites of the Levant (Beisamoun, Tell Ramad, Yiftahel, Kfar Hahoresh, and ‘Ain Ghazal; Milevski et al., 2008; Marchand, 2011–12), were found beneath house floors (Kenyon, 1957: 60–4; 1981: 77, pls. 50b–9c). Ancestor cults are apparently associated with the representation of the human figure by means of clay, marl, and water: a simply stylized human bust was found by Kenyon on the eastern side of Square DII in the upper Pre-Pottery Neolithic B layers (Kenyon, 1981: 531, pl. 72). It recalls two groups of stylized clay statues retrieved by Garstang in his North-Eastern Trench (190 and 195), in a stratigraphic spot attributable to Pre-Pottery Neolithic B or Pottery Neolithic A. Group 195 included a man, a woman, and a child (a kind of triad?), of which only the male head is renowned due to its state of preservation (Garstang, 1935: 355–6; Garstang et al., 1935: 166–7; Sala, 2006: 275–6). The statues are akin to the famous specimens from ‘Ain Ghazal in Jordan (Rollefson, 2000), and they illustrate the ability to use water and clay, anticipating the invention of pottery. The following stage in the history of Tell es-Sultan was, in fact, indelibly marked by the introduction of pottery vessels as one of the common

023-051 WaterUrban Pt1 Ch01 Nigro.indd 28

12/05/2014 11:10

The Tale of an Early City and Water Control in Ancient Palestine29

Figure 1.3.  Tell es-Sultan: Pre-Pottery Neolithic A (8500–7500 Kenyon’s Trench I.


round tower in

artifacts of ordinary life. Actually, pottery was locally produced when technological features and architecture showed a cultural regression. The Pottery Neolithic A (6000–5000 bc) village consisted of pit dwellings dug into the erosion layers (or even in previous Pre-Pottery Neolithic B strata), and flint and stone tools exhibit a cruder and less refined treatment in

023-051 WaterUrban Pt1 Ch01 Nigro.indd 29

12/05/2014 11:10


A History of Water

Figure 1.4.  Tell es-Sultan: Pre-Pottery Neolithic B (7500–6000 bc) house excavated by M. K. Kenyon in Trench III (after Kenyon, 1981: pls. 115, 138b–c).

respect of those of Pre-Pottery Neolithic. Earliest pottery was characterized by coarse, straw-tempered ware, usually decorated by burnished redon-cream painted decoration. Nonetheless, it represents an extraordinary innovation in the economy of this early rural community. Pottery Neolithic B (5000–4500 bc) Jericho exhibits once again the shift to rectilinear architecture associated with the diffusion of a new kind of mud brick (bun-shaped). Nonetheless, Pottery Neolithic A and B layers produced a few remnants of plants and charcoal, thus suggesting that the Jerichoans were basically herdsmen and hunters during these periods. Pottery Neolithic is viewed as a culturally recessive period, when the site was slightly reduced in area and hosted a less developed village in respect of the preceding Pre-Pottery Neolithic (Kenyon, 1957: 77–92). This was possibly connected with a shift and reduction in the flow of water from the spring, due to earthquakes that occurred at the end of Pre-Pottery Neolithic.1 Although it occurred in a negative sense in this case, water did significantly influence human life at Tell es-Sultan. The following Chalcolithic (4500–3400 bc) Period records a deal of marginalization of the site, possibly due to a reduction in the flow capacity of the spring. The major site in the oasis was Tell el-Mafjar (Taha et al.,

023-051 WaterUrban Pt1 Ch01 Nigro.indd 30

12/05/2014 11:10

The Tale of an Early City and Water Control in Ancient Palestine31

2004), on the northern bank of Wadi Nueima, around 2 km north-east of Tell es-Sultan. Several other camps existed on the northern bank of Wadi el-Qelt. At Tell es-Sultan, sparse ceramic finds and flint tools testify to an ephemeral occupation of the mound flanks during this period (Nigro et al., 2011a: 7–9). Water in the Early Bronze Age: from agriculture to urban culture (3400–3000 bc) During the second half of the fourth millennium bc, a group of seminomadic people, bearing a distinctly different material culture from that of the Neolithic, settled upon the tell alongside the spring of ‘Ain es-Sultan (Nigro, 2005). They arrived from the eastern highlands of Jordan, bringing their dead: disarticulated skeletons were buried in familiar tombs, using caves cut into the limestone bedrock of the plateau northwest of the spring, where skulls were usually piled in the middle of the cave, and long bones were accumulated at the sides (as illustrated by a series of tombs excavated by Kenyon: A13, A84, A94, A114, A124, A130+A61, K1, K2; Kenyon, 1960: 4–51; 1965: 3–32). The spring presumably induced this community to become sedentary and progressively to implement agricultural production, so that a new rural village grew steadily, with circular huts and annexed storage facilities and food-producing devices in Early Bronze IA (3300–3200 bc; Nigro, 2005: 198–9; 2007: 14–17). The transformation from a seminomadic/pastoralist society of hunters and herders into a flourishing agricultural community of farmers practicing intensive animal breeding in the oasis (made possible by the conspicuous water supply of the spring) is illustrated by finds and pottery, as well as by gradual changes in the burial custom, with the introduction of primary depositions and the inclusion of food offerings in open-shape ceramic vessels among the funerary equipment. Somewhat rare elements retrieved in tombs are pierced goat bones (caprovine metacarpals), incised with the schematic representation of a human face, which were interpreted as flutes, which again testify to the increasing complexity of funerary ideology related to a rural sedentary society. It is not easy to correlate sedentarization with water control as a basic economic lever supporting agricultural surplus accumulation and prompting inner social differentiation of the community. A hint at such kind of phenomena is perhaps offered by a major elongated apsidal building erected in the Early Bronze IB, overlooking the spring and the underground cultivated oasis, possibly devoted to communal activities, excavated by Kenyon (Kenyon, 1981: 322, pls. 174, 176a, 313b; Nigro, 2005: 122–6). Other distinguishing features of the earliest Early Bronze I village are the broad-room temple with raised platform and niche (Shrine 420),

023-051 WaterUrban Pt1 Ch01 Nigro.indd 31

12/05/2014 11:10


A History of Water

containing marble and stone cultic furniture, excavated by J. Garstang (Sala, 2005); and a terrace wall (a mud brick structure on stone foundations), excavated by both Garstang (in the North-Eastern Trench; Nigro, 2005: 18–25, 35) and Kenyon (Wall ZZE–ZZT in Square EIV, and Wall EO in Squares DI–FI; Kenyon, 1981: 96, 315–22, pls. 77–8, 229a, 313–14; Nigro, 2005: 111–12, 120–2). These indicate that the inner space of the village had been progressively organized according to an established layout. In the latest stage (Early Bronze IB, 3200–3000 bc), rectangular houses take the place of circular ones, compounds are more clearly delimited, and at least a street is enucleated, leading from the spring inside the dwelt area to the temple (Nigro, 2005: 200–2; 2007: 18–20). The latest stage of development of the village also shows a major increase in cereals accumulation, and a progressive specialization of pottery production. The latter were described by Kenyon, and used to identify different Early Bronze I cultural groups (Kenyon, 1960: 4–10). Actually, simple, storage, painted and redburnished wares served to different economic and symbolic functions. They vary in fabric and surface treatment, as well as in decoration (Grain Wash and Band Slip, Red-Burnished and Line-Painted Wares), and vividly depict the progressive cultural growth of the Jerichoan community during the last quarter of the fourth millennium bc. Finally, the growing presence of status symbols during Early Bronze I, such as Egyptian or Egyptianizing items (e.g. stone mace-heads and palettes), indicates the inclusion of Jericho into the Egyptian trading network and reflects a cultural influence that is typical of the “Proto-Urban” Period in the Levant. The rise of the Early Bronze II–III (3000–2300 bc) city and water control At the beginning of the third millennium bc, the flourishing Early Bronze I rural village was transformed into a strongly fortified city, thanks to a massive utilization of the water resource (Nigro, 2010: 1–5). A defensive line, consisting of a mud brick wall, encircled the dwelling area for a length of approximately 1 km, thus proclaiming the urban status of Jericho (Nigro, 2006a: 355–61; 2010: 11–36). This structure, standing upon a solid stone foundation, was 2–2.5 m wide, and reached a presumed height of 5–6 m. It was made of dune yellowish bricks with a thick ashy mortar in between, realized through the use of large quantities of water. The richness of the spring made it possible to supply both agricultural and building needs at the same time, without affecting food supply for workers and the ruling elite. Semicircular towers and massive bastions protruded from the defensive line to the west and the north, while to the east the city wall apparently included the spring and its immediate surroundings within the newly arisen city. The palace of the rulers of the city was

023-051 WaterUrban Pt1 Ch01 Nigro.indd 32

12/05/2014 11:10

The Tale of an Early City and Water Control in Ancient Palestine33

erected on the eastern flanks of the hill overlooking the spring, while on the opposite western side, the city temple was built (Nigro, 2010: 51–4). This urban layout makes it clear that the city arose when an institution imposed direct control on the main water source of ‘Ain es-Sultan. This made water available both for agriculture (irrigated land was presumably under the administration of the Jericho rulers) and for building activities (basically the erection of the city walls, where at least 1 million bricks were employed together with mortar and plaster, all requiring at least 2 million gallons of water). The urban authority was presumably responsible for the maintenance of the spring and pools created to regulate the oasis irrigation. Ceramic fragments retrieved during a survey in the area of the spring in 2009 corroborate such a hypothesis (Nigro, 2010: 57–61). It may be surmised that the subdivision of the main flow of the spring into four streams or canals was first established in this period. The improvement in agricultural production that occurred during the Early Bronze IB and that was further enhanced in urban Early Bronze II, as well as controlled access to the spring, brought about an accumulating food surplus that provided the economic foundations for the development of the early urban society of Jericho. Warehouses and storage jars at the site, as well as a number of other indicators in the material culture, point to economic growth, with the development of trade (including over long distance) for the collection and distribution of precious stuffs (salt, bitumen, sulfur from the Dead Sea, wool), and raw material (copper, wood, stone). At the heart of such a system was farming in the oasis, made luxuriant by the generous flow of ‘Ain es-Sultan. During Early Bronze III (2700–2350 bc), Jericho reached the peak of its growth in the third millennium bc: the city wall was doubled through the addition of an Outer Wall to the main Inner Wall (Figure 1.5) at a constant distance of approximately 4 m (Nigro, 2006a: 361–75; 2006b: 8–9; Nigro and Taha, 2009: 738–40). The space in between the two structures, again built of mud bricks on stone foundations, was kept free for pathways or storerooms, or filled up with crushed limestone in order to strengthen the whole defensive line. Wooden beams (tamarisk and pistachio), and reeds set across the structure, served as chains and draining devices. Such an impressive work remained for millennia one of the distinctive emerging features of Tell es-Sultan, possibly inspiring the Biblical author in the Book of Joshua (6:1–27) to mention these ruins to support the reliability of his story. The spring was again under the direct control of the palace, rebuilt on the eastern flank of the Spring Hill (Palace G, excavated by the Italian– Palestinian Expedition; Figure 1.6; Nigro et al., 2011b: 200–6); and other public buildings were excavated on the southern side (Building B1, possibly devoted to communal food processing; Marchetti and Nigro, 1998: 39–9; 2000: 130–8; Nigro, 2006b: 18–20), and at the northwest corner of the city wall (where a rectangular tower, excavated by Sellin and

023-051 WaterUrban Pt1 Ch01 Nigro.indd 33

12/05/2014 11:10

023-051 WaterUrban Pt1 Ch01 Nigro.indd 34

12/05/2014 11:10

Figure 1.5.  Tell es-Sultan: the southwestern corner of the site with the EB III (2700–2350 es-Sultan east of the tell, from the southeast.


mud brick inner wall, and the spring of ‘Ain

The Tale of an Early City and Water Control in Ancient Palestine35

Watzinger, stood; Sellin and Watzinger, 1913: 23–4, fig. 7; Nigro, 2006a: 367–9). Material culture, as evidenced also by tomb assemblages and finds, was characterized by a rich and specialized pottery inventory (a potter’s wheel was also found on the Spring Hill), by the presence of finely worked pieces of craftsmanship (namely stone and ivory bulls’ heads), and by the adoption of cylinder seals, possibly within a kind of palatial system of distribution of goods, including an exchange system for metals (copper, silver, and gold) and other precious stuff (salt, ointments, perfumes, sulfur, bitumen, etc.), as balance weights testify (Nigro, 2006b: 13–17). Jericho was at a pivotal crossroads in the early trade network of urban Palestine, as indicated by finds: seashells, exotic animals (hippopotamus), several items illustrating connection with Egypt (lotus vases, mace-heads, slate palettes, and the so-called “Abydos” Ware in Early Bronze II), and Khirbet Kerak Ware (either imported or a locally produced imitation) pointing to a northern influence (Sala, 2008; Nigro, 2009b: 69–75). Finds from the 12 excavated familiar tombs in use during Early Bronze II–III help to give a picture of this flourishing stage of Jericho’s history (Kenyon, 1960: 52–79). They were rich in pottery, including open shapes for food offerings, and jugs and juglets for ointments. Personal ornaments, cylinder seals, and other precious or rank items were also found, illustrating a variety of stones and gems traded. Among valuable finds, a crescentshaped copper axe-head (form Tomb A114[B]) and a dagger (Tomb F5) have to be mentioned, as well as a bull’s head made of fine limestone with colored shell inlays from Tomb D12. Although it is very difficult to estimate the overall population of ancient Jericho, one may surmise that at least 4,000–5,000 people lived in the city at its maximum floruit in the Early Bronze Age.

Figure 1.6.  General view of the Spring Hill of Tell es-Sultan, with EB IIIB (2500–2350 bc) Palace G (to the right), the spring of ‘Ain es-Sultan and the surrounding oasis (to the left), from the north.

023-051 WaterUrban Pt1 Ch01 Nigro.indd 35

12/05/2014 11:10


A History of Water

In spite of its massive fortifications, the Early Bronze II–III city was destroyed by an intense fire, as part of an enemy attack, at the end of Early Bronze IIIB, around 2350 bc, and did not recover for several decades. The urban interludium: Early Bronze IV village and necropolis (2300–2000 bc) In the last quarter of the third millennium bc, the spring of ‘Ain es-Sultan in the Jericho oasis once again exercised its attractive power, sustaining a rebirth of the site. A new community of semi-nomads settled on the ruins of the ancient city in Early Bronze IV (called by Kenyon “Intermediate Bronze Age”; Kenyon, 1957: 186–209), inaugurating a new burial custom in the necropolis. Rock-cut tombs, entered through vertical shafts, hosted individual primary burials with simple funerary furnishings— basically small pottery jars and copper daggers in the case of male burials; beads and other simple personal ornaments in the case of female ones. More than 350 tombs of this kind were excavated by Kenyon (Kenyon, 1960: 180–262; 1965: 33–166), who distinguished groups on the basis of tombs types (Dagger, Pottery, Bead, Square-shaft, Composite, Outsize, and Multiple Burials), and considered this new group a vanguard of the Amorites, the new population entering the Levant from the south at the end of the third millennium bc. Actually, the evidence from the tombs suggests a tribal organization in the early stage (Early Bronze IVA, 2300–2200 bc) and integration into a large rural community, incorporating northern influences at a mature stage of the period (Early Bronze IVB, 2200–2000 bc). The same cultural horizon was also excavated on the tell, and was summarized in a recent study (Nigro, 2003), distinguishing an early stage (Early Bronze IVA), when a rural village occupied the central hill of the tell, and a later stage (Early Bronze IVB), when also the slopes of the tell were occupied by houses and domestic installations. During Early Bronze IVB, Jericho was actually a huge settlement, hosting a flourishing community with a distinguished ceramic production (in fabrics, shapes, and surface treatments), characterized by the use of a fine combing for decorating jars’ shoulders, in order to hide the junction between the handmade bodies and the wheel-made necks of vessels. Copper and bronze daggers, and other items (a hoard of copper axes, including a broad fenestrated specimen, was found in the northwestern corner of the site by Sellin and Watzinger), point to a development of bronze technology during the last stage of Early Bronze IV, possibly transmitted by itinerant specialists in metallurgy.

023-051 WaterUrban Pt1 Ch01 Nigro.indd 36

12/05/2014 11:10

The Tale of an Early City and Water Control in Ancient Palestine37

Lords of Ruha, “Administrators of Canals”: the Canaanean city of Middle Bronze I–III (1900–1550 bc) At the beginning of the second millennium bc, a new city arose at Tell es-Sultan, again with its center on the Spring Hill, where a new palace (excavated by Garstang and by the Italian–Palestinian Expedition) and a temple were erected (Nigro, 2009a). Oasis irrigation and intensive cultivation made possible another flourishing of the city, with a peak in population in the eighteenth to seventeenth centuries bc that may be estimated at around 6,000–7,000 people. A new city fortification was built, running along the enlarged southern and eastern sides of the Lower City, which also included a portion of the oasis around the spring. The earliest (Middle Bronze IB) defensive structure consisted of a solid mud brick wall on stone foundations, replaced in the Middle Bronze II by a series of terrace walls supporting earthen ramparts, coated with clay and crushed limestone, regularizing the huge ruins of the collapsed monumental Early Bronze III double fortifications on the southern, western, and northern sides of the tell, the Upper City. Such ramparts, with their distinctive shiny white appearance, became a typical feature of the city. Recent Italian–Palestinian excavations have revealed a huge building in the southern Lower City, consisting of a rectangular Tower (Tower A1) with mud brick walls upon an orthostates foundation (Marchetti and Nigro, 1998: 124–35; 2000: 199–207; Nigro et al., 2011b: 187–91). The latter was the earliest building erected in the Lower City, directly upon Pre-Pottery Neolithic B layers, at the beginning of the nineteenth century bc (Middle Bronze IB). It was probably contemporary with the huge (and similar) Eastern Tower excavated by Garstang alongside the spring (Garstang, 1932: 15–17, pl. IX; Garstang and Garstang, 1948: 85–6, fig. 4), presumably devoted to its protection.2 After major destruction, possibly attributable to Pharaoh Amenemhat III’s campaign in Palestine towards the end of the nineteenth century bc, the city was further fortified through the erection of ramparts supported by stone walls, like the curvilinear stone structure brought to light by Italian–Palestinian excavations at the southwestern foot of the tell (Nigro et al., 2011b: 195–7). Earthen ramparts were strengthened by crushed limestone tongues and plastered with clayish marl, and their top was crowned by a mud brick wall. This impressive work needed enormous quantities of water, in its erection, but especially for marl and clay coating necessary in order to prevent heavy erosion. In the meantime, the fortification line had to incorporate channels to draw water outside the Lower City into the oasis. This made the eastern line of fortification a complex structure, with stone foundations incorporating such water streams. The Middle Bronze palace occupied the central and eastern part of the Spring Hill. It was excavated by J. Garstang (“Hyksos Palace”; Garstang,

023-051 WaterUrban Pt1 Ch01 Nigro.indd 37

12/05/2014 11:10


A History of Water

1934: 99–101, pl. XV [nn. 80, 81]; Garstang and Garstang, 1948: 99–101, fig. 4). Just below palace floors, a series of built-up tombs was uncovered by Kenyon and by the Italian–Palestinian Expedition (Nigro, 2009a). One of them (D.641) included the burial of a young lady bearing a scarab with the hieroglyphic inscription Adjmer Rwha—that is, the Egyptian title “administrator” (literarily “administrator of canals”—an appellation that goes back to the Old Kingdom) and the Canaanite word Ruha, possibly the ancient name of Jericho. Again, this title points to the strict relationship established at Jericho between the ruling institution and water control and administration. Not far away from the palace, on the northwestern hilltop of the southern mound at the center of the tell, stood the Middle Bronze temple, looking east. This was uncovered in the later years, although it was very badly preserved. Garstang excavated 23 Middle Bronze tombs in the necropolis (Tombs 1–5, 8–9, 12–15, 19–23, 30–2, 35, 40–2; Garstang, 1933: 4–38), collecting several Egyptian items (such as scarabs, alabaster, and faience objects). Kenyon also excavated 51 Middle Bronze tombs in the necropolis (some reused from Early Bronze IV period; Kenyon, 1960: 263–518; 1965: 167–478), subdividing them according to their pottery assemblages into five groups, ranging from Middle Bronze IB to Middle Bronze III (i.e., 1850–1550 bc). Large caves were used as familiar multiple burials; they are characterized by the extraordinary state of preservation of the finds, such as wooden trays, combs, tables, chests and bowls, ostrich eggs, leather and textiles, as well as bone inlays of wooden boxes with geometric and bird decorations. Valuable finds include an Old Babylonian cylinder seal, a bronze belt of a distinguished type with circular studs, and two equids from Kenyon’s warrior Tomb J3. A few Egyptian scarabs bear royal names and titles, including Pharaohs Hotepibra and Sobekhotep V or VI, suggesting a strong connection between Jericho and Egypt during the Thirteenth Dynasty. It is possible that Egyptians viewed Jericho as an attractive counterpart due to its oasis environment, which with its palm trees and waters echoed the Nile landscape. The Middle Bronze city was destroyed at least three times and a drastic reconfiguration occurred during the seventeenth century bc. At this time a new restricted area was encircled with a rubble rampart, supported by the huge Cyclopean Wall at its foot, and a series of terrace walls on the slope. The Cyclopean Wall was traced by the Austro–German Expedition, and extensively excavated by the Italian–Palestinian Expedition to the south. It was traced also in the oasis northeast of the spring (Figure 1.7). After the destruction at the end of Middle Bronze III, the city of Ruha was further reduced in area during the succeeding Late Bronze Age (1450– 1250 bc; Sultan V), with its focus upon the eastern flank of the Spring Hill, where the palace was replaced by a relatively small residency (called the

023-051 WaterUrban Pt1 Ch01 Nigro.indd 38

12/05/2014 11:10

The Tale of an Early City and Water Control in Ancient Palestine39

Figure 1.7.  ‘Ain es-Sultan area: boulders of the MB III (1650–1550 bc) Cyclopean Wall encircling the ancient city of Jericho at the end of the Middle Bronze Age, emerging in the oasis northeast of the ‘Ain es-Sultan.

“Middle Building”, and excavated by Garstang; Garstang, 1934: 105–17). Actually, the absence of Bichrome Ware and Cypriot imports suggested that Ruha was deserted between 1550 and 1450 bc (Kenyon, 1951). However, the scarcity of finds does not confirm absence, since the Middle Building seems to replace the previous palace directly. One of the most interesting finds from the Middle Building (just east of it on the slope) is a cuneiform tablet, attributable to the fourteenth century bc (Garstang, 1934: 116–17; Horowitz et al., 2006: 96). Pottery vessels found by Garstang in reused Tombs 4, 5, and 13 can be attributed to the same time lapse (Garstang, 1933: 14–38; Bienkowski, 1986: 32–102). Actually, Tomb 5 shows vessels as early as the second half of the fifteenth century and the beginning of the fourteenth century bc; Tombs 4 and 13, conversely, yielded vessels datable to 1375–1275 bc. Tomb 5 also gave up a scarab of Thutmosis III and one of Hatshepsut (a second scarab of Thutmosis III was found in Pit Tomb 11 dating from Iron I, see below); while Tomb 4 yielded two scarabs with the cartouche of Amenophis III. No evidence of a fortification system for this period existed and Garstang’s Wall of “City D” was incorrectly attributed to Late Bronze. This is, of course, fairly normal, since the majority of sites in Palestine are devoid of a new fortification system in this period, having been subject to the Egyptian Eighteenth Dynasty’s control.

023-051 WaterUrban Pt1 Ch01 Nigro.indd 39

12/05/2014 11:10


A History of Water

Biblical accounts: Jericho and the spring in the first millennium bc (1150–586 bc) Waters from the spring of ‘Ain es-Sultan are also the protagonists of Biblical tales concerning Jericho, which was viewed in the Bible as a paradise, and a symbol of civilization and richness. According to the Old Testament, Jericho was one of the main centers of the tribe of Benjamin, marking the border with the tribe of Ephraim. Seven well-known Biblical accounts (Old Testament and New Testament) are centered on Jericho: 1. The heavily fortified city was conquered by Joshua to the sound of rams’ horns (Joshua 6), with the sudden collapse of its walls and a curse over its reconstruction. 2. Eglon, the king of Moab, conquered Jericho at the time of Judge Ehud (Judges 3:13), who eventually killed him. 3. At the time of David, his envoys, who had been shaved off half their beards, and had their garments cut off in the middle, at their buttocks, by Hanun king of Ammon, were told to wait at Jericho until their beards had grown (2 Samuel 10:5). 4. Hiel, king of Bethel, rebuilt the city at the time of Achab, burying his first-born son (Abiram) under the city walls and his youngest one (Segub) under the city gates (1 Kings 16:34). 5. After Elijah’s ascent to heaven (by the Jordan), prophet Elisha healed the waters of the spring—which had become bitter and caused disease and death—by throwing a pot of salt in (2 Kings 2:19–22). The latter tale also gave the name of “Elisha’s Spring” to ‘Ain es-Sultan. 6. In the New Testament, Jericho is remembered due to the episode of Zacchaeus, the chief tax collector at Jericho. He climbed a sycamore tree to see Jesus, and then gave half his possessions to the poor (Luke 19:1–10). 7. The last episode is that of the blind man healed by Jesus due to his faith (Mark 10:46–52; Matthew 20:29–34). Biblical narratives thus reflect the centrality of the site not only in the history of the region, but also in its cultural imagery, especially in relation to its waters and to the image of a luxuriant, cultivated land. This image was fixed in the Hasmonean Period, when the oasis became the seat of the kings’ winter palaces and of a number of villas belonging to the Jerosolimitane aristocracy. Jericho, the “City of Palms” (Judges 2:13; 2 Chronicles 28:13) was at that time intensively cultivated for the production of ointments and perfumes. Different is the tale told us by archaeology, which has reconstructed the following history of the city and the annexed spring. In the final stage of the Late Bronze period, when Joshua’s attack is traditionally dated, Tell es-Sultan was unoccupied and only a few remains, datable to the early Iron

023-051 WaterUrban Pt1 Ch01 Nigro.indd 40

12/05/2014 11:10

The Tale of an Early City and Water Control in Ancient Palestine41

Age, could be identified. Garstang’s “Cremation Pit” (Tomb 11; Garstang, 1933: 36, fig. 11) can be dated to Iron I, a period also illustrated by pottery fragments retrieved by the Austro–German Expedition and studied by Helga and Manfred Weippert (Weippert and Weippert, 1976: 105–48). They were discovered in the foundation embankment of a monumental building erected on the eastern slope of the Spring Hill, the so-called “Hilani” (Sellin and Watzinger, 1913: 67–70, fig. 42, pls. 15–16, I, IV; Garstang, 1934: 102–4, pl. XIII), dated to Iron IIA (tenth to ninth centuries bc) by the same scholars (the period in which the two Biblical episodes of David’s envoys (2 Samuel 10:5) and the episode of Hiel king of Bethel (1 King 16:34) would be situated). A possible occupation of the site in the ninth century is also suggested by Kenyon’s Tomb A85 (Tushingham, 1965: 482–9). A continuous occupation from the eighth century to the sixth century, both on the summit of the tell and on its northern and southern slopes, was documented by all the expeditions. It was extensively excavated only by Sellin and Watzinger in the central and northern areas of the tell, where private houses and slab-paved staircases climbing the (by that time) steep mound (similar to those of Tell es-Sa’idiyeh) were brought to light. A tripartite building was also uncovered by Kenyon at the foot of Trench I to the west (Kenyon, 1981: 111–12, pls. 94, 232). A double-winged royal stamp on a jar handle (Bartlett, 1982: 537, fig. 220:1) may indicate that Jericho was included in the administration of the Kingdom of Judah in the seventh to sixth centuries bc, even though its inclusion in the Kingdom of Ammon cannot be excluded. After the Babylonian conquest of Jerusalem in 587 bc and its destruction, Jericho fell under the Neo-Babylonian and then Persian administration (586–333 bc). The site, though on a reduced scale, continued to be occupied (a barbed arrow head found in Trench I is a tangible relic of Persian warriors active in this area), even though this period experienced a decline in settlement and cultivation in the whole oasis (only one site, a ritual bath in Wadi en-Nueima, was attributed to Persian occupation; Dinur and Feig, 1986). An ostracon with an Aramaic inscription and some stamp seal impressions (Bartlett, 1982) can also be attributed to this period. Foreign powers were no longer interested in the development of the extraordinary agricultural potential of the oasis (Nigro et al., 2011a: 16–17). Triumph of waters: Jericho in the Hellenistic (Hasmonean) and Roman Periods (333 bc–ad 324) The situation reversed in the Hellenistic Period (333–64 bc), when the Jericho oasis witnessed an extraordinary flourishing, with the systematic exploitation of the springs of ‘Ain Dyuk and ‘Ain el-Auja, and the aqueduct of Wadi el-Qelt, which gave rise to the building of numerous villas and

023-051 WaterUrban Pt1 Ch01 Nigro.indd 41

12/05/2014 11:10


A History of Water

to an enhancement of productive capacities (palm date trees, wine, and opobalsamum (Balm of Mecca) for the perfume industry). The banks of Wadi Qelt, at the southern border of the oasis, were chosen as the site of a palatial complex by the Hasmonean rulers (Nigro et al., 2011a: 17–18; Figures 1.8 and 1.9). Before the construction of the palace complex, a number of fortresses controlled access to the oasis (Tell el-‘Aqaba/Cypros, Nuseib ‘Uweishira and Jebel Quruntul). The huge palatial complex at Tell Abu el-‘Alayiq North extended over an area of c.30 dunams (Netzer, 1996; 2001: 1–7, 11–174, 334–8). It included the First Hasmonean Palace, the so-called “Buried Palace”, and the “Pool Complex”, with its ritual baths, built by Johannes Hyrcanus (134–104 bc) and later enlarged by Alexander Janneus (103–76 bc), including the “Fortified Palace”, the “Twin Palaces”, as well as an industrial area with ritual baths, and a synagogue or triclinium. The palace complex became the winter residency of the Hasmonean rulers, and it was progressively enriched until the Roman Period, when it was rebuilt by King Herod after its destruction by an earthquake in 31 bc (Herod’s Second, and then Third Palace; Netzer, 2001: 7–10, 229–98, 317–24, 326–30, 339–41; fig. 10). In this period, the whole oasis flourished and became a garden with flowers, trees, and abundant water running in aqueducts and canals (Netzer and Garbrecht, 2002), with many rural

Figure 1.8.  Tulul Abu el-‘Alayiq: isometric view of Herod’s Third Palace, from the southwest (after Netzer, 2001: ill. 460).

023-051 WaterUrban Pt1 Ch01 Nigro.indd 42

12/05/2014 11:10

The Tale of an Early City and Water Control in Ancient Palestine43

Figure 1.9.  Tulul Abu el-‘Alayiq: plan of the area south and north of Wadi Qelt, including the Royal Estate, with the Hasmonean and Herodian winter palaces, and the water channels and installations (after Netzer and Garbrecht, 2002: fig. 6).

023-051 WaterUrban Pt1 Ch01 Nigro.indd 43

12/05/2014 11:10


A History of Water

villages spread in the countryside, such as Tell Abu Hindi, Khirbet en-Nitla/ Tell Jaljul, and Tell es-Samarat, as well as other marginal settlements (Qasr el-Yehud, to the east, and Suwwanet eth-Thaniya and Wadi en-Nueima, to the north). As regards Tell es-Sultan, it hosted a number of small installations (an inscribed handle of a Rhodian amphora dates from 220 to 150 bc; Bartlett, 1982: 542, fig. 220:6), while important hydraulic structures were erected by the spring of ‘Ain es-Sultan. In March 2012, the Ariha Municipality carried out rehabilitation work in the area of the spring of ‘Ain es-Sultan, involving the two major buildings of the hydraulic installations: the Ottoman Pool and the Old Mill, connected with the former oval building through an aqueduct refurbished several times in history (Dorrel, 1993: 111; Taha, 2009). Work consisted of the removal of dump and residual material from building activities related to these installations. The filling yielded a mass of archaeological material, including pottery and stone implements dating from the Neolithic onwards. The oval wall of the Ottoman Pool was excavated down to its earliest layers, exposing a stratigraphy extending to the Roman Period, to which a magnificent tunnel 2 m wide with ashlar corbelled vault belonged. This structure lies beneath the Byzantine and Mameluk aqueduct, and came to light 15 m to the east alongside the mill and its water devices. Pottery from layers under the oval wall of the Ottoman Pool included Early Bronze Age fragments dating back to Early Bronze II–III, thus suggesting the longevity of the spring after its possible shifting occurred between Pre-Pottery Neolithic and Pottery Neolithic. Actually, sparse architectural remains from the Roman Period (64 bc–ad 324) were excavated on the tell, or retrieved in secondary contexts (Zagari, 2000: 357–9). They belong to rural installations and include a wine press with an associated signinum plastered bin. A Roman Corinthian capital was found on the tell, possibly re-employed in a Byzantine church, as well as other architectural fragments. The necropolis west of the tell was also used in Roman times. Seven tombs and 14 graves were excavated

Figure 1.10.  Tulul Abu el-‘Alayiq: general view of the Hasmonean Palaces Complex (left) and of the northern wing of Herod’s Third Palace (right), from the south.

023-051 WaterUrban Pt1 Ch01 Nigro.indd 44

12/05/2014 11:10

The Tale of an Early City and Water Control in Ancient Palestine45

by Kenyon (Bennett, 1965: 516–45). A carved capstone of a monumental tomb was recovered by the Italian–Palestinian Expedition by the site. During the Roman Period, the main site in the oasis became Tell Abu Hindi, together with Tulul Abu el-‘Alayiq, where Herod’s palaces stood (Nigro et al., 2011a: 19–20). Byzantine, Islamic, and Ottoman Periods In the Byzantine Period (ad 324–636), Jericho appears on the Madaba Map, with the name and a basilica depicted in the mosaic. Already in the fourth century ad, at the dawn of the Byzantine era, Jericho was a place of pilgrimage and prayer, thanks to its religious significance; churches and monasteries flourished in the oasis, in the ‘Ain Hajla area, and on the surrounding hills (as Tell Abu Hindi, Khirbet en-Nitla/Tell Jaljul, Qasr el-Yehud, Early Bronze Jebel Quruntul, and Rujm el-Mugheifir North; Nigro et al., 2011a: 21–3). The main center of the oasis was the site of Tell el-Hassan, where present-day Jericho is. The Jericho oasis thus became a rich and variegated Byzantine enclave, where several Christian communities of monks lived (Orthodox, Catholic, Coptic, etc.), with monasteries and churches epitomizing Byzantine art in mosaics, frescoes, and stucco decoration (Donceel-Voûte, 1999). On Tell es-Sultan (on the northeastern peak of the Spring Hill), a rural village arose, also producing dust pits that cut deeply into the Iron and Bronze Ages strata (Zagari, 2000: 357–65). In the nearby area of the spring of ‘Ain es-Sultan, a basilica was erected, of which a capital and some sparse remains are preserved today.3 The Byzantine Basilica is the first documented cult place directly connected with the spring of ‘Ain es-Sultan (Cirelli and Zagari, 2000: 366–7). ‘Ain es-Sultan retained its fundamental function also in the Islamic Period (ad 636–1516) through to the Ottoman Period (ad 1516–1918). However, after the passing of the Persian army in 614 and the arrival of the Arabs in 638, the number of inhabited sites at the Jericho oasis remained severely restricted (Nigro et al., 2011a: 23–5). Remains of Islamic occupation of Tell es-Sultan are represented by ceramic fragments dating from the Omayyad Period (Abbassid and Mamluk pottery was also found; Zagari, 2000: 365–6). At the northern edge of the oasis, on the northern banks of Wadi en-Nueima, the magnificent residency of Caliph Hisham was built in the first half of the eighth century bc, including baths paved with extraordinary mosaics, and wooden and stucco-figured decorations nowadays in the Rockefeller Museum (ex-Palestine Archaeological Museum), in Jerusalem (Hamilton, 1959). The Jericho oasis thus preserved its vocation as a seat of winter palaces, in this case for the caliphs of Damascus. During the Crusader Period, some of the main installations near to Tell es-Sultan were the sugar mills of Tawaheenes-Sukkar, fed by the spring of ‘Ain el-Auja (Taha, 2009). Already during the middle and late Islamic

023-051 WaterUrban Pt1 Ch01 Nigro.indd 45

12/05/2014 11:10


A History of Water

Periods, however, the oasis and the site of Tell es-Sultan were gradually abandoned, although the spring of ‘Ain es-Sultan was preserved up until the Ottoman Period (ad 1516–1918), when a monumental pool (70 m by 15 m) was created and is still in use (although with an inadequate modern roofing) today. Jericho: the city of the spring This very brief overview of the history of Tell es-Sultan/ancient Jericho points to the central role played by the spring of ‘Ain es-Sultan, which was, as early as the Pre-Pottery Neolithic Period, the focal point of the village, and remained so for the city it had become in the Bronze Age. Fresh waters from the spring provided not only a basic resource for subsistence, but also a means to develop the economy and social organization faster, influencing land control and use, as well as animal domestication and breeding. The rise of agriculture, the key Neolithic process at Jericho (as elsewhere in the Fertile Crescent), was due here to the spring’s abundance and to the ability of the local community, whose knowledge allowed them to store and control the water. The manufacture of bricks (i.e., of modular architecture) and of pottery also follows from the abundant water supply offered by ‘Ain es-Sultan. Both achievements represent basic steps in the history of humankind. Archaeological excavations have made it possible to follow thoroughly the rise of the city in the Bronze Age, starting from the settlement of a new groups alongside the spring at the end of the fourth millennium bc, gradually following the growth of a small hamlet, which, wisely exploiting the waters of the spring, became a large rural village and then, successively, a fortified town and a prominent city in the third and second millennia bc. This development reached its peak in the Middle Bronze Age, when the rulers of Jericho bore the Egyptian title of adjmer (i.e., administrator of canals)—further proof of the social and economic role played by water control in the city of the oasis. The Lords of Jericho exercised their power over the spring and thus extended their control over the whole oasis, its surroundings, and the important routes crossing and following the southern Jordan Valley down to the Dead Sea. This established Jericho at a pivotal spot of the ancient road network on the long-distance trade routes, which the Egyptian connection also testifies. In Hellenistic and Roman times, the perfume industry became one of the main activities of the oasis (again, thanks to the spring and an exceptional system of irrigation), and it became the site for villas and the royal palaces of the Hasmoneans. It was ‘Ain es-Sultan and the other water sources of Wadi el-Kelt and Wadi en-Nueimal that made such flourishing development possible. This ancient city thus owes its historical parable to

023-051 WaterUrban Pt1 Ch01 Nigro.indd 46

12/05/2014 11:10

The Tale of an Early City and Water Control in Ancient Palestine47

the spring it soon incorporated, and which, as the site of several biblical tales, has enjoyed everlasting fame. Notes 1 The presence of earthquakes fracturing and affecting Pre-Pottery Neolithic B layers and structures, registered by both Garstang’s and Kenyon’s Expeditions, points to repeated seismic events affecting Neolithic Jericho. In particular, a deep crack detected by J. Garstang in his North-Eastern Trench, crossing the floors of buildings at level X (the uppermost PPNB layer; Garstang et al., 1936: pl. XL, a; Garstang and Garstang, 1948: pls. VI [right end of the section, loci 453/297], IX), may appear to offer probative evidence that an earthquake occurred at the end of Pre-Pottery Neolithic. 2 At Jerusalem in the same period, the Gihon Spring on the eastern flank of the hill, where the Middle Bronze City arose, was protected with the erection of a couple of huge towers (Reich and Shukron, 2004, 2008). 3 Also two synagogues arose in the oasis: the Synagogue of Shahwan, in the Tell el-Jurn area, and the Synagogue of Khirbet Na‘aran.

references Bartlett, J. R. (1982). “Appendix A. Iron Age and Hellenistic stamped jar handles from Tell es-Sultan”, in K. M. Kenyon and T. A. Holland, Excavations at Jericho. Volume Four. The Pottery Type Series and Others Finds, pp. 537–45, London: The British School of Archaeology in Jerusalem. Bennett, C. M. (1965). “Tombs of the Roman Period”, in K. M. Kenyon, Excavations at Jericho. Volume Two. The Tombs Excavated in 1955–1958, pp. 516–45, London: The British School of Archaeology in Jerusalem. Bienkowski, P. (1986). Jericho in the Late Bronze Age, Warminster: Aris & Phillips. Cirelli, E. and F. Zagari (2000). “L’oasi di Gerico in età bizantina ed islamica. Problemi e proposte di ricerca”, Archeologia Medievale, 27, pp. 365–76. Dinur, U. and N. Feig (1986). “WadiNu’eima”, Excavations and Surveys in Israel, 5, pp. 110–11. Donceel-Voûte, P. (1999). “Jericho aux époques Byzantine et Omeyyade”, Dossier d’Archéologie, 240, pp. 113–21. Dorrel, P. (1993). “The Spring of Jericho from early photographs”, Palestine Exploration Quarterly, 125, pp. 95–114. Garstang, J. (1932). “Jericho: city and necropolis”, Liverpool Annals of Archaeology and Anthropology, 19, pp. 3–22, 35–54. ——— (1933). “Jericho: city and necropolis. 4. Tombs of MBAii. 5. Tombs of MBAii and LBAi. 6. The Palace Area”, Liverpool Annals of Archaeology and Anthropology, 20, pp. 3–42. ——— (1934). “Jericho: city and necropolis. 6. The Palace Area (cont.). Palace and Store Rooms, MBii Pottery and Houses LBi. Upper Stone Building, EBi”, Liverpool Annals of Archaeology and Anthropology, 21, pp. 99–136. ——— (1935). “L’art néolithique à Jéricho”, Syria, 16, pp. 353–7.

023-051 WaterUrban Pt1 Ch01 Nigro.indd 47

12/05/2014 11:10


A History of Water

Garstang, J., J. Crowfoot, and J. P. Droop (1935). “Jericho: city and necropolis (fifth report)”, Liverpool Annals of Archaeology and Anthropology, 22, pp. 143–84. Garstang, J., I. Ben-Dor, and F. M. Fitzgerald (1936). “Jericho: city and necropolis (report for the sixth and concluding season, 1936)”, Liverpool Annals of Archaeology and Anthropology, 23, pp. 67–100. Garstang, J. and J. B. E. Garstang (1948). The Story of Jericho, London: Marshall, Morgan & Scott. Hamilton, R. W. (1959). Khirbet al-Majfar, an Arabian Mansion in the Jordan Valley, Oxford: Oxford University Press. Helms, S. W. (1988). “Jericho which was once a little city, but is now destroyed, and is but a little village”, in S. Shaath (ed.), Studies in the History and Archaeology of Palestine (Proceedings of the First International Symposium on Palestine Antiquities), Vol. III, pp. 83–102, Aleppo: Alecso. Horowitz, W., T. Oshima, and S. L. Sanders (2006). Cuneiform in Canaan. Cuneiform Sources from the Land of Israel in Ancient Times, Jerusalem: Israel Exploration Society and the Hebrew University. Kenyon, K. M. (1951). “Some notes on the history of Jericho in the second millennium B.C.”, Palestine Exploration Quarterly, 83, pp. 101–38. ——— (1957). Digging Up Jericho, London: Ernest Benn Limited. ——— (1960). Excavations at Jericho. Volume One. The Tombs Excavated in 1952–1954, London: The British School of Archaeology in Jerusalem. ——— (1965). Excavations at Jericho. Volume Two. The Tombs Excavated in 1955–1958, London: The British School of Archaeology in Jerusalem. ——— (1981). Excavations at Jericho. Volume Three. The Architecture and Stratigraphy of the Tell, London: The British School of Archaeology in Jerusalem. Marchand, F. (2011–12). “The modelling skulls in the Ancient Near-East”, Tiempo y sociedad, 6, pp. 5–41. Marchetti, N. and L. Nigro (1998). Scavi a Gerico, 1997. Relazione preliminare sulla prima campagna di scavi e prospezioni archeologiche a Tell es-Sultan, Palestina (Quaderni di Gerico 1), Rome: Università degli Studi di Roma “La Sapienza”. ——— (2000). Excavations at Jericho, 1998. Preliminary Report on the Second Season of Excavations and Surveys at Tell es-Sultan, Palestine (Quaderni di Gerico 2). Rome: Università degli Studi di Roma “La Sapienza”. Milevski, I., H. Khalaily, N. Getzov, and I. Hershkovitz (2008). “The plastered skulls and other PRE-POTTERY NEOLITHIC B finds from Yiftahel, Lower Galilee (Israel)”, Paléorient, 34(2), pp. 37–6. Netzer, E. (1996). “The Hasmonean Palaces in Palestine”, in H. Hoepfuer and G. Brands (eds), Die Paläste der Hellenistischer Könige. Internationale Symposion, Berlin von 16. 12. 1992 bis 20. 12. 1992, pp. 203–8, Mainz amRhein: P. von Zabern. ——— (2001). Hasmonean and Herodian Palaces at Jericho: Final Reports of the 1973–1987 Excavations. Volume I: Stratigraphy and Architecture, Jerusalem: Israel Exploration Society. Netzer, E. and G. Garbrecht (2002). “Water channels and a royal estate of the Late Hellenistic Period in Jericho’s western plains”, in D. Amit, J. Patrich, and Y.

023-051 WaterUrban Pt1 Ch01 Nigro.indd 48

12/05/2014 11:10

The Tale of an Early City and Water Control in Ancient Palestine49

Hirschfeld (eds), The Aqueducts of Israel (Journal of Roman Archaeology Supplementum 46), pp. 367–77, Portsmouth: Journal of Roman Archaeology. Nigro, L. (2003). “Tell es-Sultan in the Early Bronze Age IV (2300–2000 bc). Settlement vs necropolis—a stratigraphic periodization”, Contributi e Materiali di Archeologia Orientale, IX, pp. 121–58. ——— (2005). Tell es-Sultan/Gerico alle soglie della prima urbanizzazione: il villaggio e la necropoli del Bronzo Antico I (3300–3000 a.C.) (Rome “La Sapienza” Studies on the Archaeology of Palestine & Transjordan, 1), Rome: “La Sapienza” Expedition to Palestine and Jordan. ——— (2006a). “Sulle mura di Gerico. Le fortificazioni di Tell es-Sultan come indicatori della nascita e dello sviluppo della prima città di Gerico nel III millennio a.C.”, in F. Baffi, R. Dolce, S. Mazzoniand, and F. Pinnock (eds), Ina Kibra ¯ tErbetti. Studi di Archeologi orientale dedicati a Paolo Matthiae, pp. 349–97, Rome: Università degli Studi di Roma “La Sapienza”. ——— (2006b). “Results of the Italian–Palestinian Expedition to Tell es-Sultan: at the dawn of urbanization in Palestine”, in L. Nigro and H. Taha (eds), Tell es-Sultan/Jericho in the Context of the Jordan Valley: Site Management, Conservation and Sustainable Development (Rome “La Sapienza” Studies on the Archaeology of Palestine & Transjordan, 2), pp. 1–40, Rome: “La Sapienza” Expedition to Palestine and Jordan. ——— (2007). “Aside the spring: Byblos and Jericho from village to town in the second half of the 4th millennium bc”, in L. Nigro (ed.), Byblos and Jericho in the Early Bronze I: Social Dynamics and Cultural Interactions. Proceedings of the International Workshop held in Rome on March 7th 2006 by Rome “La Sapienza” University (Rome “La Sapienza” Studies on the Archaeology of Palestine & Transjordan, 4), pp. 1–45, Rome: “La Sapienza” Expedition to Palestine and Jordan. ——— (2009a). “The built tombs on the Spring Hill and the Palace of the Lords of Jericho (‘dmrRha) in the Middle Bronze Age”, in J. D. Schloen (ed.), Exploring the longue durée. Essays in Honor of Lawrence E. Stager, pp. 361–76, Winona Lake, IN: Eisenbrauns. ——— (2009b). “Khirbet Kerak Ware at Jericho and the EARLY BRONZE III change in Palestine”, in E. Kaptijn and L. P. Petit (eds), A Timeless Vale. Archaeological and Related Essays on the Jordan Valley in Honour of Gerrit van der Kooij on the Occasion of his Sixty-fifth Birthday (Archaeological Studies Leiden University 19), pp. 69–83, Leiden: Leiden University Press. ——— (2010). Tell es-Sultan/Jericho in the Early Bronze II (3000–2700 bc): The Rise of an Early Palestinian City. A Synthesis of the Results of Four Archaeological Expeditions (Rome “La Sapienza” Studies on the Archaeology of Palestine & Transjordan, 5), Rome: “La Sapienza” Expedition to Palestine and Jordan. Nigro, L. and H. Taha (2009). “Renewed excavations and restorations at Tell es-Sultan/Ancient Jericho. Fifth season – March–April 2009”, Scienzedell ’Antichità, 15, pp. 731–44. Nigro, L., M. Sala, and H. Taha (2011a). Archaeological Heritage in the Jericho Oasis. A Systematic Catalogue of Archaeological Sites for the Sake of Their Protection and Cultural Valorization (Rome “La Sapienza” Studies on the Archaeology of Palestine & Transjordan, 7), Rome: “La Sapienza” Expedition to Palestine and Jordan.

023-051 WaterUrban Pt1 Ch01 Nigro.indd 49

12/05/2014 11:10


A History of Water

Nigro, L., M. Sala, H. Taha, and J. Yassine (2011b). “The Early Bronze Age palace and fortifications at Tell es-Sultan/Jericho. The 6th–7th seasons (2010– 2011) by Rome “La Sapienza” University and the Palestinian MOTA-DACH”, Scienzedell ’Antichità, 17, pp. 185–211. Reich, R. and E. Shukron (2004). “The history of the Gihon Spring in Jerusalem”, Levant, 36, pp. 211–23. ——— (2008). “Jerusalem. The Gihon Spring and eastern slope of the City of David”, in E. Stern (ed.), The New Encyclopedia of Archaeological Excavations in the Holy Land, Vol. V (Supplementary Volume), pp. 1801–7, Jerusalem: The Israel Exploration Society & Carta. Rollefson, G. O. (2000). “The statuary from ‘Ain Ghazal, Jordan”, in A. Weyer (ed.), Saving Cultural Heritage. Projects Around the Mediterranean (Hornemann Institute Series 3), pp. 101–10, Hamburg: Glöss Verlag. Sala, M. (2005). “Il Sacello 420 (‘BabylonianShrine’): luogo di culto di un quartiere abitativo. Struttura e funzioni”, in L. Nigro, Tell es-Sultan/Gerico alle soglie della prima urbanizzazione: il vilaggio e la necropolis del Bronzo Antco I (3300–3000 a.c.) (Rome “La Sapienza” Studies on the Archaeology of Palestine & Transjordan, 1), pp. 42–7, Rome: “La Sapienza” Expedition to Palestine and Jordan. ——— (2006). “Garstang’s north-eastern trench: archaeological evidences and Potential”, in L. Nigro and H. Taha (eds), Tell es-Sultan/Jericho in the Context of the Jordan Valley: Site Management, Conservation and Sustainable Development (Rome “La Sapienza” Studies on the Archaeology of Palestine & Transjordan, 2), pp. 267–87, Rome: “La Sapienza” Expedition to Palestine and Jordan. ——— (2008). “Khirbet Kerak Ware from Tell es-Sultan/ancient Jericho: a reassessment in the light of the finds of the Italian–Palestinian Expedition (1997– 2000)”, in J. M. Córdoba, M. Molist, M. C. Pérez, I. Rubio, and S. Martínez (eds), Proceedings of the 5th International Congress on the Archaeology of the Ancient Near East, Madrid, April 3–8 2006, Vol. III, pp. 111–33, Madrid: Publications Service of the Universidad Autónoma of Madrid. Sellin, E. and C. Watzinger (1913). Jericho. Die Ergebnisse der Ausgrabungen (Wissenschaftliche Veröffentlichung der Deutschen Orient-Gesellschaft 22), Leipzig: Hinrichs. Taha, H. (2009). “Some aspects of sugar production in Jericho, Jordan Valley”, in E. Kaptijn and L. P. Petit (eds), A Timeless Vale. Archaeological and Related Essays on the Jordan Valley in Honour of Gerrit van der Kooij on the Occasion of his Sixty-fifth Birthday (Archaeological Studies Leiden University 19), pp. 181–91, Leiden: Leiden University Press. Taha, H. and A. Qleibo (2010). Jericho, A Living History, Ten Thousand Years of Civilization, Ramallah: The Palestinian Ministry of Tourism and Antiquities. Taha, H., N. Anfinset and J. Yassine (2004). “Preliminary report on the first season of the Palestinian-Norwegian excavation at Tell el-Mafjer, Jericho”, Orient Express, 2004(2), pp. 40–4. Tushingham, A. D. (1965). “Tombs of the Early Iron Age”, in K. M. Kenyon, Excavations at Jericho. Volume Two. The Tombs Excavated in 1955–1958, pp. 479–515, London: The British School of Archaeology in Jerusalem. Weippert, H. and M. Weippert (1976). “Jericho in der Eisenzeit”, Zeitschrift des Deutschen Palästina-Vereins, 92, pp. 105–48.

023-051 WaterUrban Pt1 Ch01 Nigro.indd 50

12/05/2014 11:10

The Tale of an Early City and Water Control in Ancient Palestine51

Zagari, F. (2000). “Appendix H: remarks on the Byzantine occupation of Tell es-Sultan”, in N. Marchetti and L. Nigro, Excavations at Jericho, 1998. Preliminary Report on the Second Season of Archaeological Excavations and Surveys at Tell es-Sultan, Palestine (Quaderni di Gerico 2), pp. 355–81, Rome: Università degli Studi di Roma “La Sapienza”.

023-051 WaterUrban Pt1 Ch01 Nigro.indd 51

12/05/2014 11:10


Mohenjo-Daro, Indus Valley Civilization: Water Supply and Water Use in One of the Largest Bronze Age Cities of the Third Millennium bc1

Michael Jansen Introduction The beginning of the third millennium bc saw the first urban civilizations of the world formed, in the large river systems of the Euphrates–Tigris in Mesopotamia, the Nile in Egypt, and the Indus Valley in the northwestern Indian subcontinent. Mohenjo-Daro (Sindhi: “hill of the dead”), the probable capital of the Indus civilization, is one of the largest cities of the third millennium Bronze Age. What appear today as hills in the alluvial plain of the Indus, approximately 1 million m2 in area, seems to be only the “tip of an iceberg”. The major part of it has been covered by the alluvial silts of the river over the past 4,500 years. Out of the 1 million m2 of visible surface, 100,000 m2 have been excavated so far, most of it between 1922 and 1936 by the Archaeological Survey of India (ASI) of the British Raj. Discoveries by Mortimer Wheeler (Wheeler, 1968) in 1950, through deep soundings close to the western side of the “citadel”, gave clear indications that the ancient surroundings of the city were much lower than those of today. Over the past 4,500 years, the Indus seems to have silted up its valley—at least, close to Mohenjo-Daro—by more than 7 m. Recent drillings and further diggings have proved that a larger part of the city today is covered by the alluvium of the Indus. It is presently estimated that the major part of the city of another 3 million m2 is covered by the alluvium of the past 4,500 years. In the 1920s, excavators were taken by surprise by the seeming homogeneity of the material culture and the architectural structures themselves, all being built in both burnt and unburnt bricks of today’s size and standards: approximately 6 × 12 × 24 cm (i.e., proportions 1:2:4). An exception was found in round cylindrical structures sunk vertically into the ground with diameters ranging from 60 cm to 2.4 m. Their circular “tubes” had once been formed by conically shaped bricks, allowing the construction of a perfectly built cylindrical form. Deep diggings inside the “tubes”

052-070 WaterUrban Pt1 Ch02 Jansen.indd 52

16/05/2014 15:21

Mohenjo-Daro, Indus Valley Civilization53

Figure 2.1.  Site plan of Mohenjo-Daro showing the excavated areas by name and visiting tours. Source: © M. Jansen, Aachen Conservation Center.

by the author showed that these went down more than 18 m below the present surface. They were perfectly constructed wells for the vertical supply of water inside the city itself. It is estimated that more than 700 wells had existed at one time in the city of Mohenjo-Daro—the highest density ever found for a city worldwide, and the earliest one. In attempting to understand how these wells could have been constructed in a densely built urban context, it was concluded that either they must have been constructed bottom up, with the earliest strata of the settlement, or they were later sunk “top down” (German: abteufen), a technique that is still in use here and there in the villages of Sindh. In the first case, the planning of the water supply must have coincided with the earliest planning of a city layout. In the second case, the digging of

052-070 WaterUrban Pt1 Ch02 Jansen.indd 53

16/05/2014 15:21


A History of Water

Figure 2.2.  The citadel mound with the so-called stupa from south, 1981. Source: © M. Jansen.

the vertical shafts of the wells through many layers of occupation deposits must have been a crucial work. Water supply (wells) As mentioned above, for the construction of the round well shafts, a special form of brick was developed. Corresponding to the smaller inside circumference of the shaft, these bricks were wedge-shaped. If such a vertically built cylinder made of wedge-shaped bricks were split down the middle and laid horizontally, the result would have been the first barrel vault, yet there is no evidence to show that the Harappans took this technological step forward. As far as we know, the corbelled vault was the only method they used to bridge larger spans using bricks, as attested by the roofing construction of drains and doorways. Nevertheless, the cylindrical wells still stand as witnesses to the high standard of hydro-engineering attained by the Harappans, as their circular form is statically ideal for withstanding the lateral pressure exerted on the shafts, which were sunk to a depth of at least 20 m. The cylindrical, brick-lined wells were probably invented by the Indus peoples of the alluvial plain. No wells have been found to date in any of the pre-Harappan or early Harappan settlement sites. Even in Mesopotamia

052-070 WaterUrban Pt1 Ch02 Jansen.indd 54

16/05/2014 15:21

Mohenjo-Daro, Indus Valley Civilization55

and Egypt, vertical water supply systems were virtually unknown within the settlements. It is all the more astonishing, therefore, that the radius of the mean catchment area supplied by each well in Mohenjo-Daro was calculated to be a mere 17 m. In view of the tiny size of the rooms housing the wells, and the extraordinary shaft depth of more than 20 m, the only technically feasible manner of construction would be the “shaft sinking” method. This method has been employed up until recent times, and involves laying the first courses of bricks on a stout wooden ring set into the ground. The compressed earth is then removed from under the ring and the core forming inside it, while simultaneously the shaft is extended upwards. The result is that the growing brick cylinder gradually sinks down into the earth under its own weight, until water-bearing soil is reached. Whether this was really the construction method used by the Harappan well-builders awaits proof from further excavation in the near future. An estimated 700 draw wells with an average catchment area radius of 17 m represented a frequency of wells; this is unparalleled in the history of water supply systems. The actual innovation in Mohenjo-Daro was the provision of a network of water supply points within the built-up urban area, from which water could be fetched conveniently as required, and then brought to the place of consumption (Jansen, 1991a: 17). Wells on the “citadel”2 To date, only six wells have been located in the entire “citadel” sector, four in the SD Area and two in the L Area. With an average catchment area of 3.175 m2 per well, and an average distance of 56.3 m between wells, the SD Area figures are almost twice as high as the equivalent values for the Lower City. Even the average diameter of the wells here is almost twice that estimated for the Lower City. As a rule, the wells were accessible to the general public. As far as location is concerned, convenience was not necessarily a deciding factor. Thus Block 1 in the SD Area, a complex building known as the “priests’ college”, does not have a single well, although the existence of several bathing platforms implies intensive use of water. The corresponding values for the L Area regarding well density (every 56.5 m) and distribution pattern are similar. Generally speaking, the “citadel” area was, in its prime, the site of a number of imposing structural complexes. The “Great Bath” is undoubtedly the most outstanding of these today. Compared with the Lower City, water supply points in the form of wells are less frequent on average, but larger in diameter. These and other indications weigh in favor of a public function for the wells in the “citadel” area. On the other hand, bathing platforms and effluent disposal drains are found just as densely distributed as in the Lower City.

052-070 WaterUrban Pt1 Ch02 Jansen.indd 55

16/05/2014 15:21


A History of Water

In sum, these observations are interpreted as pointing towards the public nature of the functions performed on the “citadel”, possibly in connection with veneration rites, as posited by Marshall (1931) and Mackay (1938). It only remains to say that there is nothing in the archaeological record to date to justify Wheeler’s use of the term “citadel”, with its military connotations, for this section of the city. Table 2.1.  List of wells on “citadel” Location SD Area 1. B1.6, R1725 2. B1.1, R1626 3. B1.4, R1327 4. B1.728 L Area4 1. Bl.C, R8530 2. Bl.D, R8031


Size (m)

Level (amsl) (m above msl)

Access point

Oval Round Round Oval3

1.85 × 1.30 2.07 1.77 1.71 × 1.62

54.86 51.24 51.03 51.70

From Main Street Passage 12 Street? ?

Round Round

2.29 1.78

53.16 54.47

Street? Street?

Note: Mean diameter of wells in the SD Area and the L Area = 1.91 m; amsl = above mean sea level (meters above mean sea level).

Table 2.2.  Wells in the DK-G Area5 No.


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

1 1 5 6 7 7 7 8 8A 9 10 10 11 11 11 12 14 15 18 23 24



Diameter (m) Max. height level6



19 20 12 3 19 75 68, 77 19 42 78 26, 27 87, 88 38

0.85 0.85* 0.65 0.60 0.90* 0.80 0.90* 0.67 0.90 0.80 1.31 1.04 0.70* 0.80* 0.80* 0.90* 0.90 0.76 1.68 1.60* 1.04

:50 :50 :71 :72 :77 :78 :78 :87 :92 :95 :107 :109 :119 :119 :119 :121 :143 :144 :150 :154 :156



95 4 3 95 4

−2.56 −2.01 −3.96 −2.59 −1.46 −2.53 w.r. 2.04 w.r. −4.08 −1.58 −2.56

−2.44 −2.44 −2.22 −0.02 1.16

51.90 52.46 50.50 51.87 53.00 51.94 52.43 50.39 52.89 51.90 w.r. w.r. w.r. 53.57 52.02 52.25 54.45 w.r. 53.30

Note: Diameters marked * were measured at a later date; the mean diameter of the DK-G Area wells is 0.93 m; w.r. = without reference.

052-070 WaterUrban Pt1 Ch02 Jansen.indd 56

16/05/2014 15:21

Mohenjo-Daro, Indus Valley Civilization57

Table 2.3.  Wells in the VS Area No.


VS1 2 3 4 5 6 7 8 9 10

A Area



Diameter (m)

Max. height level Source8 (above amsl)


23 6 61 39 22 5 4 1 7.8 2

0.60* 1.37 0.60* 0.90 0.60* 1.10* 0.90* 0.80* 0.80* 0.80*

52.94 54.01 53.60 54.2 53.58 52.10 53.74 49.15 52.84 54.18

:216 :218 :219 :221 :223 :225 :231 :230

Note: Diameters marked * were measured at a later date; the mean diameter of the VS Area wells is 0.95 m.

Table 2.4.  Wells in the HR Area No.


HR1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

A Area 2 2 3



Diameter9 (m)

Max. height level10 (above amsl)



10 18 6 19 51 2 7 58 81 26 24 88 11 11 38 10 6 16 58 103 1 1 6

1.07 0.95* 0.90* 0.80* 0.80* 1.31m 2.05m 0.95* w.r. 0.75* w.r. w.r. 0.50* 0.50* 0.75* 0.90* 1.07* 0.80* 1.l0* 0.80* 0.60* 0.70* 0.90*

54.12 53.20 55.60 53.08 54.04 52.75 53.93 55.20 53.28 53.63 55.00 54.90 52.97 53.56 55.50 55.14 54.50 55.00 53.72 53.98 55.83 55.25 55.30

:179 :180 :182 :183 :188 :189 :189 :190 :192 :199 :199 :200 :201 :202 :203 :204 :205 :206 :207 :207

Note: Diameters marked * were measured at a later date; the mean diameter of the HR Area wells is 0.91 m.

052-070 WaterUrban Pt1 Ch02 Jansen.indd 57

16/05/2014 15:21


A History of Water

Table 2.5.  Wells in the Moneer (MN) Area No.




Diameter13 (m)

Max. height level (amsl)


1 2 3 4 5



10 3 9 2 5

0.60* 0.80* 0.80* 0.70* 1.00*

55.02* 53.78* 53.80* 54.18* 53.95*



Note: Diameters marked * were measured at a later date; the mean diameter of the Moneer Area wells is 0.78 m.

Table 2.6.  Mean well density in individual excavated sectors Excavated area “Citadel” SD L Lower City DK-G DK-A,B,C VS Moneer HR

Surface area



Average freq.15

12.700 6.400 19.100

4 2 6

3.175 3.200 3.183

56.3 m 56.5 m 56.4 m

28.000 12.200 13.000 7.200 20.600 71.000

21 8 (10) 10 5 23 67

1.333 1.525 (1.220) 1.300 1.440 895 1.060

36.5 m 39.0 m (34.9) 36.0 m 38.0 m 32.0 m 33.3 m

As can be seen in the comparative chart in Table 2.6, in the Lower City the average area per well is 1.060 m2, one-third of that of the “Citadel”. Also, the mean diameter of the “Citadel” wells is 1.91 m, much wider than the one in the Lower City (which is below 1 m). If we assume a similar density of wells for the whole city and take the amount of 73 wells for 10 percent of the excavated area, we would have a hypothetical 730 wells for the conventional calculation of 100 ha for today’s visible mounds above the present alluvium. Taking the hypothetical size of 300 ha (200 ha of which is buried under the alluvium), we would reach a total of 2,190 wells. No city in history has ever had even the smaller amount of 73 wells as proven by excavation, not to mention the hypothetical 2,190. Water use areas Almost every house in Mohenjo-Daro was equipped with a “bathroom”. This consisted of a shallow basin or platform, approximately 1 m2 in size, built of sharp-edged bricks which sloped towards an outlet connected to the street drain outside. Animal figurines retrieved from soakpits

052-070 WaterUrban Pt1 Ch02 Jansen.indd 58

16/05/2014 15:21

Mohenjo-Daro, Indus Valley Civilization59

Figure 2.3.  Left: Aerial view of the citadel showing the Great Bath and the so-called stupa, 1982. Right: The Great Bath from north (top is south) with the Granary to the west and residential structures to the east. Aerial view from hot air balloon. 1982. Sources: Both © M. Jansen.

incorporated in house drainage systems may suggest a ritual significance for these elaborate bathing facilities, which could not have served the interests of hygiene alone. This interpretation is also the obvious one for the first known swimming pool in history: the Great Bath of Mohenjo-Daro. Consisting of a rectangular brick basin with a capacity of 160 m3, and entered via a flight of steps at each narrow end, the Great Bath formed the center of an open inner courtyard enclosed within an imposing complex, some 1,800 m2 in area. The very fact that such a large pool was installed within the city points towards a veneration of water in a way that is familiar from other early developed urban civilizations. The city’s effluent disposal system was the logical technological consequence of the townspeople’s extraordinarily high consumption of water. In

052-070 WaterUrban Pt1 Ch02 Jansen.indd 59

16/05/2014 15:21


A History of Water

the houses, bathing platforms and toilet outlets discharged into catchment vessels, which were often connected to the public drain network. Used water from the Great Bath was disposed of through an enormous drain 1.8 m high, big enough to walk through, and roofed with a corbelled vault. Bathing and toilet facilities For technical accomplishment and sheer numbers, the bathing platforms in Mohenjo-Daro were unique in the ancient world. Even in Mesopotamia, where the use of a standard brick as the smallest building element points to certain parallels with Harappa culture, such facilities were practically unknown. A typical bathing platform was installed inside the dwellinghouse, either as a reserved area within a larger room or as a separate, custom-built “bathroom”. To allow for drainage, it consisted of a slightly raised, sloping platform floor built of several layers of bricks, laid in a staggered or criss-cross pattern, polished deep red from frequent use. The close-fitting, sharp-edged bricks were precisely formed in order to keep the joints as narrow as possible; these joints were grouted with clay mortar. The platform was edged in by an outside rim of standing bricks, creating a shallow basin. The platform floor sometimes sloped towards the middle of one side, but in most cases it sloped towards a corner where the effluent was guided either along a gutter or through a wall outlet, from where it discharged into the street drain outside or into a catchment vessel. As a rule, such a bathing platform was installed against an outside wall to facilitate drainage by means of a direct connection to the municipal network. A toilet is often found incorporated into the outside wall of the platform, sometimes fitted with its own vertical chute for discharging the effluent into the street drain or cesspit. A pair of raised side brackets held a seat made of a longer brick or a wooden plank. However, other toilets were also found that consisted simply of a hole in the floor immediately above the effluent chute, in the manner still familiar in the area to this day. The Great Bath in SD Area16 In 1925, a large complex covering approximately 1,800 m2 was unearthed. It soon became known as the “Great Bath” after the sunken basin dominating its central courtyard (cf. Jansen, 1979: 105ff; Jansen, 1983). It appears that the Great Bath was not conceived as part of the original structural concept for the “citadel”, but was built at a later date, as its colossal effluent drain cuts diagonally across the northeastern corner of the lower and older section of the “granary” adjoining the Great Bath on the west. Furthermore, the technical construction of the tank-like basin

052-070 WaterUrban Pt1 Ch02 Jansen.indd 60

16/05/2014 15:21

Mohenjo-Daro, Indus Valley Civilization61

Figure 2.4.  Top left: A curved brick drain after excavation in DK-G Area, 1927. Top right: Bathing platform with water outlet to street in Moneer Area after cleaning, 1982. Bottom left: Reconstruction of a bathing platform with attached well and drain to the street. Bottom right: Brick lined well in HR B, badly damaged, 1979. Sources: Top left: copy from original, M. Jansen. Top right: © M. Jansen. Bottom left: des.R. Bunse, © M. Jansen. Bottom right: © M. Jansen.

with an outer retaining wall and inner bulkhead-type props indicates that it was sunken into older construction layers. The outer retaining walls, up to 2 m thick and at the inner side inclining by 8°, are built in English bond and taper upwards. Originally, a public street up to 5 m wide ran round the perimeter of the complex and set it apart from neighboring structures, thus making the Great Bath the only known freestanding, detached single building in Mohenjo-Daro. This walkway was largely blocked by structures superimposed at a later date, when alterations were also carried out to the interior of the complex. The heart of the complex was a rectangular tank measuring approximately 12 m × 7 m and 2.4 m deep, to which two flights of ten steps each led down: one flight at the northern end and one at the southern end.

052-070 WaterUrban Pt1 Ch02 Jansen.indd 61

16/05/2014 15:21


A History of Water

From the south, the pool courtyard was entered through a surrounding colonnaded gallery, its pillars set at intervals of two-pillar widths. Two entrances in the south of the complex led into a narthex-like anteroom, and from there through nonaligned passages to the gallery and the pool courtyard. From here, further rooms to the east and north could be reached. In all, three concentric walkways enclosed and provided access to the central basin. The innermost ran round its perimeter inside the courtyard, the middle one was the colonnaded gallery itself, and the third was the public street marking the outside boundary of the entire complex. Such a layout, with the pool as its centerpiece, begs the question of ritual processions, or at least of the possibility that the concentric walkways did not serve solely as access routes. A pillared hall, partly blocked up at a later date, once faced the entire northern side of the pool courtyard. But neither here nor elsewhere could any material evidence be produced in support of a posited cult function for the Great Bath complex. Apparently the Great Bath was supplied with fresh water from a large, double-walled well that was built of wedge-shaped bricks; this well was located in one of the compartments to the east. The top of the well shaft was 51.24 m (168’1”) above amsl and served as the benchmark level for the SD Area. A second tube built of standard bricks was laid 33 cm outside the first one as original lining of the well, supporting the hypothesis of a late construction of the whole setting. There is reason to doubt whether the basin with its capacity of 150 m3 was filled by hand, especially as the Indus people were familiar with hydraulic principles. Nevertheless, no trace of any aqueduct-type facility could be ascertained. Just how sophisticated their leveling technique must have been is shown by the basin floor, which inclined towards the effluent outlet in its southwestern corner with a gradient of 1.4–1.7 percent. This rectangular outlet of 391 cm2 (17 cm high × 23 cm broad; cf. Marshall, 1931: 131) at 49.02 m above MSL discharged the used water into a channel (50 cm broad, 10 cm deep), which traversed an underground chamber (6.8 m E–W, 1.73 m N–S) that could be entered from above. From here the effluent flowed into a drain 73 cm broad, which was roofed over with a corbelled vault 2.07 m high. The difference in level between the basin outlet and the point where the corbelled drain breaks off at a distance of approximately 18 m was 0.33 m, the equivalent of a gradient of about 1.8 percent. The extraordinary height of this mighty drain may have been necessitated by the difference in level between that of the bottom of the drain (in turn predetermined by the depth of the basin) and that of the surface of the public street around the outside perimeter of the complex.17 Running west first, the drain then turned north and then west again, thereby cutting diagonally across the northeastern corner of the lower basis of the “granary”. After a few meters running west towards the western edge of the “citadel”, the drain was discontinued. Later, in his excavation in 1950, Wheeler discovered the obvious continuation of the drain (only its

052-070 WaterUrban Pt1 Ch02 Jansen.indd 62

16/05/2014 15:21

Mohenjo-Daro, Indus Valley Civilization63

floor) some 40 m further west. However, he did not recognize it as such (Jansen, 1986: pl. 82). The tank construction of the pool is a technical masterpiece, an outstanding monument to Harappan engineering skills. Its 1.35 m thick inner “shell” was built of precisely formed bricks, laid so close-fittingly in stretcher bond that the joints were only a few millimeters wide. Remains of gypsum18 in the joints led to the assumption that the builders used gypsum mortar which is unlikely, as gypsum is not useful under water. For additional insulation, a 3 cm thick layer of bitumen was placed between this inner 1.35 m thick “shell” and another 0.3 m thick outer brick wall of lower quality. Often attested in Mesopotamia, this use of bitumen for waterproofing purposes in the Great Bath is the only known instance in the Indus civilization. As the Great Bath can be dated into the middle–late urban phase (around 2300 bc), it might coincide with the later appearance of the stone figures (priest king, kneeling figures, ram figures, etc.), also unique in the Indus civilization, which might indicate closer contacts with Western cultures. In turn, the sandwich tank construction was enclosed by an external wall which, as it also served as the foundations for the pillars of the surrounding gallery, was reinforced to the inner side with bulkheadlike buttresses to withstand lateral pressure. Presumably the ultimate function of this outermost retaining wall was to consolidate the actual foundation pit, sunk into older strata, before the freestanding, bitumensealed tank was installed inside. Finally, the remaining intervening gaps were filled in with stamped clay. The Great Bath later underwent substantial alterations. To the north and west, the outermost concentric walkway was reduced to a narrow passage, while the inner colonnaded gallery was subdivided in such a way that it became impossible to walk around the pool. The pillared hall adjoining the north side of the pool was blocked up and filled in, and used as a foundation base for an upper structure reached by a flight of steps. By this time, the chief functions for which the complex was originally designed seem to have been abandoned. When it was excavated in 1925, the entire northwestern portion of the external wall, as well as most of the western row of pillars, was missing. The Great Bath as we know it today is in fact a hypothetical reconstruction in major parts, completed as early as 1927. As in many other parts of the ruined city, remains of later overbuilding with associated Harappan artifacts19 were found among the deposits covering the Great Bath. Obviously this and the other large, non-domestic structures underwent changes of use in the post-urban period. This is taken as a further indication of the city’s loss of urban quality and, consequently, of the collapse of its municipal infrastructure. On the other hand, the presence of Harappan artifacts in the uppermost layers of deposits weighs in favor of the survival of the traditional crafts at least at local level, and hence against the sudden, total abandonment of the settlement. In

052-070 WaterUrban Pt1 Ch02 Jansen.indd 63

16/05/2014 15:21


A History of Water

this regard it can no longer be determined, for instance, whether the destruction of the northwestern portion of the Great Bath revealed by the excavations in the 1920s dates to the Harappan period or whether it is of more recent origin—for example, the legacy of brick robbers or simply the result of erosion. Water sewerage system Perhaps the most impressive engineering feat accomplished by the people of the Indus civilization in Mohenjo-Daro over 4,500 years ago is the network of effluent drains built of brick masonry, which served as the city’s sewerage system. The drains mostly ran along past the houses on one side of the generally unpaved streets, some 50 or 60 cm below street level. U-shaped in cross-section, the sides and bottoms of the drains were built of bricks set in clay mortar, while the open top could be covered in various ways. Obviously the width of the open top was dictated by the dimensions of the covering bricks, which ranged from 25 × 13 × 5.75 cm to 29.5 × 14.6 × 7.6 cm. The drains built of the smallest size bricks vary in width from 17 to 25 cm and in depth from 15 to 50 cm—that is, the equivalent of between two and eight brick courses. Thus the drains range from 260 cm2 to 1,200 cm2 in cross-section. The loose roofing could be removed for cleaning as required. The drains sloped at a gradient of about 2 cm per meter and met at varying levels, depending on cross-section and period. Constructions on curves were sited in such a way that frictional loss was minimized. Wherever a drain had to traverse a longer distance or several drains met, a brick cesspit was installed; this was the simplest method to avoid clogging caused by solids settling. The effluent flowed into such a brick shaft at a high level, filled it, then flowed out the other side at a slightly lower level. The suspended matter gradually formed a deposit which could be removed via steps leading down into the pit, which was likely covered by a loose wooden roof. Besides the closed drains and cesspits for disposing of domestic effluent, open soakpits were also in common use, especially where small lanes opened onto bigger streets. They had to be cleaned of settled deposits from time to time. In places where the street drain was too far away from the houses, closed catchment vessels were installed instead. These were positioned under the vertical chutes on the outside walls and had to be tipped out regularly. In some cases, they were fixed permanently into the walls so that their contents had to be scooped out. Other vessels had perforated bases and worked much like soakpits. It was not unusual for the houses to be connected to the municipal sewerage system by means of individual branch drains, as the bathing and

052-070 WaterUrban Pt1 Ch02 Jansen.indd 64

16/05/2014 15:21

Mohenjo-Daro, Indus Valley Civilization65

toilet facilities were generally situated on the street side of the houses where the effluent could flow straight down through a chute into the public drain or else into a catchment vessel. These built-in chutes very often stopped short some distance above the relevant drain or recept­ acle. In one instance (House 49, HR-B Area),20 a vertical chute serving two toilets ended thus, which must anticipate the original existence of at least one upper storey. Occasionally, vertical pottery drainpipes, formed of stacks of close-fitting, conical sockets approximately 60 cm long, were built into outside walls to discharge effluent from the roof or the floor above. The noxious smells emitted by the effluent chutes, soakpits, cesspits, and drains must have been a serious public nuisance, which could only have been alleviated somewhat by a constant high rate of flow and regular cleaning. Both measures must have implied a large workforce engaged in fetching water for flushing and in clearing out deposits. Analysis of effluent disposal facilities in the DK-G Area21 The DK-G south Area The excavators marked out this sector into 12 structural complexes (blocks), based on surface evidence of brick walls. In digging deeper, further subdivisions of the sectors were made based on identified streets and lanes. For identification purposes, each block was divided into houses and each house into rooms, so that even the smallest compartment can be located.22

Figure 2.5.  Excavated well in DK-G south Area.

052-070 WaterUrban Pt1 Ch02 Jansen.indd 65

16/05/2014 15:21


A History of Water

The following analysis of the effluent disposal system is based on the five horizontal section plans of the southern DK-G Area published by Ernest Mackay in 1938. These plans show the structural situation at the respective development phases corresponding to the selected levels. Studied together with the accompanying detailed descriptions, these section plans provide us with a three-dimensional reconstruction of the development of the drain network. This illustrates not only the changes down through time, but also, at least in part, the anticipated revision of the excavators’ site chronology. Flow capacity estimates The first attempt to calculate the volume of effluent discharged by the drains was carried out in 1977.23 For this test, the drain network of Intermediate II Phase was selected. However, it was not possible to calculate the flow capacity of an interconnected network, as none was found completely intact. As a result, the aim of the investigation— namely to establish by means of capacity calculation whether the system discharged mixed or separated effluent—could not be achieved. In all probability, the drain network served principally to dispose of domestic effluent from the buildings. No evidence of any kind was found that the drains had also served to channel off rainwater from the streets and other public places which, significantly, were left unpaved. Apparently rainwater was simply allowed to seep into the silty ground. In the semi-arid climate24 of the region, with its brief but dramatic seasonal downpours, this must have turned the city into a sea of mud.25 Only in First Street, at a depth of 18’ below the surface, did the excavators encounter indications of a form of street paving consisting of brick and pottery fragments. Ultimately, it may have been the nature of the extremely fine-grained, dense alluvial silt that inspired the Harappans to devise this sewerage system to channel off the everyday household effluent in such a densely populated city as Mohenjo-Daro. Once this was developed, they then installed similar systems in other settlements, so that an effluent drain network has become a characteristic feature of the urban period of the Indus civilization. The dimensions of the drains do not seem to comply with any rational considerations, but rather to have been dictated by the material and construction norms that were then current. Thus, while rarely exceeding 25 cm in width, the drains vary greatly in depth, and it was this latter factor that determined their effluent volume capacity. No doubt the width was fixed by the standard brick length, as only a cover formed of single bricks spanning the entire width would have been adequate to withstand flow pressure. Consequently, the rare examples known of drains wider than 25 cm required a different roof construction (e.g. the corbelled vault roofing the effluent drain of the Great Bath).

052-070 WaterUrban Pt1 Ch02 Jansen.indd 66

16/05/2014 15:21

Mohenjo-Daro, Indus Valley Civilization67

Concluding remarks The purpose of this analysis of the drain networks in the DK-G Area is, first and foremost, to give an account of the three-dimensional distribution of all the ascertainable hydraulic installations in this sector, thereby illustrating the high standard of water supply and effluent disposal services enjoyed at this early period (Early I Period). In addition, a description is given of the layers of drain networks installed in successive occupation levels which could provide the information necessary for a revised stratigraphy. Because of the absence to date of such a three-dimensional reference framework, it has not yet proved possible to revise the stratigraphy worked out by John Marshall (1931) and Ernest Mackay (1938). They based their stratigraphies on the mistaken (as is now known) assumption that the city as a whole grew horizontally layer by layer constantly, and not, as is customary today, on a comparative study of section profiles. Table 2.7 exemplifies the excavators’ false “method” of dating structural remains and their associated finds on the basis of the data of similar heights. The questioning of the traditional stratigraphic interpretation by Mackay (early intermediate, late) accepted to date means, of course, that all interpretations based on it become equally invalid. Consequently, we can make no detailed statements of any kind regarding the development of artifacts, etc., as their relative dating was based on the now rejected stratigraphy, unless we succeed in construing a “realistic” stratigraphy. As a rule, once an excavation has been completed, it is subsequently almost impossible to reinterpret its stratigraphic profile in any way other than that proclaimed by its excavator, because the only material preserved itself consists solely of Table 2.7.  A stratigraphic sequence of DKG in Mohenjo-Daro based on the (false) assumption of even growth in horizontal layers measured at thresholds, etc. (Mackay, 1938) Period Benchmark Late Ib IIa IIb III Intermediate II III Early




Ia −3.2 −5.0 −7.0 −9.9 I −15.9 −20.4 I II (assumed) III (assumed)

+0.8 −0.98 −1.52 −2.13 −3.02 −13.0 −4.84 −6.22 −24.0

+0.24 53.49 52.95 52.34 51.45 −3.96 49.99 48.25 −7.32

Meters above MSL 54.47 54.71

50.51 47.1526

Note: The table showing the stratigraphic sequence of Mohenjo-Daro according to Mackay (1938: XIV– XV) with conversion of elevations from feet to meters by M. Jansen, including the calculation in amsl.

052-070 WaterUrban Pt1 Ch02 Jansen.indd 67

16/05/2014 15:21


A History of Water

interpreted records, while the primary source context, the first-hand excavation in progress, no longer exists. In the case of Mohenjo-Daro, however, there is some hope that at least a rudimentary stratigraphy can be reconstructed. Here the drain networks come to our assistance, as they not only furnish us with telling insights into the highly developed urban infrastructure of the period, but also divulge significant chronological clues. On the assumption that structures dating to a particular settlement phase were serviced by interconnecting drain networks—that is, the existence of the latter is an indication of the at least broad contemporaneity of the buildings thus linked—we investigated the drain situation in a selected area. Our choice fell on the drains exposed by the deep dig in the southern DK-G sector, as this is the only site where periodical alterations to the drain network can be proven archaeologically by means of vertical section profiles. A complete reconstruction of a revised stratigraphy of the DK-G Area would break the bounds of this attempt. In addition to an initial structural analysis, here barely outlined, such a reconstruction would entail an exhaustive investigation of the three-dimensional find distribution of the over 12,000 entries registered in the excavators’ field books. Nevertheless, even the initial study produced enough evidence to show that a major revision of the relative chronology of individual structural clusters or blocks is indeed feasible. Over and beyond their former primary function of effluent disposal, therefore, the drain networks as connecting links between otherwise detached structural units may well prove indispensable for the archaeological reconstruction of the site stratigraphy. Here it must be repeated that the traditional classification of the occupation periods (or levels) (e.g. “Late I–III” or “Intermediate I–III”) no longer has any stratigraphic significance whatsoever—apart from the absolute elevations given in Table 2.7. Consequently, Mackay’s “horizontal section plans” (Mackay, 1938: pls. 101–6), far from corresponding to actual stages in the construction of the city as he believed, represent nothing but horizontal cross-sections at the respective depths. A more realistic stratigraphy based on all data available is feasible. Notes   1 Major parts of this text have been taken from Jansen (1993).   2 The name “citadel” was introduced by Wheeler (1968), having induced the idea of a “priest-king” political system with an elite reigning on the “citadel”. It hosts the following excavations: Banerjii’s excavation, SD Area (after Siddiqui), L Area (Mackay), REM Area (after R. E. M. Wheeler).   3 This well is oval only in its upper portion, presumably due to lateral pressure from the surrounding soil. The lower portion measures 1.7 m in diameter. The Research Project Mohenjo-Daro carried out a deep excavation in this well in 1986.

052-070 WaterUrban Pt1 Ch02 Jansen.indd 68

16/05/2014 15:21

Mohenjo-Daro, Indus Valley Civilization69

 4 The L Area was excavated in 1927 by Ernest Mackay. Why it was named “L” is unknown.   5 The DK Area was named after the excavator K. N. Dikshit. The DK-G Area, however, was excavated after 1927 by Ernest Mackay (Mackay, 1938).   6 The elevations refer to the benchmark level of 178’ 7” above msl (=54.47 m above msl) as marked in Block 9, House I, Room 27. Cf. Mackay (1938: XIV).   7 Cf. Mackay (1938).   8 Marshall (1931).   9 In the excavator’s report (Marshall, 1931), almost no reference is made to the diameter of the wells. The figures marked * were taken from the site documentation of the Research Project Mohenjo-Daro. 10 Likewise omitted from the excavation records are absolute elevation figures. When given, they are “below surface”, so that they are unusable for our purpose. The height measurements given have also been borrowed from the leveling carried out by the Research Project Mohenjo-Daro. 11 Marshall (1931). 12 This Block/House/Room terminology for the structures excavated in the Moneer Area was introduced by the Research Project Mohenjo-Daro; the excavators’ original terminology, preserved only in the field book entries, cannot be reconstructed. 13 The figures marked * were taken from the site documentation of the Research Project Mohenjo-Daro. 14 Aachen Research Project. 15 Calculated as the root of the average catchment area, which is measured as a square. 16 The SD Area was named after the archaeologist Siddiqui. The Great Bath was excavated by John Marshall himself. 17 Marshall speculated, because of its height, about the drain passage being an escape for men. The technical explanation seems to be much more reasonable. 18 This piece of information by the excavators is puzzling in view of the total unsuitability of gypsum mortar for water-bearing structures. Composition according to Marshall (1931: 132): gypsum 43.75 percent, lime 13.78 percent, sand 40.00 percent, alkaline salts 2.4 percent. 19 A sherd bearing a Brahmi inscription, unearthed during the clearing of the basin, was located in a secondary position. Cf. Jansen et al. (1983b). 20 The HR Area was named after H. Hargreaves, another archaeologist who excavated in Mohenjo-Daro between 1925 and 1927. 21 As mentioned in note 5, the DK-G Area, in the north of the lower city, was named after K. N. Dikshit, one of the early excavators. It was excavated from 1927 to 1931 by Ernest Mackay (Mackay, 1937). DK-G south is the only area in which deep digging (down to 6 m below the former surface) took place. 22 This horizontal orientation system differs from those normally used by archaeologists (e.g. grid/squares or coordinate points). The disadvantage of this system (as can be seen in the Moneer Area) is that if the reference system (map) is missing, an orientation is impossible. 23 Cf. Jansen (1977: 157). The capacity calculation was carried out by Herr Ucker, Koln, as part of a publication by the information service of the ceramics industry.

052-070 WaterUrban Pt1 Ch02 Jansen.indd 69

16/05/2014 15:21


A History of Water

24 The climate of the region was probably damper during the peak Indus civilization period than it is today. 25 Even today, seasonal rains make Mohenjo-Daro and the surrounding countryside almost impassably muddy. Our campsite was swamped, and crates, books, and everything else on the ground was simply washed away. 26 It may come as a surprise that the lowest level in DK-G with 47.15 m amsl has the same height as the surrounding plain has, indicating that the “deep dig” by E. Mackay started from a hill. The deepest level of this dig did not reach lower than the surface of today’s plain outside the hill.

references Jansen, M. (1977). “Mohenjo-Daro—5000 Jahre Keramik im Kanalbau part I, II”, in Steinzeug Information Ausgabe 24 & 25, Hrsg.: Fachverband Steinzeugindustrie e. V. Köln (Cologne). ——— (1979). “Architectural Problems of the Harappa Culture”, South Asian Archaeology (SAA) Conference 1977, Naples, pp. 405–31. ——— (1983). “An approach towards the replacement of artifacts into the archaeological context of the Great Bath in Mohenjodaro”, in G. Urban and M. Jansen (eds), Dokumentation in der Archologie. Techniken, Methoden, Analysen. Veröffentlichungen der Seminarberichte vom 5-6. Dezember 1981, Aachen, Geodätisches Institut der RWTH Aachen, Nr. 34, 1983, Aachen, pp. 43–70. ——— (1986). Die Indus-Zivilisation–Wiederentdeckung einer frühen Hochkultur, Köln: Dumont Dokumente Archäologie. Mackay, E. (1938). Further Excavations at Mohenjo-Daro, 2 vols, Delhi. Marshall, J. (1931). Mohenjo-Daro and the Indus Civilization, 3 vols, London. Wheeler, M. (1968). The Indus Civilization, 3rd edn, Cambridge.

052-070 WaterUrban Pt1 Ch02 Jansen.indd 70

16/05/2014 15:21


Xi’an: Water Management and the Development of City Planning in History

Xiaochang C. Wang and Rong Chen History and hydrological features This chapter analyzes the water infrastructure developments from the earliest dynasties to today. Xi’an (in ancient times, Chang’an) was the capital city of China, dating back to the West Zhou Dynasty (1046–771 bc). Until ad 907, Xi’an had been the political center of China for 13 dynasties, of which the Qin Dynasty (221–206 bc) signified the beginning of the unification of China with the ruins of the Terracotta Army left as one of the wonders in the world. Tang Chang’an City, built in the Sui Dynasty (ad 581–618) and further expanded in the following Tang Dynasty (ad 618–907), was the greatest ancient city in Chinese history—it had an urban area as large as 83.1 km2 and a population of up to 1 million in its most prosperous times. However, such a great city was almost completely destroyed during the peasant uprising around ad 900. As a result, the Tang Dynasty was brought to an end and the rebels built the short-lived Hou-Liang Dynasty (ad 907–23) with its capital city at Kaifeng (Henan Province), which was about 600 km eastward from Xi’an. Although Xi’an was no longer a capital city afterwards, it was still a regional central city and performed important roles in politics, culture, and economy in China. The current Xi’an City Wall surrounding the central city area was built during the Ming Dynasty (1368–1644) and is the only well-preserved city wall in China. The name of the city was changed from Chang’an to Xi’an at that time (Hou, 2010). Table 3.1 summarizes the history of Xi’an as a capital and/or local central city in China. Figure 3.1 shows the location of Xi’an within the Chinese territory. It is now the capital city of Shaanxi Province. As the Chinese population is largely concentrated in the eastern provinces, and the region to the west of Xi’an is characterized by large territory but very low population density, Xi’an is actually at the edge of northwestern China, where low annual rainfall and dry climate are the main features (Jiang, 2009). Figure 3.2 shows the surface water system in the Xi’an area (including the central city and suburban area). It was well known that the ancient

071-088 WaterUrban Pt1 Ch03 Wang.indd 71

12/05/2014 11:12

72A History of Water

Table 3.1.  The history of Xi’an as a capital and/or local central city in China Period


Name of the city

Type of city

1046–771 bc 221–06 bc 206 bc–ad 9 ad 265–316 ad 318–29 ad 351–94 ad 384–417 ad 537–57 ad 557–81 ad 581–618 ad 618–907 ad 960–1115 ad 1115–1234 ad 1368–1644 ad 1644–1911 ad 1912–49 ad 1949–Now

West Zhou Qin West Han West Jin Qian-Zhao Qian-Qin Hou-Qin West Wei North Zhou Sui Tang Song Jin Ming Qing Republic of China People’s Republic of China

Feng-Jing/Hao-Jing Xianyang Chang’an Chang’an Chang’an Chang’an Chang’an Chang’an Chang’an Chang’an Chang’an Yong-xing-jun Cheng Jing-zhao Fu Xi’an Fu Hong-man Cheng Xi’an Xi’an

Capital city Capital city Capital city Capital city Capital city Capital city Capital city Capital city Capital city Capital city Capital city Local central city Local central city Local central city Local central city Local central city Local central city

Figure 3.1.  Location of Xi’an in China (plotted by the authors).

071-088 WaterUrban Pt1 Ch03 Wang.indd 72

12/05/2014 11:12

Xi’an: Water Management and the Development of City Planning in History73

capital of Chang’an was surrounded by “Eight Rivers”, namely the Jing, Wei, Ba, Chan, Feng, Hao, Lao, and Jue, shown in Figure 3.2. Of these, the Wei River is the largest tributary of the Yellow River; the others are all sub-tributaries with their origins from the Qingling Mountain Range in the south. The river basin where Xi’an is located is geographically unique because the Qingling Mount is the dividing crest between two watersheds with very different climatic features. The northern watershed belongs to the Yellow River basin, which features a dry climate and low annual rainfall, while the southern watershed belongs to the Yangtze River basin, which has a wet climate and plenty of rainfall. Although Xi’an is in the northern watershed and back against the Qingling Mount, all these rivers were able to receive sufficient mountain runoff so that the eight rivers had almost permanent flows throughout the year and with minor influence from the variation of local climates at that time (Wang, 2001). The richness of surface water also supplemented sufficient groundwater recharge and made this area very suitable for irrigated agriculture from ancient times. For this reason, the city, as indicated by its name of Chang’an in Chinese meaning “long-lasting safety and prosperity”, was well known as a place of mild climate and perennial harvest, and was almost free from flooding or drought. This attracted thousands of people, who thronged into Chang’an from far away until the end of the most prosperous Tang Dynasty. However, as time passed, many rivers gradually changed their routes or lost their original rich flows due to climate changes and overexploitation

Figure 3.2.  River systems in the Xi’an area (plotted by the authors).

071-088 WaterUrban Pt1 Ch03 Wang.indd 73

12/05/2014 11:12

74A History of Water

Table 3.2.  Hydrological parameters of the rivers as surface water resource for Xi’an Rivers

Catchment area (km2)

Hei River* Lao River Feng River Ba River Other streams Total

2,283 665 1,460 2,564 3,171 10,143

Average annual runoff (million m3/yr) 817 179 480 740 271 2,487

Note: * Not included in the ancient eight rivers, but currently used as a main water resource for the domestic water supply to Xi’an city.

of forest resources in the Qingling Mount (Wang, 2001). As a result, the eight rivers gradually lost their rich flow. The Chan became the branch flow of the Ba, and the Hao and Jue became branch flows of the Feng. Due to the limited availability of the main stream of the Wei River and its tributary Jing River from the north, only the tributaries that originated from the southern mountain range can provide surface water resources to Xi’an city. Table 3.2 summarizes the hydrological parameters of these rivers based on the early data from 1956 to 1980 (Editing Committee of Water Resources in Xi’an City, 1999). Groundwater has been used as an important water resource in Xi’an for a long time. It is estimated that the total groundwater resource amounts to 1.695 billion m3/yr. However, there is an overlap between surface water and groundwater resources estimated as 1.036 billion m3/yr. Therefore, the total water resource (surface water + groundwater – overlap) can be evaluated as about 3.146 billion m3/yr (Editing Committee of Water Resources in Xi’an City, 1999). Historical experiences of urban water supply and water management As Xi’an has been a national or regional central city for a long time, it also has a long history of urban water supply (Du and Qian, 1998). Figure 3.3 shows the location of Xi’an City in different dynasties (Editing Committee of Water Resources in Xi’an City, 1999). In the West Zhou Dynasty (1020–771 bc), Xi’an became a capital city for the first time, on a limited scale. For the convenience of water supply, the Imperial Palace was constructed on low land adjacent to the Feng River, where surface water was directly carried to the Palace by simple earth ditches—the oldest and easiest way for urban water supply. However, the relatively small flow of the Feng River could not meet the needs of increasing water supply caused by the expansion of the city. Therefore, in the later Han Dynasty (202 bc– ad 8), the Imperial Palace was relocated to a place nearer to the Wei River

071-088 WaterUrban Pt1 Ch03 Wang.indd 74

12/05/2014 11:12

Xi’an: Water Management and the Development of City Planning in History75

Figure 3.3.  Location of Xi’an City in different dynasties (plotted referring to Editing Committee of Water Resources in Xi’an City, 1999).

where larger water supply and drainage systems were constructed for obtaining water from more rivers and discharge floods to the main river channel. In the following Tang Dynasty (ad 618–907), the city was moved further south to a higher location, and even larger urban water systems were constructed. Urban water supply for the Han Chang’an City The construction of the capital city in the Han Dynasty experienced three periods. In 202 bc, when several palaces were newly constructed, the urban area was very small. In the period of 194–189 bc, with the completion of the city walls, the urban area grew to 35 km2. In the following years, various palaces on a large scale were successively constructed, and the urban population gradually grew to more than 300,000 as a result of migration of people from the surrounding area. For such a large city, the ancient way of fetching water directly from the nearby natural streams could no longer meet the demand of urban water use; it therefore became necessary to store and regulate surface flows for the provision of sufficient water to the city. The famous Kunming Lake was constructed at that

071-088 WaterUrban Pt1 Ch03 Wang.indd 75

12/05/2014 11:12

76A History of Water

Figure 3.4.  Kunming Lake and the water supply and drainage system for Han Chang’an City (202 bc–8 ad) (plotted referring to Editing Committee of Water Resources in Xi’an City, 1999).

time as the first urban water reservoir for water storage and regulation. As shown in Figure 3.4 (Editing Committee of Water Resources in Xi’an City, 1999), the Kunming Lake, with a storage capacity of 30 million m3, received the flow from the upstream of the Feng River through a waterway. From the outlet at its north end, another waterway could transfer water to the open channel system for both water distribution and drainage during flooding seasons. Smaller channels and ditches (not shown in Figure 3.4) were also provided for sending water to the city area. The average daily flow from the Feng River could amount to 500,000 m3/day—sufficient to cover the various water demands in the city. Urban water supply for Tang Chang’an City The Tang Dynasty (ad 618–907) was the most prosperous time in Xi’an’s history. It was also a period of construction of large-scale surface water systems to meet the demands of such an ancient megacity, which had more than 1 million residents at this time. Tang Chang’an City was built 10 km southeast of Han Chang’an City (Figure 3.3) in order to prevent flooding from the lowlands adjacent to the Wei River.

071-088 WaterUrban Pt1 Ch03 Wang.indd 76

12/05/2014 11:12

Xi’an: Water Management and the Development of City Planning in History77

Figure 3.5.  Tang Chang’an City in the most prosperous period (ad 618–907) (available at http://image.baidu.com).

Tang Chang’an City was a well-planned, huge city (Figure 3.5). It covered 83.1 km2 and was surrounded by city walls with a perimeter of about 36 km. The city plan was like a chessboard—with 11 main streets running from south to north, and 14 avenues stretching from east to west. Ensuring water supply to this walled huge city was in any sense an extraordinarily onerous task at that time. Figure 3.6 shows the river and channel systems for water supply and drainage for Tang Chang’an City (Hou, 2010, 2012) after long years of construction and expansion to meet the growing needs of the whole city. The system, when ultimately completed, consisted of three main channels to supply water to the city: the Longshou Channel at the east which led water from the Chan River; the Qingming Channel at the south which transferred water from the Jue River; and the Yong’an Channel at the southwest which brought water from the Hao River. Some other channels were also constructed for either water supply or drainage. All together, they formed a well-planned water system for the city. As Tang Chang’an City was at a higher elevation than the previous Han Chang’an City (Figure 3.4), Kunming Lake could no longer perform the function of water storage and regulation. However, because several natural streams were utilized in the water system, the water supply was well maintained without large storage facilities. Urban water supply for Xi’an in the Ming and Qing Dynasties Tang Chang’an City’s prosperity disappeared quickly after the Tang Dynasty was destroyed as a result of a peasant uprising. Afterwards,

071-088 WaterUrban Pt1 Ch03 Wang.indd 77

12/05/2014 11:12

78A History of Water

Figure 3.6.  River and channel systems for Tang Chang’an City (ad 618–907) (plotted referring to Editing Committee of Water Resources in Xi’an City, 1999).

China entered the period of warlords. The ancient megacity of Chang’an suddenly became ruins and the water systems, shown in Figure 3.6, were almost completely abandoned. Although Chang’an was set as a prefecture (the second rank of cites following the capital city) from the beginning of the 1000s in the Song Dynasty, its prosperity was not recovered much until the Ming Dynasty (which started in 1368), when the city received its current name of Xi’an. The walled central city (Figure 3.3) was constructed almost at the original place of Tang Chang’an City, but occupied only about a sixth of the original area. In 1379, the prefecture governor decided to repair the Longshou Channel, which was used to supply water to the newly built city from the east side. Closed brick channels were constructed for distributing water to different city areas. In 1476, the Tongji Channel was built for transferring water from the Jue River into the city from the west side. Figure 3.7 shows an old sketch of the brick channel system in Xi’an City at that time (Editing Committee of Water Resources in Xi’an City, 1999). The main channel and branch channels were built along the main streets. A number of outlets were built along the channel, each distributing

071-088 WaterUrban Pt1 Ch03 Wang.indd 78

12/05/2014 11:12

Xi’an: Water Management and the Development of City Planning in History79

Figure 3.7.  Channel systems for water supply in Xi’an City in the Ming and Qing Dynasties (1368–1911) (plotted referring to the graph available at http://image.baidu.com/i?tn=baidu image&ipn=r&ct=201326592&cl=2&fm=index&lm=-1&st=-1&sf=2&fmq=&pv=&ic=0 &nc=1&z=&fb=0&istype=2&ie=utf-8&word=%E6%98%8E%E6%B8%85%E8%A5%BF%E 5%AE%89%E5%9F%8E%E5%86%85%E4%BE%9B%E6%B0%B4%E6%B8%A0%E9%81%93).

water to every ten families. In the Xi’an “Tablet Forest” Museum, there is a tablet from that time on which the “Provisions of Water Management” are depicted. The items in the provisions included those on water source protection, water channel maintenance, water tariffs, and so on. Such a channel system was used as the sole facility for water supply throughout the Ming and Qing Dynasties (1368–1911). However, it was difficult for the channel systems to be expanded to meet the increasing demand for water supply, and their maintenance and repair required heavy labor and unaffordable costs at that time. Near the end of the nineteenth century and the beginning of the twentieth century, people began to dig shallow wells to mitigate the shortage of water supply from the channel systems (Editing Committee of Water Resources in Xi’an City, 1999). Fast city expansion from the 1950s Historically, the city wall of Xi’an was a clear boundary between the city and the countryside. However, fast urbanization was seen in Xi’an after the establishment of the People’s Republic of China. Since then, the urban area has grown across its previous boundary and the city wall has become only a symbol of ancient Xi’an. In order to understand how fast Xi’an City has been growing, a comparison of city plans from over the past six decades is necessary. It should be pointed out that although in many countries, a city plan should be valid for a long timespan counting in decades or even centuries, it is not a condition

071-088 WaterUrban Pt1 Ch03 Wang.indd 79

12/05/2014 11:12

80A History of Water

Figure 3.8.  Urbanization in China since 1949 (plotted by the authors using statistical data).

in China, due to uncertainties of urban development in this previously agricultural-based country. As can be seen from Figure 3.8, in 1949 only 10.64 percent of Chinese people were living in non-agricultural areas. A fast increase in this percentage occurred until 1960, but this then decreased in the following years. Until 1978, urban populations in China totaled less than 18 percent. The percentage of urban populations began to increase from 1979 to 1995, at an almost constant rate. However, this accelerated after 1995. In 2009, 46.59 percent of people were living in cities and towns. Table 3.3 compares the scales of urban plans for Xi’an City from 1953 to 2010 (NDRC, 2009; Xi’an Municipality Planning Bureau, 2009). The First Urban Development Plan was formulated at the beginning of the 1950s for urban development over a period of 20 years from 1953 to 1972. It was estimated that the urban area would cover 131 km2 and the population would grow to 1.2 million. This plan played an important role in the development of the city from 1953 to 1959. However, due to natural disasters and famine at the beginning of the 1960s and the Cultural Revolution from 1966 to 1976, city development was very slow and did not follow any specific plan. The Second Urban Development Plan was formulated when the Cultural Revolution ended and the central government started an “open door policy” for economic reform. This was another 20-year plan towards the end of the twentieth century. However, the plan was relatively conservative, as can be seen from the planned urban area and population in comparison with the first plan. Ultimately, this plan did not meet the demands for the city’s development, because the scale of Xi’an City (in terms of both urban area and population) became much larger than had

071-088 WaterUrban Pt1 Ch03 Wang.indd 80

12/05/2014 11:12

Xi’an: Water Management and the Development of City Planning in History81

Table 3.3.  Comparison of the urban plans for Xi’an City Plan


Urban area (km2)

Urban population (million)

First Urban Development Plan Second Urban Development Plan Third Urban Development Plan Fourth Urban Development Plan International Metropolitan Plan

1953–72 1980–2000 1995–2010 2008–20 2010–20

131 162 275 490 800

1.20 1.80 3.10 5.28 10.00

been forecast at the beginning of the 1990s—about ten years before the target year of 2000. In order to keep pace with urban growth, the Third Urban Development Plan was put forward five years earlier than it should be. This new plan, with a timespan of 15 years from 1995 to 2010, was a great advance from the previous one to keep pace with the speed of urban development at that time. Nevertheless, the plan was still unable to meet the demands of fast development. Five years before the target year of 2010, Xi’an City had already expanded larger than the planned scale set for the target year. In such a situation, the Fourth Urban Development Plan, covering ten years, was again a great advance from the third. Towards the target year of 2020, the urban area and population were estimated to be 490 km2 and 5.28 million, respectively. Shortly after the Fourth Urban Development Plan was put forward, a new regional development scheme, the “Guanzhong–Tianshui Economic Zone”, was approved by the central government (NDRC, 2009; Wang, 2011). This is an interprovincial development scheme covering the Guanzhong (Central Shaanxi Basin) area within this province and the Tianshui area in the neighboring Gansu Province. It is comprised of seven cities (six in Shaanxi and one in Gansu), and covers a total area about 80,000 km2 and a population over 30 million. The establishment of such an economic zone is part of the national actions for “Developing the Northwest China”. As Xi’an has long been a central city in the northwest region, its leadership functions are important for regional development. Therefore, within the Guanzhong– Tianshui Economic Zone Development Scheme, Xi’an and another city, Xianyang, will be merged into a larger city, called the Greater Xi’an International Metropolitan (Zhang, 2010). This will cover an urban area of 800 km2 with a population of over 10 million by the year 2020 (Pei, 2011). City planning with water availability as a limiting factor Availability of water resources in the Xi’an area The total water resource in the Xi’an area has tended to decrease. From the earlier hydrological data series of 1956 to 1980, it was estimated that

071-088 WaterUrban Pt1 Ch03 Wang.indd 81

12/05/2014 11:12

82A History of Water

Table 3.4.  Comparison of water consumptions in the Xi’an area in 1993 and 2005 (million m3/yr) Year 1993 2005

Area Central city Suburban Whole Central city Suburban Whole

Total 591.5 943.2 1,534.7 759.6 998.9 1,758.5

Domestic Urban


122.5 15.1 137.6 164.3 35.0 199.3

10.8 54.6 65.4 7.8 55.8 63.6

Agriculture Industry Service Ecology 103.0 709.3 812.3 99.2 686.6 785.8

321.5 159.5 481.0 343.1 212.5 555.6

33.7 4.7 38.4 104.2 7.4 111.6

41.0 1.7 42.7

the total water resource could amount to 3.146 billion m3/yr (Editing Committee of Water Resources in Xi’an City, 1999). However, recent analysis based on the data series of 1981 to 2005 resulted in an estimation of 2.347 billion m3/yr—about 25 percent lower than the earlier estimation (Water Authority of Xi’an City, 2010). Improvement of the hydrological analysis method may have erased the inaccuracy of the previous estimation, but climate change is also an important reason for the decrease. Of the total water resources, the quantity of surface water resource is 1.973 billion m3/yr. Taking 40 percent as the limiting safety level for surface water withdrawal, exploitable surface water resources can be estimated as 790 million m3/yr, while exploitable groundwater resources are estimated as 900 million m3/yr. Therefore, total exploitable water resources in Xi’an can be valued as 1.69 billion m3/yr. Water demand and water supply in the Xi’an area Discussion on the supply and demand relationship for water has to cover not only the central city, but also the satellite towns and villages in suburban areas. By using available data, we can compare water consumptions in 1993 and 2005 to see how water use varied in different sectors with the expansion of urban areas. As can be seen from Table 3.4 (Water Authority of Xi’an City, 2010), in 1993, when the total and urban populations in the Xi’an area were about 4.5 million and 2.5 million, respectively, the total water consumption was 1.535 billion m3/yr—a value lower than that of its total exploitable water resources. In 2005, when the total and urban population in the Xi’an area grew to about 8 million and 5.2 million, respectively, total water consumption increased to 1.759 billion. This surpassed the total exploitable water resources, indicating that water shortage had begun to be a problem in the Xi’an area. It is noticed that as a result of urbanization, the percentage of agricultural water consumption has much decreased, while a substantial increase has been seen in domestic water consumption. There has also been a great increase in the percentage of water consumption for urban service.

071-088 WaterUrban Pt1 Ch03 Wang.indd 82

12/05/2014 11:12

Xi’an: Water Management and the Development of City Planning in History83

Ecological water use became a new category recently, due to the increasing needs for the restoration of urban water environments, such as urban lakes and streams. Projection of water demand to the target year of 2020 Based on the Fourth Urban Development Plan for Xi’an City (Xi’an Municipality Planning Bureau, 2009), the annual water demand to the target year of 2020 was projected as 2.592 billion m3/yr. In comparison with the exploitable water resources, calculated at 1.69 billion m3/yr, there is a gap of 902 million m3/yr between the demand and the resources available. The details of the water demand projection are shown in Table 3.5 (Water Authority of Xi’an City, 2010). Inter-basin water transfer for solving water shortages in the Xi’an area From the above discussion, it is clear that there has been an imbalance between water consumption and exploitable water resources since 2005, or before. The latest data (unpublished) show that in 2010, the total water consumption in the whole Xi’an area amounted to 2.1 or 2.2 billion m3—30 percent more than the exploitable resources available. Water supply is in fact supported by overexploitation of surface water and groundwater in the Xi’an area. For this reason, most of the streams in this area often have a very limited base flow in their channels, providing insufficient ability to absorb and/or dilute pollutants from point and/or non-point sources. Surface water pollution thus becomes a serious problem, especially for the Wei River, because many of its branch streams cannot carry sufficient flow to its main channel. In extremely dry seasons, much of the water in the Wei River is diverted from the river channel in its upstream agricultural area for irrigation use. When the stream reaches Xi’an with very limited base flow, domestic or industrial effluent from the urban area may take a large quantity of the water volume in the river channel. Depletion of the groundwater table is also a serious problem in this area, due to overpumping (Zhang et al., 2011).

Table 3.5.  Water demand projection in the Xi’an area for 2020 (million m3/yr) Area Central city Suburban Whole

Total 1,200.9 1,390.7 2,591.6

071-088 WaterUrban Pt1 Ch03 Wang.indd 83

Domestic Urban


280.6 105.6 386.2

0 56.5 56.5


Industry Service Ecology

 90.8 891.9 982.7

534.7 277.7 812.4

199.7  18.1 217.8

 95.1  40.9 136.0

12/05/2014 11:12

84A History of Water

In order to solve both the water shortage and water pollution problems, inter-basin water transfer is set as the main measure in the Fourth Urban Development Plan and the Greater Xi’an International Metropolitan Plan (Tian, 2010; Zhang and Wang, 2010). The territory of Shaanxi Province stretches over two river basins—the Yellow River Basin in the north and the Yangtze River Basin in the south. The Qingling Mountain range is the line of demarcation between the two basins. As the annual rainfalls in the north and south of the mountain range differ much from each other, the northern part of Shaanxi Province only possesses 21.1 percent of the provincial water resource, though its territory and population take 68.5 percent and 77.3 percent, respectively. In contrast to this, the southern part of Shaanxi Province possesses 78.9 percent of the provincial water resource, but covers a much smaller territory and has a smaller population. The inter-basin water transfer project will be implemented within the provincial territory by constructing a tunnel across the Qingling Mountain range to transfer water from the Han River, which is the tributary of the Yangtze River, to the Wei River. The construction cost is estimated as 2.5 billion US dollars and the construction work will take about 11 years. When it is completed, an annual flow of 1.55 billion m3 will be sent to the Wei River upstream of Xi’an (Research Institute of Yangtze River’s Water Resource Protection Science, 2011). Such a large flow is considered sufficient to cover the gap between the water demand and exploitable water resources in the Xi’an area. Meanwhile, the base flow in the main channel of the Wei River can be supplemented so that its quality can be much improved. Consideration on alternative options If the abovementioned trans-basin water transfer project is an “active” measure to solve the water shortage problem, there are other “conservative” options under discussion among city planners and environmentalists. We say these options are conservative because they are based on alternative measures without large-scale water transfer from other basins. Option 1: water saving Water saving is a topic that has been discussed for a long time. The basic idea is to take effective measures for decreasing water demand and consumption by water saving from sectors where large percentages of water resource are used. From Tables 3.3 and 3.4, we can see that agriculture has been and will probably continue to be the largest sector for water use. Although agricultural water consumption decreased from 52.9 percent in 1993 to 44.7 percent in 2005, and is predicted to decrease further to 38 percent in 2020, it is still possible to save water from this sector. In the Xi’an area, irrigated agriculture by far has depended mainly on flooding irrigation. If

071-088 WaterUrban Pt1 Ch03 Wang.indd 84

12/05/2014 11:12

Xi’an: Water Management and the Development of City Planning in History85

the irrigation method can be innovated, such as by practicing trickling irrigation, at least half of the irrigation water can be saved. Only by this may there possibly be a reduction of more than 400 million m3/yr from the agricultural water demand shown in Table 3.4, which is almost equivalent to the total domestic water demand. As for the industry, by practicing water recycling, it is possible to realize “increase in production without increasing water consumption”, so that industrial water demand in 2020 may be kept at the 2005 level. This is another reduction of about 250 million m3/yr. By water saving from the agricultural and industrial sectors, the gap between total demand and exploitable water resources may be reduced from 902 million m3/yr to about 250 million m3/yr; this may be solved by water reuse or some other measures to be discussed below. Practicing agricultural and industrial water saving will need additional investment. However, the investment may not be as high as the cost of the construction of the trans-basin water transfer project. Option 2: water reclamation Currently, about 90 percent of the domestic waste water (including waste water from the third industry for urban service) from urban areas is collected and sent to domestic waste water treatment plants. It is planned that by 2020, the urban waste water collection and treatment system will cover the whole urban area (including the central city and towns in the suburban area). If the total water consumption for domestic and urban service is referred, the collectable waste water will amount to about 500 million m3/yr (80 to 85 percent of the water supply). If the effluent from the domestic waste water treatment plants can be reused for non-potable purposes, such as irrigation (direct use of the secondary effluent), industry, urban ecology, and other municipal uses (after tertiary treatment by additional facilities), the maximum capacity of water reclamation will be more than 400 million m3/yr (80 to 85 percent of the collected waste water). Rainwater harvesting is another means of usable water reclamation. If 5 percent of the annual rainfall can be harvested, it can produce about 140 million m3/yr water for non-potable water supply, as mentioned above. Therefore, the total potential of water reclamation can be evaluated as 540 million m3/yr. In China, a policy has already been set by the central government that for all cities in the northern water-deficient region, at least 40 percent of the treated domestic waste water should be reused by 2015 or 2020. In this sense, water reclamation is not merely an option but something that must be done in Xi’an City. Option 3: adjustment of water allocation in the whole river basin In the discussion above, the available water resources in the Xi’an area did not include the main stream of the Wei River at all, though it is the largest river among the “Eight Rivers Surrounding the Old Capital”. This

071-088 WaterUrban Pt1 Ch03 Wang.indd 85

12/05/2014 11:12

86A History of Water

is because the Guanzhong (Central Shaanxi Basin) area has long been the main food production area in Shaanxi Province and water in the Wei River, after entering the province from the west, is allocated mainly to agricultural irrigation through an open canal system, which diverts water from a location about 200 km upstream of Xi’an. The total water resource of the Wei River within the provincial territory is estimated as 6.02 billion m3/yr on average, while the total water withdrawal amounts to 4.62 billion m3/ yr (Gu, 2009; Wang et al., 2012) which is 76.7 percent of the total, indicating an abnormally high percentage of water resource development beyond the sustainable level. Of the total water consumption, that consumed for agricultural irrigation takes 50–60 percent (2.31–2.77 billion m3/yr). Similar to the situation in the Xi’an area, flooding irrigation is the main irrigation method in the whole Wei River basin area. If 20 percent of the irrigation water (462–554 million m3/yr) can be saved by adopting more advanced irrigation technology, and this amount of water can be allocated to Xi’an, the water shortage problem may also be alleviated to a great extent. Option 4: restriction of the urban development scale If the exploitable water resources of 1.69 billion m3/yr are taken as a limiting factor for determining the urban development scale, it can be roughly estimated that an urban scale no larger than that of the 2005 level—namely an urban population of about 5 million and a total population of about 8 million—may be supportable by the water resources (see Table 3.3 for water consumption in 2005). From the earliest times for humankind, “living near the water” has been a principle for people to choose a location to build their homes. Such a principle implies humankind’s harmonic existence with nature or the sustainable support from nature. Unfortunately, this principle has been forgotten since urbanization became the symbol of social and economic development in modern society. From the viewpoints of sustainable utilization of natural resources and environmental conservation, we would like to propose the restriction of the urban development scale as an option for solving the water shortage problem. This idea may not be welcomed by the decision makers who are seeking ever faster development and the miracle of economic growth, but it is one thing we more or less have to consider. Concluding remarks The history of city planning is closely related to water management. Xi’an, as an ancient capital city, has a rich history of exploring water resources and supplying water to the city area under different conditions and in different ways. The most ancient way was to build the city as close to a stream as possible, so that water supply would become very simple. The city also

071-088 WaterUrban Pt1 Ch03 Wang.indd 86

12/05/2014 11:12

Xi’an: Water Management and the Development of City Planning in History87

has had a very unique experience of supplying water to over 1 million residents by a channel system during the Tang Dynasty (ad 618–907), due to the very rich surface water resources. A harmonic coexistence of human beings with water flowing in a natural way was realized in the most prosperous time of Tang Chang’an City. This provided an example of good water management in a well-planned, ancient megacity. The city did not face difficulties in the acquisition of water to meet the demand of various water uses until it suddenly expanded after the 1980s; this can be seen by comparing the frequently updated urban development plans. The very fast urbanization almost totally changed the relationship between people and water in this area, resulting in a big gap between water demand and exploitable water resources in the local basin. However, people are still dreaming about the restoration of the old scenery of a city surrounded by flowing streams (Ecological Society of Shaanxi Province, 2004). How can we make such a dream a reality with the loading on the water environment surpassing its carrying capacity (Varis and Vakkilainen, 2001)? Is trans-basin water transfer a good solution for the sustainable development of this city? These questions may be difficult to answer, but we have to consider them when the urban development plan is to be updated again. references Du, P. F. and Y. Qian (1998). “Water supply of the cities in ancient China”, China Historical Materials of Science and Technology, 19(1), pp. 3–10. Ecological Society of Shaanxi Province (2004). Research Report on the Restoration of Eight Rivers Surrounding Chang’an, Xi’an: Ecological Society of Shaanxi Province. Editing Committee of Water Resources in Xi’an City (1999). Water Resources in Xi’an City, Xi’an: Shaanxi People’s Press. Gu, M. X. (2009). “Analysis of the situation and trends of the water resources and water consumption mount in recent years in Weihe River Basin of Shaanxi province”, Journal of Water Resources & Water Engineering, 20(1), pp. 143–5. Hou, Y. J. (2010). “Ideas and pursuits for the book series—Urban History of Xi’an”, Journal of Chang’an University (Social Science Edition), 12(4), pp. 16–20. ——— (2012). “Preliminary analysis of the vitality of Xi’an city”, Jianghan Forum, (1), pp. 13–19. Jiang, Y. (2009). “China’s water scarcity”, Journal of Environmental Management, 90, pp. 3185–96. National Development and Reform Commission (NDRC) (2009). Guanzhong— Tianshui Economic Zone Development Plan, Beijing: National Development and Reform Commission (NDRC) of China. Pei, C. R. (2011). “Research on the strategy of the urban structure and international metropolitan development for Xi’an”, The Journal of Humanities, (1), pp. 186–9. Research Institute of Yangtze River’s Water Resource Protection Science (2011). Environmental Impact Assessment Report on the Trans-Basin Water Transfer

071-088 WaterUrban Pt1 Ch03 Wang.indd 87

12/05/2014 11:12

88A History of Water

Project, Wuhan: Research Institute of Yangtze River’s Water Resource Protection Science. Tian, Y. (2010). “Suggestions on the international metropolitan plan of Xi’an”, Xi’an Social Science, 28(6), pp. 69–70. Varis, O. and P. Vakkilainen (2001). “China’s 8 challenges to water resources management in the first quarter of the 21st century”, Geomorphology, 41, pp. 93–104. Wang, J. (2001). “Reflection correlated with the reasons that lead to the river system’s exhaustion of Ancient Capital Chang’an”, Journal of Northwest Institute of Architecture Engineering (Natural Science), 18(4), pp. 97–100. Wang, S. H. (2011). “My suggestions for constructing Xi’an into an internationally metropolitan city”, Journal of Chang’an University (Social Science Edition), 13(1), pp. 7–15. Wang, T., W. H. Zeng, and M. C. He (2012). “Study of the seasonal water environmental capacity of the central Shaanxi Reach of the Wei River”, Procedia Environmental Sciences, 13, pp. 2161–8. Water Authority of Xi’an City (2010). Middle and Long Term Water Demand and Supply Plan, Xi’an: Water Authority of Xi’an City. Xi’an Municipality Planning Bureau (2009). The Fourth Urban Development Plan of Xi’an City, Xi’an: Xi’an Municipality Planning Bureau. Zhang, M. J. and F. Wang (2010). “Xi’an: Toward the international metropolitan”, West China Development, (5), p. 46. Zhang, P., F. X. Cheng, and T. Tian (2011). “Research on Xi’an urban culture system under the background of international metropolis”, Urbanism and Architecture, (3), pp. 122–4. Zhang, Y. B. (2010). “Consideration on the international metropolitan plan of Xi’an”, The New West, (24), p. 11.

071-088 WaterUrban Pt1 Ch03 Wang.indd 88

12/05/2014 11:12


Phoenician Cities and Water: The Role of the Sacred Sources in the Urban Development of Motya, Western Sicily

Federica Spagnoli1 Introduction Water played a primary role in Near Eastern religious beliefs and cults. The Phoenicians,2 one of the foremost populations living in the Levant at the beginning of the first millennium bc, erected their cult buildings close to natural spring waters, and the water was collected inside huge sanctuaries, such as Bostan esh-Sheikh in Lebanon, Amrit in Syria, and many other sacred compounds. Moreover, the presence of fresh water has been one of the necessary conditions for the rising of a city since early urbanization. This chapter aims to present and discuss the case study of the Phoenician site of Motya, where a sacred compound, named “Temple of the Kothon”, was erected close to a fresh water spring by the first settlers coming from the homeland. This highlights how the creation of this sacred area molded the later urban plan of the city.3 The Temple is strictly linked to an artificial basin, the so-called “Kothon”,4 which collects the water of the under-hearth spring. This arrangement has close similarities to some of the most important sacred compounds of the Phoenician homeland, both from an architectural and a cult point of view. After a brief introduction about the Phoenician origins of the city of Motya, the main architectural phases of the Temple of the Kothon and the principal finds will be described, and the cult installations found inside the Temple, and the role the water held in the cult, will be interpreted in light of the results of recent excavations. The Phoenician expansion in the Mediterranean Sea: the foundation of Motya The Phoenician culture5 grounds its roots in the Bronze Age Levantine cultures, with which it shares politics, social structures, material culture, and religiosity, but it had its floruit in the first half of the first millennium bc,

089-106 WaterUrban Pt1 Ch04 Spagnoli.indd 89

16/05/2014 15:08

90A History of Water

from the twelfth century to the sixth century bc. During this period, a milestone of Phoenician history was the rising of the Assyrian Empire and the conquest of the Levant in the eighth century bc. The new political layout of the region caused Phoenician cities to prefer to pay tributes to Assyrians over losing their political independence. Moreover, the Assyrian control of the principal trade-routes from the Egypt and the Red Sea to Anatolia made Phoenicians increase maritime commerce through the west. Many ports of trade, also called emporia, rose up on the coasts of the Mediterranean.6 Therefore, the presence of Phoenicians along the coasts of the Mediterranean Sea, already testified by archaeological evidence from the beginning of the Iron Age (eleventh and tenth centuries bc), grew in importance and influence from the middle of the eighth century bc, with the settlement of new foundations in Italy (Sicily and Sardinia), Iberia, and North Africa.7 Western Phoenician cities are characterized, as in Phoenicia, the present-day Lebanon and coastal Syria, by the presence of a neighboring spring, a river, or a lake basin that could assure survival and population growth. These conditions occurred at Motya, an island city of western Sicily, where the first Phoenician settlement, dating back to the first half of the eighth century bc, arose in the place where water emerged from underground, forming a natural lake. The first founding action of the new settlers was probably to put water under the control of a divine authority, making the area consecrated to a deity by offering rites and cult activities in favissae8 and cult pits. In the second half of the eighth century bc the fresh water spring was collected to a sacred building, the Temple of the Kothon, by underground canals (Nigro, 2004b). Therefore, this temple represents an important attestation of Phoenician culture in the Western Mediterranean, indicating the sacred use of water in this central region of the Phoenician world. Archaeology of Motya: old excavations and recent discoveries Motya is a little island 45 ha in size and 1 km off the coast of Sicily. It is located between Marsala in the south and Trapani in the north, in the center of the Marsala Lagoon; it is enclosed and protected on the western and northern sides by several islands, of which the largest and most important is called Isola Lunga (Figure 4.1). Motya hosts one of the most important Phoenician sites of the western Mediterranean since the first phases of Phoenician expansion into the West at the beginning of the eighth century bc. Motya is an exceptional case in Phoenician archaeology. It was a flourishing city from the end of the eighth century to the start of the fourth century bc. At that point it was conquered and razed by the Greeks, and afterwards was gradually abandoned.9 As a consequence, there is a lack of

089-106 WaterUrban Pt1 Ch04 Spagnoli.indd 90

16/05/2014 15:08

Phoenician Cities and Water91

Figure 4.1.  Motya (Sicily) in the center of the Marsala Lagoon. In the background, Eryx, an Elymian site, is located on the top of Mt San Giuliano; down the western slopes, the city of Trapani: view from southeast. Source: Rome “La Sapienza” University Expedition to Motya.

later superimposing occupation levels and so there is broad preservation of the Phoenician layers.10 The site has been investigated since the end of the nineteenth century, but only in the second half of the last century have methodological stratigraphic excavations been carried out. In the 1960s the archaeological mission of the University of Leeds, directed by Professor B. S. J. Isserlin, explored the southwestern region of the island, near the South Gate. A residential quarter was found here,11 and a rectangular basin carved into the rock, the so-called “Kothon”, where a small portion (just the monumental entrance) of a building located on the east of the artificial basin emerged (Isserlin et al., 1974). The activities of Professor Isserlin paved the way for the joint Archaeological Expedition of La Sapienza University and the Soprintendenza of Western Sicily, which carried out meaningful research activities at Motya from 1964 to 1993. Since 2002, after a hiatus of nearly a decade, the Archaeological Mission to Motya of the University of Rome “La Sapienza” resumed investigations on the island (Figure 4.2).12 Excavations in the southern part of the island (Area C) revealed, since the first archaeological field carried out in 2002, the presence of a monumental sacred compound (Figure 4.3). This was delimited by a Circular Temenos 118 m in diameter, including the Kothon—actually a sacred pool

089-106 WaterUrban Pt1 Ch04 Spagnoli.indd 91

16/05/2014 15:08

92A History of Water

Figure 4.2.  Excavation areas of the Archaeological Mission of “La Sapienza” University of Rome since 2002 (after Nigro, 2007c: fig. 1.3).

of fresh water flowing from a spring—and a large sacred building, the “Temple of the Kothon”, erected in the eighth century bc,13 and razed to the ground after the siege and the overall destruction brought about by Dionysius of Syracuse in 397/396 bc. After this dramatic event, the ruins of the temple were transformed in an open cult place called “Sanctuary C3”, which was used until the end of the fourth century bc (Figure 4.4). Ten seasons of excavations in Area C (2002–11) made it possible to reconstruct the architecture and stratigraphy of the Temple thoroughly. The earliest occupational layers (Phase 9, 770–750 bc) show votive pits and

089-106 WaterUrban Pt1 Ch04 Spagnoli.indd 92

16/05/2014 15:08

Phoenician Cities and Water93

Figure 4.3.  The temple area (Area C) located at the southwestern corner of Motya: view from west. Source: Rome “La Sapienza” University Expedition to Motya.

cult installations, associated with Phoenician Red Slip Ware (Nigro, 2012b: 8–10) and coarse handmade simple vessels like pots and trays; Greek imports are Corinthian skyphoi and kotylai, while Attic pottery is not yet attested (Spagnoli, 2012: 18–23). In the following phase (Phase 8, 750–650 bc), the earliest Temple (C5) is built up with almost the same layout it will also preserve in the following reconstructions (Temple C1 and C2). In its original layout, the Temple of the Kothon adopted the so-called “Four-room Building” plan developed from a Levantine tradition (Shiloh, 1970; Ottosson, 1980: 66–71; Wright, 1985: 275–80; Nigro, 1994: 203–91, 436–52; Sharon and Zarzecki-Peleg, 2006), with a central courtyard with an obelisk and two other baetyls.14 It has a main cella on its northern side, ending with an adyton or sancta sanctorum including a platform or shallow altar, a small stele behind it, and a circular depression flanked by a couple of big

089-106 WaterUrban Pt1 Ch04 Spagnoli.indd 93

16/05/2014 15:08

94A History of Water

Figure 4.4.  The Temple of the Kothon: view from west. Source: Rome “La Sapienza” University Expedition to Motya.

greyish mud bricks used as eschara—an installation for burning offerings and especially perfumes or incense. Beside the altar, there was a funnelshaped sinkhole (mundus), devoted to sacred libations. Such holes and the sacred well suggest that an underground deity was worshipped in the temple, such as Baal ‘Addir, god of the underwater and netherworld (Nigro, 2012b: 48).15 The main entrance of the temple was on the southern side, facing the South Gate and the city walls. In its first monumental reconstruction, in the middle of the sixth century bc (Temple C1), the entrance shows some typical Phoenician architectural features. The passage was framed by two semi-columns, each surmounted by a Proto-Aeolic order capital,16 and two freestanding pillars in front of them.17 Another important feature, located north of the temple, is the circular favissa F.864 (Nigro, 2010: 18–24, 35–7), holding the cult installations, such as stelae and baetyls, which discharged from the temple itself after the destruction of the building by the Greeks in 397/396 bc. This deposit seals up the area where a flow of fresh water sprang out from the soil (Figure 4.5). In autumn and winter rainfalls are high and the spring emerges like a little lake higher than in the dry season. In the same way during the last occupation of the area, in the fourth century bc, the well was sealed with a pillar base, according to the cult practice of defunctionalization of sacred devices, such as wells,

089-106 WaterUrban Pt1 Ch04 Spagnoli.indd 94

16/05/2014 15:08

Phoenician Cities and Water95

Figure 4.5.  The fresh water spring on the northern side of the Temple. In the foreground are the bamboo canes: view from northwest (after Nigro, 2012b: fig. 11).

pits, and favissae. This operation involves filling them with broken pottery, cult objects, and sand, and covering them with stones and architectural elements. This ritual action is usually associated with an animal sacrifice (often a dog). The Temple of the Kothon and the role of sacred areas into the urban arrangement The study of the Temple of the Kothon in its different architectural phases, compared to the other cult places of the island, indicates the role of the sacred areas during the urbanization process and the developing of the city. Reconsidering the architectural phases of the temple, it seems that, from its first construction, it significantly influenced the urban plan of the city. Before the arrival of the Phoenicians, the southwestern sector of the island of Motya looked like a gentle slope leaning toward the coast; 30 m from the seashore, a large natural depression hosted a small lake fed on a natural fresh water spring. The Phoenicians settled in that part

089-106 WaterUrban Pt1 Ch04 Spagnoli.indd 95

16/05/2014 15:08

96A History of Water

of the island, which had previously been occupied by an indigenous community,18 and cut the slope to level off the surface and to capture the water in numerous wells, exploiting the impermeability of the soil of this part of the island. The site of the natural lake was also modified: a rock-cut basin collecting water from the spring and a temple linked to it by a complex system of underground channels were built as the focal point of the rising city. On the other side of the island, to the north and at the nearest point to the coast of Sicily, another sacred building, the Sanctuary of “Cappiddazzu”, was erected a little later. The first sacred use of the area is attested from the end of the eighth century bc by several ritual favissae, where offerings and votive objects were found. Just as for the Temple of the Kothon, the Sanctuary of “Cappiddazzu” also knows two main monumental phases: the sixth and the fifth centuries bc, and a significant rebuilding after 397/ 396 bc.19 The “Cappiddazzu” and the Temple of the Kothon are the twin sacred poles of the island, both a peripheral and liminal place in respect of the residential quarters located at the center of the island. The position of those religious poles probably influenced also the arrangement of the two principal city gates,20 which were connected by an important street that intercepts both sacred areas crossing the island north–south. The Kothon, a sacred basin connected to the Temple Archaeological evidence allows us to reconsider the function of the Kothon itself. This can now be interpreted as a sacred pool instead of a dry-dock, as believed in the past. The survey of the Kothon confirmed that the built-up basin was completely enclosed by a continuous wall made of ashlar blocks. The southern wall of the basin, in its lower courses, also belongs to the original unitary ashlar structure enclosing the pool on all sides. In Late Roman Period the southern wall was partially dismantled in order to connect the basin to the sea by a channel to use it as a fishpond. Further on, in the twelfth century ad the Kothon was used as a “salina”, a salt-producing device up to the eighteenth century ad (as already noticed by Isserlin, 1971: 185). During the 2005 season, the emptying of the basin21 revealed the presence of a flow of fresh water from the northern wall of the pool, where a series of blocks protrude from the edge of the perimeter wall.22 This device proved to be the structural element through which fresh water flowed into the pool. A basic element required to understand the water systems linking the basin to the temple was obtained by geological investigations and paleo-environmental studies in the Marsala Lagoon23 where Motya lies. In antiquity, in fact, the sea level was 0.8–1 m lower,24 allowing fresh water, present in the underground marl strata, to erupt. Moreover, geological investigations demonstrated that the sacred well in the central

089-106 WaterUrban Pt1 Ch04 Spagnoli.indd 96

16/05/2014 15:08

Phoenician Cities and Water97

cult space of the temple received fresh water from the same source (Nigro, 2009a: 552, figs 306–7, 319). The Kothon and the temple were thus connected by an underground system, which can be easily related to classic ideological conceptions that the Phoenicians held. The sacred pool and the sacred well were both communicating directly with the world of underground waters. The Levantine character of the Temple of the Kothon and the role of water in Phoenician culture Recent discoveries at the Temple of the Kothon allow us to understand better the significance and the importance of this sacred building linked to the sacred basin, especially in the light of the religious architectural tradition to which this Phoenician temple is ascribable (Nigro, 2009c). The deep relationship that links Mediterranean cult places and water sources, especially in the Phoenician homeland, descends from one of the most typical Near Eastern religious conceptions. Underworld water in the Levant, as well as in earlier Mesopotamia, is always the water from which the world had its beginning in the Creation, and it is from the same water that, by a divine act, human civilization emerged. The presence of such water thus gives to a place the status of sacred space and, at the same time, makes it suitable for human settlement, therefore enabling the foundation of a city and the seat of the temple (Nigro, 2012a: 303). The major Phoenician cities and their main sanctuaries arose in direct connection with important water sources (Tyre with Ras el ‘Ain, Sidon with Nahar al-Awali and Bostan esh-Sheikh, Arwad with Amrit and the source of Naba’ el-Tell). Byblos, in particular, since the earliest origins of the settlement,25 had been focused on the central source and the nearby “sacred lake”, located in between the two major temples of the city: Balaat Gebal and the so-called “Temple en L”, successively reconstructed as the Obelisks Temple (Dunand, 1950–8: 644–52, fig. 767; Finkbeiner, 1981; Saghieh, 1983: 14–25; Nigro, 2012a: 304). The Levantine character of the Temple of the Kothon is further demonstrated by comparing it with some illustrious Phoenician sanctuaries. The obelisk in the central court of the temple recalls the earlier Temple of the Obelisks in Byblos, which comprised a sacred well and various alignments of obelisks and baetyls possibly related to libation activities. These vertical elements were discarded and accumulated in a favissa in Byblos (as in the Temple of the Kothon). The Phoenician temple that is most strikingly similar to the Temple of the Kothon from an architectural point of view is Ma’abed of Amrit, ancient Marathos, in Syria.26 Some general correspondences between the two religious complexes can be observed: both temples were erected in the sixth century bc, and they were both created by cutting and modeling bedrock. They were both built of ashlar structures; they are

089-106 WaterUrban Pt1 Ch04 Spagnoli.indd 97

16/05/2014 15:08

98A History of Water

both centered on a rectangular pool with the corners oriented according to the cardinal points (Dunand and Saliby, 1985: fig. 2). Finally, in both cases the pool was connected with a water source and some other structures. The Ma’abed is connected with the spring of Naba’ el-Tell through a channel: this channel led the waters into a cave, on the eastern side of the basin, and from the latter into the pool. Another branch of the channel runs along the eastern and southern edges of the pool, possibly to be used in ritual acts as libations; this channel may explain why that connecting the sacred well to the Kothon at Motya emerged on the eastern quay of the latter. At the middle of its northern side a structure made of blocks protrudes towards the pool, probably serving as a dock for the boat of the priests to reach the shrine in the center of the basin (Nigro, 2012a: 306). The similar protruding structures visible on the northern side of the Kothon at Motya was never fully excavated, and it was, thus, not possible to establish its function (Isserlin, 1971: 179). Concluding remarks The foundation of the Temple of the Kothon occurred together with that of the city itself, in a favorable spot of the island, due to the presence of fresh water not far away from the sea-shore serving as a dockland. For the Phoenician colonies, the availability of fresh water was a basic factor for the location of the new settlement, together with other typical features of the Phoenician landscape, such as coastal lagoons, spurs overlooking a bay, river mouths, and so on. The Marsala Lagoon is actually a very particular aquatic ecosystem where, in its southern side,27 the high-saline waters meet the sea. This leads to a propitious environment for the breeding of several ichthyic species and shellfish, such as mullet and lobsters. A fishy sea and the presence of the fresh water in the island favored the proliferation of wild animals (ducks and other birds, but also ungulates, such as deer) that had a primary role in the local diet.28 Similar environmental features are also attested in Gades (modern Cádiz, in Spain), a western colony beyond the Pillars of Hercules, hosting two sacred compounds. Described in Greek and Latin texts, the Temple of Heracles-Melqart (also called “Herakleion”), in the south, and the Temple of Baal Hammon, in the north,29 were erected at the beginning of Phoenician occupation of the island.30 Gades is traditionally one of the most ancient colonies of Tyre (as Arrianus recalls: Arr., Anabasis, II, 16: 4), dating back to the beginning of the first millennium bc.31 The myth of its foundation is strictly related to the erection of the Temple of Herakles-Melqart, mentioned by classical authors. According to Strabo (III, 5: 7–8) and Silius Italicus (Sil. Ital., Punica, III, 29–31), a Phoenician cult of a deity represented in an aniconic way was worshipped in the temple: two baetyls, three altars, and two fresh

089-106 WaterUrban Pt1 Ch04 Spagnoli.indd 98

16/05/2014 15:08

Phoenician Cities and Water99

water springs, a wellspring and a smaller temporary one, as in the Temple of the Kothon, were the foremost cult installations (Plácido, 1993: 75). The Temple of Heracles-Melqart at Gades rose in the southeastern part of the island in the place where, coming from the East, navigators could safely reach the coast.32 The similarities between the temple at Gades and the Temple of the Kothon, compared to the main cult places of Phoenicia, show the strong cultural unity of Syro-Palestinian, Canaanite, and Phoenician sacred architecture and religious beliefs connected to the presence of fresh water, and highlight the importance of the water as one of the principal reasons for the development of Phoenician sites along the coast of the Mediterranean Sea.

Notes   1 I would like to thank the Director of the Archaeological Mission to Motya of the “La Sapienza” University of Rome Professor Lorenzo Nigro, for giving me precious advice in the study of this topic.   2 The name “Phoenicians”, descending from the Greek word Phoinikes (plur. of “purple”. Textile industry is one of the most typical Phoenician handicraft), is employed for the first time in the Odyssey (Hom. Od., XIV, 288: δὴ τότε Φοῖνιξ ἦλθεν ἀνὴρ ἀπατήλια εἰδώς; XV, 415: ἔνθα δὲ Φοίνικες ναυσίκλυτοι ἤλυθον ἄνδρες) to indicate this population. Actually the Phoenicians called themselves “Sidonian”, from the name of one of the most powerful Phoenician cities, but also “Canaanites”. In spite of the lack of direct written sources, indirect sources testify the strong character and identity of this population. In the fifth century ad, St Augustine writes about farmers of the Hippo region, in Algeria: “Unde interrogati rustici nostri quid sint punici repondentes: Canani” (Augustine, Expositio ad Rom., 13).   3 Nigro (2010) with a summary of “La Sapienza” University of Rome excavations in the sacred area of the Kothon.  4 Kothon is a Greek word that means “ring-shaped cake”, and so it could be more appropriate for a circular basin.   5 The developing area of this culture encompasses the region including coastal Syria, Lebanon, and the north of Israel. It is delimited by the Orontes River to the north and the Anti-Lebanon Mountains to the south. The principal cities, which are actually city-states politically independent of each other, set up on the coast and controlled a wide part of the neighboring region at the back, which furnished alimentary goods, water, raw materials, and the workforce.  6 Emporion, a Greek word meaning “port of call, market”, is the first nucleus of a Phoenician colony. It is an independent political settlement, and not always has a mother city. The emporion is characterized by three principal features: the temple (usually consecrated to Melqart, the deity that, in the homeland, especially in Tyre, represents the power of the king of the city), the dock, and the storehouse (Aubet, 1997: 85–6). The emporion is the place where Phoenician merchants met native populations to obtain raw materials in exchange for various merchandise, often luxury goods.

089-106 WaterUrban Pt1 Ch04 Spagnoli.indd 99

16/05/2014 15:08

100A History of Water

  7 The most important and probably the first “colony” (in the Greek meaning of the word, implying an ideological and political reliance on the mother city) is Carthage, founded by Tyre in 814/813 bc, based on the written sources. In particular, Flavius Josephus in Against Apion (1.18) says that the city was founded during the seventh year of the reign of King Pygmalion (who ruled from 820 to 774 bc). His source is an earlier (and nowadays lost) work of Menander of Ephesus about the history of Tyre, with a list of the kings of Tyre from Abibaal to Pygmalion. This assertion follows the theory of Timeus of Taormine, which states that Carthage was founded by the Tyrians 38 years before the first Olympiad (performed in 776 bc).  8 A favissa is an underground reservoir or cellar near the temples, for water, for sacred utensils or votives no longer in use. The noun, descending from the Latin favissae, -a¯rum, corresponds to the Greek word θησαυροί.   9 After the destruction, the city continued for about two centuries, but it lost the strategic and economic role it had played before. From the following periods, there are sporadic traces of Roman and Byzantine occupation and a re-use of the ruins of the Sanctuary of “Cappiddazzu” (see infra) during the Norman Period. 10 Intensive agricultural works (especially grapevines), and a steady despoliation of the ruins in order to obtain building materials, have been carried out from Medieval times up to the beginning of the twentieth century. Those activities heavily damaged the stratigraphy and the architectural structures of the Phoenician city (Nigro, 2004c: 43; Nigro, 2005: 26–9; Nigro, 2007c, 14–18; Nigro, 2011: 27–30). 11 This area was formally investigated by Joseph Whitaker in 1906 by taking several soundings; findings were later deepened by the archaeological mission directed by Isserlin. The area has been occupied with residential and commercial/storage purposes since the seventh century bc. The area partially changed its plan after the erection of the city walls with the South Gate and probably the Circular Temenos of the Temple of the Kothon in the middle of the sixth century bc, and the structures were reused after the overall destruction of the city at the beginning of the fourth century bc (Whitaker, 1921: 145–78; Isserlin, 1962–63: 116–17; Nigro and Lisella, 2004: 81–2). 12 Archaeological explorations center on six different points of interest. On the western side of the coast (Area F), next to the West Gate and the adjacent large line of city walls, a defensive building (the “Western Fortress”) was found. This edifice was originally a military building, established for defensive purposes and contemporary to the erection of the city walls (mid-sixth century bc). After the destruction of Motya, brought about by Dionysius of Syracuse in 397/396 bc, it was reused as a luxury residency (excavation results were published in 2011 in the volume Mozia—XIII). In this later phase, a small shrine, sacred to Astarte, one of the principal Phoenician deities, was added south of the palace (Nigro, 2010). The southwestern and southeastern slopes of the Acropolis, at the center of the island (Area D and Area B, respectively), hosted two important residential quarters. These areas were occupied before the incoming of the Phoenician colonists, at least since the Middle Bronze Period (fourteenth century bc), as shown by the stratigraphic soundings in Area D (Caltabiano, 2007: 105–9; Caltabiano and Spagnoli, 2010). These areas were so early occupied because of their favorable position,

089-106 WaterUrban Pt1 Ch04 Spagnoli.indd 100

16/05/2014 15:08

Phoenician Cities and Water101

which allowed a wide control of the sea, and because of the presence of drinking water drawn from deep wells. In the last phases of the Phoenician occupation, coinciding with the second half of the fifth century bc until the destruction in 397/396 bc, areas B and D were occupied by three patrician residences, the “House of the Domestic Shrine” and the “House of the Horn of Triton” in Area D, and the “House of the Square Well” in Area B, where ivory furniture, bronze jewellery and tools, attic black-painted and red-figured pottery, and stone and terracotta small sculptures were found. The Tofet, an open cult place at the northwestern side of the island, in use from the end of the eighth century to the end of the fourth century bc, is characterized by the deposition of urns containing buried remains of infants and children, commemorated by stelae or gravestones. The sanctuary was excavated by Antonia Ciasca (“La Sapienza” University of Rome) in several campaigns carried out from 1964 to 1993. Ciasca gave a complete stratigraphic sequence of the deposition inside the sanctuary and a plausible hypothesis of reconstruction of the compound, which is articulated in two principal architectural phases. The first one is from the end of the eighth century to the middle of the sixth century bc, and the second phase is from the latter half of the sixth century to the fourth century bc. Since 2009, during the investigations of the southern perimeter wall carried out by the Archaeological Mission to Motya of “La Sapienza” University, a series of rectangular cult-rooms, erected during the rearrangement of the sacred area after the destruction at the beginning of the fourth century bc, was brought to light. Renewed field activities also interested the Necropolis area, where six archaic tombs dating back to the beginning of the seventh century bc, were brought to light. The last area of investigation is Area C, located in the southern part of the island. This includes the South Gate Area, with the artificial basin called “Kothon” and the sacred building east of it (Nigro, 2004a: 19–32). 13 The Temenos encircled the whole sacred area, including the Temple of the Kothon, marking its major reconstruction in Phase 5 in the mid-sixth century bc. In the same phase, Shrine C6-C4 and Sanctuary C7 were erected west of the sacred pool. 14 Cult furnishings recall similar installations in the renowned Temple of the Obelisks at Byblos (Dunand, 1950–8: 644–52, fig. 767). The erection of stelae and baetyls in cult places is largely documented in Syria and Palestine during the Bronze and Iron Ages. This religious practice lasts in use until the Roman Period, as shown by a coin of Macrinus (ad 217), representing a temple with a sacred precinct and an obelisk/baetyl in the middle (Jidejian, 1968: fig. 121). 15 Baal ‘Addir, attested at Bylos in the sixth century bc with the epithet “Mighty Lord”, has in the Phoenician Pantheon a chthonian and agrarian character. In later periods, especially in the Punic regions of North Africa, he is considered similar to Baal Hammon. Baal Hammon and Baal ‘Addir are both versions of Baal, one of the most important deities of the Semitic Pantheon. Baal, son of Dagan, the god of crop fertility and vegetation in the third and second millennia bc, is the god of the storm, the rain, and the dew, and also the god of life and fertility. In Semitic mythology, he engaged in mortal combat with Mot, the god of death and sterility, and then came back to life as the god of vegetation (Lipi´nsky, 1992: 55–6, 126).

089-106 WaterUrban Pt1 Ch04 Spagnoli.indd 101

16/05/2014 15:08

102A History of Water

16 One of those capitals was found thrown into the sacred well at the center of the temple, while its base was reused and set up on the closed top of the well at the time of Sanctuary C3, in the fourth century bc. 17 Two bronze pillars, which had been made by Tyrian metallurgists, also flanked the entrance of Solomon’s Temple in Jerusalem, according to the Biblical account (1 King 7:15–22; 2 Chronicles 3:15–17; Busink, 1970: 299–321). 18 The indigenous settlement of the Middle and Late Bronze Ages (fifteenth to eleventh centuries bc) is testified in this area only by potsherds and objects. No architectural structures were found, probably because of the later rearrangement of the area, carried out by the Phoenicians in the middle of the eighth century bc. 19 Architectural phases of the Sanctuary of “Cappiddazzu” are discussed in Tusa (2000), Nigro and Spagnoli (2004), Nigro (2009b: 243–51). 20 The distance between the South Gate and the Temple of the Kothon is on average 30 m, while the Sanctuary of “Cappiddazzu” is 100 m from the North Gate. 21 The Kothon was partly emptied by J. Whitaker in its western part, while the perimeter structure, in the corners and in some spots of the northern and southern sides, was explored by the British Expedition directed by B. S. J. Isserlin (1971: 184–6). 22 The protruding structure has been interpreted in the past as an inner berth (Isserlin, 1970: 565; 1971: 185, pl. XXIXb; Famà, 1995: 178; Tusa, 2004: 448). 23 The presence of numerous fish and birds of various kinds concentrating in the Kothon and in its immediate neighborhood, as already observed by Whitaker (1921: 190), indicated the presence of a flow of fresh water. 24 As testified by the nowadays submerged quays and docks all around the island of Motya (Isserlin, 1971: 179; Tusa, 2004: 450, fig. 9; Caltabiano, 2011: 447–8). 25 The earliest sacred building at Byblos (the Enceinte Sacrée) arose just aside the central well by the end of Early Bronze IA (around 3300 bc) (Dunand, 1973: 235–41, fig. 143, pl. J,c; Dunand, 1982: 195; Nigro, 2007a: 1–3, 26–31; Sala, 2007: 48–58). 26 The similarities between the Kothon of Motya and the Ma’abed of Amrit were stressed for the first time by P. Mingazzini (Mingazzini, 1968: 105–12). 27 In antiquity, Isola Lunga was joined to the coast, closing the northern side of the lagoon because of the lower level of the sea. 28 As in the Levant, the island city obtained any foodstuffs from the indigenous communities, such as the Sicanians and the Elymes settled on the fertile hills of oriental Sicily (Falsone, 1988: 43–5). Those populations were able to offer a wide range of agricultural products in a favorable exchange system (Tusa and Morris, 2004). 29 The temple is mentioned by Strabo (III: 5, 3) and Pliny the Elder (Nat. Hist. 4. 120). A Proto-Aeolic capital found under the sea, west of the city, is the unique architectural remains of the building (Aubet, 1997: 236–7). 30 In ancient times, the archipelago of Gades was composed of three islands: Erytheia, where the Phoenician city was built up, to the north; the island of San Fernando to the west; Kotinoussa, a long and narrow island ending in two promontories. The Temple of Herakles-Melqart was founded south of it (Plinius, Nat. Hist., 4: 22). Nowadays the modern city lies at the end of the promontory that closes at the west of the Gulf of Cadiz. The geography of

089-106 WaterUrban Pt1 Ch04 Spagnoli.indd 102

16/05/2014 15:08

Phoenician Cities and Water103

that area was modified by the alluvium of the Guadalete River flowing into the Gulf. 31 M. Velleius Paterculus asserts the city was founded 80 years after the War of Troy, in 1104/1103 bc (Hist. Rom., 1: 2, 1–3). 32 A good example is the usual locations of Melqart’s and Astarte’s temples in the Mediterranean (Bernardini, 2003: 112–19).

references Arrianus, L. Flavius (2010). Anabasis Alexandri, edited by A. G. Roos, Florence. Aubet, M. E. (1997). Tiro y las Colonias Fenicias de Occidente, Bellaterra, Ediciones S.A., Barcelona. Bernardini, P. (2003). “I Fenici ai confini del mondo: le isole erranti e le colonne di Melqart”, in Sardinia, Corsica et Baleares Antiquae, I, pp. 112–21. Busnik, Th. A. (1970). “Der Tempel von Jerusalem von Salomo bis Herodes: Eine archäologish-historische Studie unter Berücksichtigung des westsemitischen Tempelbaus 1: Der Tempel Salomos”, Leiden: Studia Francisci Scholten Memoriae Dedicata 3. Caltabiano, A. (2007). “La ceramica preistorica dal Sondaggio I”, in L. Nigro (ed.), Mozia—XII, pp. 105–9, Missione Archeologica a Mozia, Rome. ——— (2011). “Gli approdi di Porta Sud e Porta Ovest a Mozia”, in L. Nigro (ed.), Mozia—XIII, pp. 441–56, Missione Archeologica a Mozia, Rome. Caltabiano, A. and F. Spagnoli (2010). “Mozia: La ceramica fenicia arcaica dal sondaggio stratigrafico III nella Zona D”, in L. Nigro (ed.), Motya and the Phoenician Ceramic Repertoire Between the Levant and the West, 9th–6th Century bc. Proceedings of the International Conference held in Rome, 26th February 2010, Quaderni di Archeologia Fenicio-Punica, V, pp. 117–49, Missione Archeologica a Mozia, Rome. Dunand, M. (1950–8). Fouilles de Byblos. Tome II, Paul Geuthner, Paris. ——— (1973). Fouilles de Byblos. Tome V. L’architecture, les tombes, le matériel domestique, des origines néolithiques à l’avènement urbain, Paul Geuthner, Paris. ——— (1982). “Byblos et ses Temples après la pénétration amorite”, in H.-J. Nissen and J. Renger (eds), Mesopotamien und Seine Nachbarn. Politische und kulturelle Wechselbeziehungen im Alten Vorderasien vom 4. bis 1. Jahrtausend v. Chr. XXIV. Rencontre Assyriologique Internationale Berlin 3. bis 7. Juli 1978, pp. 195–201, Dietrich Reimer Verlag, Berlin. Dunand, M. and N. Saliby (1985). Le Temple d’Amrith dans la Pérée d’Aradus, BAH CXXI. Paris. Falsone, G. (1988). “The Bronze Age occupation and Phoenician foundation at Motya”, Bulletin of the Institute of Archaeology London, 25, pp. 31–53. Famà, M. L. (1995). “Il porto di Mozia”, Sicilia Archeologica, 87–9, pp. 171–9. Finkbeiner, U. (1981). “Untersuchungen zur Stratigraphie des Obeliskentempels in Byblos. Versuch einer methodischen Auswertung”, Baghdader Mitteilungen, 12, pp. 13–69. Isserlin, B. S. J. (1971). “New light on the Kothon at Motya”, Antiquity, 45, pp. 178–86.

089-106 WaterUrban Pt1 Ch04 Spagnoli.indd 103

16/05/2014 15:08

104A History of Water

Isserlin, B. S. J. and J. du Plat. Taylor (1974). Motya. A Phoenician and Carthaginian City in Sicily. A report of the excavations undertaken during the years 1961–65 on behalf of the University of Leeds, the Institute of Archaeology of London University and Fairleigh Dickinson University, New Jersey, vol. I. Field Work and Excavation, E. J. Brill, Leiden. Jidejian, N. (1968). Byblos through the Ages, Dar el-Machreq Publishers [distribution Librairie Orientale], Beirut. Lipi´nsky, E. (1992). Dictionnaire de la Civilisation Phénicienne et Punique, Brepols, Paris. Mingazzini, P. (1968). “Scopo e natura del cosiddetto Kothon di Mozia”, in A. Ciasca, G. Garbini, P. Mingazzini, B. Pugliese, and V. Tusa. Mozia—IV. Rapporto preliminare della Missione archeologica della Soprintendenza alle Antichità della Sicilia Occidentale e dell’Università di Roma, Studi Semitici, 29, pp. 105–12, Istituto di Studi del Vicino Oriente, Università di Roma, Rome. Nigro, L. (1994). Ricerche sull’architettura palaziale della Palestina nelle Età del Bronzo e del Ferro. Contesto archeologico e sviluppo storico, Contributi e Materiali di Archeologia Orientale V, Università degli studi di Roma “La Sapienza”, Rome. ——— (2004a). Mozia—X. Rapporto preliminare della XXII campagna di scavi—2002 condotta congiuntamente con il Servizio Beni Archeologici della Soprintendenza Regionale per i Beni Culturali e Ambientali di Trapani, Quaderni di Archeologia Fenicio-Punica, I, Università degli studi di Roma “La Sapienza”, Rome. ——— (2004b). “Gli scavi di Antonia Ciasca al Tofet e alle mura (1964–1993)”, in L. Nigro and G. Rossoni (eds), “La Sapienza” a Mozia. Quarant’anni di ricerca archeologica (1964–2004), pp. 20–9, Università degli studi di Roma “La Sapienza”, Rome. ——— (ed.) (2005). Mozia—XI. Il Tempio del Kothon. Rapporto preliminare delle campagne di scavo XXIII e XXIV (2003–2004) condotte congiuntamente con il Servizio Beni Archeologici della Soprintendenza Regionale per i Beni Culturali e Ambientali di Trapani, Quaderni di Archeologia Fenicio-Punica, II, Università degli studi di Roma “La Sapienza”, Rome. ——— (2007a). “Aside the spring: Byblos and Jericho from village to town in the second half of the 4th millennium bc”, in L. Nigro (ed.), Byblos and Jericho in the Early Bronze I: Social Dynamics and Cultural Interactions. Proceedings of the International Workshop held in Rome on March 7th 2006 by Rome “La Sapienza” University, Rome “La Sapienza” Studies on the Archaeology of Palestine & Transjordan, 4, pp. 1–45, Università degli studi di Roma “La Sapienza”, Rome. ——— (2007b). “Il lago sacro e l’obelisco”, Kalós, 19, pp. 24–9. ——— (ed.) (2007c). Mozia—XII. Zona D. La “Casa del sacello domestico”, il “Basamento meridionale” e il Sondaggio stratigrafico I. Rapporto preliminare delle campagne di scavi XXIII e XXIV (2003–2004) condotte congiuntamente con il Servizio Beni Archeologici della Soprintendenza Regionale per i Beni Culturali e Ambientali di Trapani, Quaderni di Archeologia Fenicio-Punica, III, Università degli studi di Roma “La Sapienza”, Rome. ——— (2009a). “Recenti scoperte dell’Università di Roma ‘La Sapienza’ a Mozia (2002–2006): il Tempio del Kothon, la ‘Casa del sacello domestico’, il ‘Basamento meridionale’ e la Fortezza Occidentale”, in C. Ampolo (ed.),

089-106 WaterUrban Pt1 Ch04 Spagnoli.indd 104

16/05/2014 15:08

Phoenician Cities and Water105

Immagine e immagini della Sicilia e di altre isole del Mediterraneo antico, Vol. II. Atti delle seste giornate internazionali di studi sull’area elima e la Sicilia occidentale nel contesto mediterraneo. Erice 12–16 ottobre 2006, pp. 551–9, Università degli studi di Roma “La Sapienza”, Pisa. ——— (2009b). “Il Tempio del Kothon e il ruolo delle aree sacre nello sviluppo urbano diMozia dall’VIII al IV secolo a.C.”, in S. Helas and D. Marzoli (eds), Phönizisch und punisches Städtewesen. Akten der internationalen Tagung in Rom vom 21. bis 23. Februar 2007, Iberia Archaeologica Band 13, pp. 241–70, Università degli studi di Roma “La Sapienza”, Mainz am Rhein. ——— (2009c). “Il Tempio del Kothon e le origini fenicie di Mozia”, in A. Mastino, P. G. Spano, and R. Zucca (eds), Naves Plenis Velis Euntes, Tharros Felix 3, pp. 77–118, Università degli studi di Roma “La Sapienza”, Rome. ——— (2010). “Alle origini di Mozia: stratigrafia e ceramica del Tempio del Kothon dall’VIII al VI secolo a.C.”, in L. Nigro (ed.), Motya and the Phoenician Ceramic Repertoire between the Levant and the West, 9th–6th Century bc. Proceedings of the International Conference held in Rome, 26th February 2010, Quaderni di Archeologia Fenicio-Punica, V, pp. 1–48, Università degli studi di Roma “La Sapienza”, Rome. ——— (ed.) (2011). Mozia—XIII. Zona F. La Porta Ovest e la Fortezza Occidentale. Rapporto preliminare delle campagne di scavi XXIII–XXVII (2003–2007) condotte congiuntamente con il Servizio Beni Archeologici della Soprintendenza Regionale per i Beni Culturali e Ambientali di Trapani, Quaderni di Archeologia Fenicio-Punica, VI, Università degli studi di Roma “La Sapienza”, Rome. ——— (2012a). “The Temple of the Kothon at Motya, Sicily: Phoenician religious architecture from the Levant to the West”, in M. Gruber, S. Ahituv, G. Lehmann, and Z. Talshir (eds), All the Wisdom of the East. Studies in Near Eastern Archaeology and History in Honor of Eliezer D. Oren, Orbis Biblicus et Orientalis, 255, pp. 293–331, Università degli studi di Roma “La Sapienza”, Friburg. ——— (2012b). “La favissa di Baal ‘Addir”, in L. Nigro and F. Spagnoli, Alle sorgenti del Kothon. Il rito a Mozia nell’Area sacra di Baal ‘Addir—Poseidon. Lo scavo dei pozzi sacri nel Settore C Sud-Ovest (2006–2011), Quaderni di Archeologia Fenicio-Punica/CM 02, pp. 8–12, Università degli studi di Roma “La Sapienza”, Rome. Nigro, L. and A. R. Lisella (2004). “Il Quartiere di Porta Sud”, in L. Nigro and G. Rossoni (eds), “La Sapienza” a Mozia. Quarant’anni di ricerca archeologica, 1964–2004. Catalogo della mostra, Università di Roma «La Sapienza», Facoltà di Scienze Umanistiche, Museo dell’Arte Classica, 27 febbraio–18 maggio 2004, pp. 78–83, Università degli studi di Roma “La Sapienza”, Rome. Nigro, L. and F. Spagnoli (2004). “Il Santuario del ‘Cappiddazzu’”, in L. Nigro and G. Rossoni (eds), “La Sapienza” a Mozia. Quarant’anni di ricerca archeologica, 1964–2004. Catalogo della mostra, Università di Roma “La Sapienza”, Facoltà di Scienze Umanistiche, Museo dell’Arte Classica, 27 febbraio–18 maggio 2004, pp. 56–61, Università degli studi di Roma “La Sapienza”, Rome. Ottosson, M. (1980). Temples and Cult Palaces in Palestine, Acta Universistatis Upsaliensis, Boreas, 12, Vandenhoek and Ruprecht Göttingen, Uppsala. Paterculus, C. Velleius (2010). Historiae Romanae, Libri Duo, Notis Adiectis, Nabu Press, Florence.

089-106 WaterUrban Pt1 Ch04 Spagnoli.indd 105

16/05/2014 15:08

106A History of Water

Plácido, D. (1993). “Le vie di Ercole nell’estremo Occidente”, in A. Mastrocinque, Ercole in Occidente, Labirinti. Collana del Dipartimento di Scienze Filologiche e Storiche, 2, pp. 63–80, Università degli Studi di Trento, Trento. Pliny the Elder, Naturalis Historia, edited by K. F. Th. Mayhoff, Perseus Digital Library, Tufts University, G. R. Crane editor-in-chief, www.perseus.tufts.edu Saghieh, M. (1983). Byblos in the Third Millennium B.C.: A Reconstruction of the Stratigraphy and a Study of the Cultural Connections, Aris & Phillips, Warminster. Sala, M. (2007). “Early shrines at Byblos and Tell es-Sultan/ancient Jericho in the Early Bronze I (3300–3000 bc)”, in L. Nigro (ed.), Byblos and Jericho in the Early Bronze I: Social Dynamics and Cultural Interactions. Proceedings of the International Workshop held in Rome on March 7th 2006. by Rome “La Sapienza” University. Rome “La Sapienza” Studies on the Archaeology of Palestine & Transjordan, 4, pp. 47–68. Sharon, I. and A. Zarzecki-Peleg (2006). “Podium structures with lateral access: authority ploys in royal architecture in the Iron Age Levant”, in S. Gitin, J. E. Wright, and J. P. Dessel (eds), Confronting the Past. Archaeological and Historical Essays on Ancient Israel in Honor of William G. Dever, pp. 145–67, Eisenbrauns, Winona Lake, IN. Shiloh, Y. (1970). “The four-room house—its situation and function in the Israelite city”, Israel Exploration Journal, 20, pp. 180–90. Silius Italicus, Punica, edited by W. C. Summers and J. P. Postgate, Perseus Digital Library, Tufts University, G. R. Crane editor-in-chief, www.perseus.tufts.edu Spagnoli, F. (2012). “Ceramica dei primi utilizzi della favissa F.2950”, in L. Nigro and F. Spagnoli, Alle sorgenti del Kothon. Il rito a Mozia nell’Area sacra di Baal ‘Addir—Poseidon. Lo scavo dei pozzi sacri nel Settore C Sud-Ovest (2006–2011), Quaderni di Archeologia Fenicio-Punica/CM 02, pp. 18–23, Missione Archeologica a Mozia, Rome. Strabo, Geography, Perseus Digital Library, G. R. Crane editor-in-chief, www. perseus.tufts.edu, Tufts University. Tusa, S. (2004). “Il sistema portuale di Mozia. Il Kothon”, in L. Nigro (ed.), Mozia—X, pp. 445–64, Missione Archeologica a Mozia, Rome. Tusa, S. and I. Morris (2004). “Scavi sull’acropoli di Monte Polizzo, 2000–2003”, Sicilia Archeologica, 102, pp. 35–90. Whitaker, J. I. S. (1921). Motya, a Phoenician Colony in Sicily, G. Bell and Sons, London. Wright, G. R. H. (1985). Ancient Building in South Syria and Palestine, Handbuch der Orientalistik 7, E. J. Brill, Leiden-Köln.

089-106 WaterUrban Pt1 Ch04 Spagnoli.indd 106

16/05/2014 15:08


Waters at Babylon

Olof Pedersén Introduction For some 2,500 years Babylon has often been taken as the archetypal ancient example of a leading metropolis. Classical, biblical, renaissance, and modern traditions about Babylon, situated at the Euphrates on the flood plain in lowland Mesopotamia, have influenced world literature and visual arts. Large-scale excavations and a huge amount of cuneiform texts found at Babylon have contributed to a modern understanding also shown in large exhibitions, like the one in Berlin 2008 with more than half a million visitors (Marzahn and Schauerte, 2008). On the spot in the remains of Babylon, both the 2003 allied military camp and the expansion of modern buildings and infrastructure have created problems, as has also the water situation until modern times. The complex historical water management at Babylon will be discussed from different comparative and historical perspectives, indicating both present limitations in our understanding and possibilities for future research. First, the larger perspectives are applied in the section dealing with the Mesopotamian flood plain, surveying both the modern and the ancient situations. Then the surroundings of Babylon is discussed in some more detail. The final section gives a somewhat fuller treatment of selected aspects of waters in the city of Babylon itself. The last two sections focus especially on the period when Babylon was the main metropolis during the reign of Nebuchadnezzar II, in the first half of the sixth century bc. This preliminary survey is part of a re-examination of the city of Babylon. In order to get a better understanding of various aspects of this enormous ancient city. Several pilot projects have been attempted; one of them presented here deals with an overview of the water situation at Babylon. The basic sources used consist of archaeological materials from the excavations in Babylon, including the cuneiform texts unearthed at the site as well as other ancient texts. Satellite images and old maps, as well as geological and hydrological investigations, have also been used as sources (cf. maps). Within the project, the mapping called ANE.kmz of sites in

107-129 WaterUrban Pt1 Ch05 Pedersen.indd 107

12/05/2014 11:14


A History of Water

the Ancient Near East on Google Earth has been further developed, and a preliminary version of a digital model of Babylon has been integrated with the material in order to allow new interpretation attempts (cf. digital references). Mesopotamian flood plain Despite low and, during most historic periods, insufficient precipitation, the Mesopotamian flood plain in the southern half of modern Iraq has been the home for some of the oldest human cultures based on irrigation from the Euphrates and the Tigris, with towns and cities appearing before 5000 bc. These include the Sumerian, Akkadian, Babylonian, Aramaic, and Arabic-speaking cultures, arranged under a number of different political systems. The flood plain is the low area in the southern part of the Iraqi central depression between the Arabian Plate and the Turkish and Iranian Plates. It has been built up of river deposits from the Quaternary, consisting of a heavy Pleistocene period deposit covered by a Holocene period deposit. The Holocene deposits, dated from the last 10,000 years or so, have often a depth up to some 10–20 m. The complete Quaternary deposits making up the flood plain are some 50–250 m deep, with more depth in the eastern than in the western part of the plain. A number of borings and examinations of deep wells, especially in the 1980s, have contributed to the knowledge of the geological circumstances. Rivers and river arms have partly been distinguished in several levels of the flood deposits, but their chronological distribution has still to be better established (Aqrawi et al., 2006; Al-Jiburi and Al-Basrawi, 2011; Yacoub, 2011). The Mesopotamian lowland flood plain comprises some 54,000 km2. It is watered by the southern sections of the Euphrates and the Tigris. The Euphrates, with a total length of 2,700 km (of which 1,213 km is in Iraq), has an average flow rate of only 1,100 m3/s. Almost all of the waters come from the Turkish mountains. The Tigris, with a length of 1,718 km (of which 1,418 km is in Iraq), has a considerably larger mean runoff of 4,000 m3/s. Approximately half of the Tigris water is from the Turkish mountains; the rest is from Iranian and Iraqi tributaries. In flood periods in spring, the Euphrates can reach 50,000 m3/s and the Tigris 80,000 m3/s (Verhoeven, 1998; Partow, 2001; Krásný et al., 2006). Due to the extremely low precipitation, humans on the Mesopotamian flood plain are dependent on water from the rivers and on systematic irrigation. After a wetter Early and Middle Holocene, the climate during the periods of interest here was variable but drier (e.g. Wilkinson, 2012), comparable with the present situation, having less than 200 mm rain per year. Most water in the rivers flows from the mountain areas; the addition from precipitation, except from occasional heavy rainstorms, is of minor

107-129 WaterUrban Pt1 Ch05 Pedersen.indd 108

12/05/2014 11:14

Waters at Babylon109

importance. The plain is also the discharge zone of groundwater from all surrounding areas. The depth of the groundwater of the flood plain varies between 10 m below surface in the north and only about 1 m in the southeast. Heavy evaporation of the near-surface saline groundwater leads to salt accumulation in the soil. Typically, the salinity of the groundwater increases when moving in a southeastern direction on the plain and with depth. Groundwater near the rivers may sometimes be useable for irrigation purposes (salinity 1–5 g/l), but as drinking water (less than 1 g/l), the Euphrates is traditionally preferred (Krásný et al., 2006; Al-Jiburi and Al-Basrawi, 2011). From a historical perspective, the traditional Euphrates and Tigris water regime on the flood plain will interest us especially here (Figure 5.1). From the 1950s, the whole traditional water system, and thereby also much of the landscape, has been changed fundamentally. A Third River with networks of canals for drainage has been constructed over the last

Figure 5.1.  Babylon and surroundings. White, the fresh water system: Modern Euphrates and Tigris with main river arms and selected canals. Black, the drain system: Third River since the 1950s with selected drain canals. Source: Google Earth Pro, with additions.

107-129 WaterUrban Pt1 Ch05 Pedersen.indd 109

12/05/2014 11:14


A History of Water

60 years. According to the new, not yet fully implemented system, every single field on the flood plain should not only have the traditional water inlet canal on the uppermost side but also have a drain on the lowest side. The drain will eventually lead the drainage water to the Third River, which should bring it through its separate outlet into the sea. The prospects of getting salty soil cleaned with this drainage system were quite good when the project started, but modern dams reducing the water flow complicate the system. These dams have a far larger capacity than there is water in the rivers. The gross storage capacity of all existing modern hydraulic works (dams and reservoirs) on the Euphrates is five times the river’s average annual flow, and will lead to future problems (Lebon, 1964; Partow, 2001: 9). We now turn to the older system. A useful working tool in order to deal with the traditional system in future historical research is the declassified Corona satellite images from the 1960s to the 1970s. The Corona images of the Middle East at http:// corona.cast.uark.edu (still in beta) are in the process of being conveniently digitized, orthorectified, and put together in a layer over Google Earth images by the University of Arkansas. In addition to cuneiform texts, older maps (e.g. Selby and Collingwood, 1859; Selby et al., 1885) and descriptions of the landscape in classical, Arabic, and early western literature (e.g. Le Strange, 1895; Cadoux, 1906; Janssen, 1995) are also of great value for understanding the traditional water system. In the traditional system, annual floods of high but varied magnitude occurred usually in April–May with unpredictable inundations. The flood first appeared in the Tigris area and was then followed by the Euphrates area. The lowest discharge of the rivers happened in August–September. There was far more water than was needed or possible to use in the flood periods and too little in the low discharge periods, so irrigation and storage systems had to be constructed and maintained (Rzóska et al., 1980; Verhoeven, 1998; Bagg, 2003; Bagg, 2012). In the northern part of the flood plain, the rivers are meandering, whereas in the southern part, they show anastomosing patterns. On the flat flood plain, the rivers, always carrying sediment, deposit it and create levees higher than the surrounding landscape. Overflowing water deposits coarser materials next to the river and finer silt farther away from it, thereby building up the levees more than the surroundings. The result is a landscape with river channels and canals on levees with lower areas in between. Rivers and canals supplying the water needed for irrigation could also be used for boat transportation, especially when bringing heavy goods. Breaks in the levees can be used for irrigation, but can also create new river arms or even, in extreme situations, develop into a new course of the main river (Brown, 1997; Verhoeven, 1998; Wilkinson, 2012). The lower areas between the levees had a tendency to be swampy or marshy. A large number of marshes, either seasonal or permanent, were spread over the flood plain. The largest marshes were in the southeast,

107-129 WaterUrban Pt1 Ch05 Pedersen.indd 110

12/05/2014 11:14

Waters at Babylon111

where remains of the wetlands can still be found and where there are attempts to preserve, conserve, and return them to earlier sizes. This will probably only be possible for limited parts, because not enough water is available. A number of other marshes in the surroundings of Babylon will be discussed below. During the flood periods, uncontrolled inundations could cover large areas in the landscape. Many of the marshes served as water reservoirs storing excess water during the inundation periods. In modern times, this has instead been regulated with canals and lakes—for example, the Tharthar, Habbaniya, and Abu Dibbis lakes, and a number of river reservoirs (Cole, 1994; Partow, 2001; Eger, 2011). This modern development, similar to what has happened in many other modern countries, can be seen when the present situation is compared with maps from the nineteenth century (e.g. Selby and Collingwood, 1859; Selby et al., 1885) or even the Corona satellite images. Ancient levees and sites can often be seen on the modern surface, and such remains started to be surveyed in the 1950s. There have been a number of well-known surveys of the distribution of archaeological sites, especially on the central flood plain. The more recent have evolved into more detailed reconstructions of ancient watercourses (Adams, 1965; Adams, 1972; Adams and Nissen, 1972; Gibson, 1972; Adams, 1981; Gasche and Tanret, 1998). The recent surveys have better registered the remains of ancient levees and reconstructed the ancient watercourses according to these remains using ancient texts (Gasche and Tanret, 1998; Gasche et al., 2002). Other studies have been trying to reconcile the results of archaeological surveys with information from ancient texts (e.g. Steinkeller, 2001). The southwestern and southeastern sections of the flood plain have never been surveyed in detail, and therefore have obvious possibilities for future research. Remote sensing using satellite images, aerial photographs, and Digital Elevation Models (DEMs) has started to open up new possibilities, but has also given rise to new questions that can only be answered by “ground truth” in the form of surveys, excavations, and borings (Hritz, 2010). The main principle for reconstruction of the ancient landscape has been to establish a line of sites dated to the same period and stringed along remains of old levees. However, levees and sites can be buried in the landscape without visible traces on the surface, or be totally lost due to erosion or human activity. Levees are much more visible in the northern part of the flood plain, whereas they often disappear under the surface in the south, with some visible exceptions (especially in the east). Excavation or drilling is often required in order to get secure results when surface evidence is not compelling. However, very few excavations and drillings of ancient watercourses have been conducted. The most well known are the ones at Sippar and Nippur (Gasche and Tanret, 1998; Wilkinson, 2003). A number of specialized agricultural soil surveys have been conducted with the intention to develop the modern agricultural sector. The more

107-129 WaterUrban Pt1 Ch05 Pedersen.indd 111

12/05/2014 11:14


A History of Water

detailed of these surveys have also registered ancient levees and settlements. The soil analysis has dealt in depth with the well-known negative effect of irrigation, alluded to above, with the deposit of salty silt in the warm climate leading to increased salt content in the soil (Buringh, 1960; Buringh, 1961; Poyck, 1962; Wirth, 1962; Lebon, 1964). There are still many basic aspects of the historical water regime of the Mesopotamian flood plain that have to be studied in detail with modern methods. The present geological interpretations lack much of the wanted precision for reconstruction of the ancient water systems. Geomorphological and archaeological examinations of watercourses are represented by a few good examples, but are likewise only at the beginning. The promising use of remote sensing needs the ground truth in the form of surveys, excavations, and borings properly interpreted. There are many basic questions that may be answered in the future. Surroundings of Babylon The preliminary discussion of the surroundings of Babylon will focus especially on the time of Nebuchadnezzar II (604–562 bc) and the years following his reign. At that time, Babylon was the leading metropolis in the world. The area around Babylon may be delimited and divided in different ways, but essentially it consisted of a series of cities, Kish, Borsippa, Dilbad, and Kutha, within 15–30 km distance, as well as Sippar and Marad, 60 km away in an agricultural landscape in the northwestern part of the Mesopotamian flood plain, watered by means of the Euphrates and its river arms, and a large number of canals of various sizes. In this chapter, the designation “Euphrates” will be used also for river arms like Araktu found in some ancient texts (Figure 5.1). There has never been any modern archaeological survey of the area immediately around Babylon. However, the area east and northeast of Babylon was included in the pioneering preliminary Akkad survey from the 1950s, although this lacked many details (Adams, 1972). More information for the area around the eastern neighboring city Kish has been produced by a more detailed study (Gibson, 1972). The area west of Babylon was included within detailed agricultural surveys; however, these had quite a different focus than the archaeological ones. The identified sites in the agricultural studies seem hardly ever to have been noted by archaeologists (Buringh, 1960; Buringh, 1961; Poyck, 1962; Wirth, 1962). The most modern survey has been produced for the area of Sippar north of Babylon (Gasche and Tanret, 1998; Gasche et al., 2002). All these surveys have paid attention to ancient watercourses in addition to archaeological sites. Remains of levees were used for the reconstruction of main river arms and canals. Strings of archaeological sites datable to a specific period situated along a levee were taken as a strong indication

107-129 WaterUrban Pt1 Ch05 Pedersen.indd 112

12/05/2014 11:14

Waters at Babylon113

that the channel on the levee was used during the same period. Lack of evidence of sites along a levee has been interpreted as no longer a working channel. There is a tendency to place more weight on the actual remains of levees in the most recent works, and some 2 km northwest of Sippar, a 25 km auger transect was followed perpendicular to old levees. This, having a depth of c.7 m, is the only published transect in the larger Babylon area. However, further southeast, there are also in the Nippur and Abu Salabikh area c.5 m deep transects (Wilkinson, 1990; Gasche and Tanret, 1998; Wilkinson, 2003). The common agricultural units on the Mesopotamian flood plain are situated as a large number of slices along river arms and canals. The upper short side along the water can often be only some 20 m, but the long sides perpendicular to the short side along the water may extend to several 100 m from the water source. Along the water, there is the garden with palm trees (with 20 m corresponding to two rows of palm trees), followed on lower sections by fields often used for barley. In the lowest areas, there is more marginal land in use for pasture; further away, there may even be marshy or salty land. The principle is the same both according to ancient texts (e.g. Wunsch, 2000) and modern detailed maps and satellite images. The only difference is that in ancient times, the agricultural units could be swept over and cleaned by inundations, whereas in the modern system, there is a more permanently regulated water inlet and drainage. In depressions between the levees, either seasonal or permanent swamps and marshes could develop as referred to above. There were traditionally several marshy areas in the greater Babylon surroundings as seen on nineteenth-century maps (Selby and Collingwood, 1859; Julius, 1862; Selby et al., 1885), as well as on modern geological maps (cf. maps). The largest and more permanent wetlands lay southwest of Babylon, but marshy areas at least seasonally could also be found north, east, and south of the city. The situation was similar in ancient times (Cole, 1994; Partow, 2001; Eger, 2011). Nebuchadnezzar II proudly claimed that he had arranged for marshes around Babylon as defensive installations (Langdon, 1912; Reade, 2010). A main wall was reported in ancient texts to have been built by Nebuchadnezzar II between the Euphrates and the Tigris, in the north, in the area of Sippar. The western section of the wall, situated some 3–5 km north of the ancient Royal Canal (with an approximate alignment like that of the later Yusufiya Canal from the Euphrates in the direction of the Tigris, south of Baghdad), has been documented in excavations. A similar wall from Babylon to Kish and further to the Tigris, according to inscriptions by Nebuchadnezzar II, has still to be found (Gasche et al., 1987, 1989; Gasche, 2010). It has not yet been established if traces of this wall could be found in one of the ridges from south Babylon in an eastern direction towards Kish.

107-129 WaterUrban Pt1 Ch05 Pedersen.indd 113

12/05/2014 11:14


A History of Water

Royal roads, as main roads are called in Neo-Babylonian cuneiform documents, were often situated parallel to rivers and canals, sometimes immediately at the side of them, sometimes at a distance. The royal road from Babylon in an eastern direction to Kish was situated immediately south of the Banitu Canal, probably on the canal levee. In a southern direction, a royal road was positioned some 30 m west of the marshy area along the west side of the Euphrates. In a northern direction, another royal road is attested textually, east of the Euphrates at some distance, separated from the river by means of gardens (Wunsch, 2000). The Euphrates and canals that were not too small were also used for transportation. Huge bulk transports would especially fit the system of boat transportation (e.g. building materials), but also barley in large quantities for storage in silos next to the palace in Babylon (Pedersén, 2005). In connection with the New Year festival, the main gods (i.e., the cult statues) were transported on special processional boats, quite luxurious and built for this specific purpose. In this way, gods were brought from the temples in diverse cities on the special boats under jubilation to the New Year festival house in northern Babylon (e.g. Pongratz-Leisten, 1994). The Euphrates has seen several significant changes over the millennia, in addition to the constant smaller-scale movements (Figure 5.2). Surveys (Adams, 1972; Gasche and Tanret, 1998) and the study of Digital Elevation Models (DEMs) have shown, east of the Iskandariya Terrace, the remains of what seems to be a main river arm of the Euphrates, surrounded by a number of tells. Most agree that this was the ancient course of the river during an extended period, even if detailed datings are lacking and there has so far not been any corresponding survey in the area west of the Iskandariya Terrace, where the Euphrates now runs. Accepting the eastern course of the river as the main stream of the Euphrates during the Neo-Babylonian period lets the river approach Babylon from the north and not from the northwest, as the river arm has done in more recent times. The landscape north of Babylon (Figure 5.3) has not been the object of any published detailed survey. However, the middle and eastern sections were partly included in the preliminary (for this area) Akkad and Kish surveys (Adams, 1972; Gibson, 1972). A series of modern or pre-modern canals between Sippar and Babylon are, or have until recently been, leading the Euphrates water inlands in the direction of the Tigris. These include from the north the Latifiya, Iskanderiya, Musayib, and Mahawil canals. These west–east canals may well correspond at least to some extent to ancient canals, even if their beginnings at the modern Euphrates are 10–20 km further to the west of the ancient Euphrates. The Old Kutha Canal, mentioned in ancient texts, probably leading to the city Kutha some 30 km northeast of Babylon, may approximate one of the last-mentioned modern canals. The ancient Pallukkatu Canal, known as Pallakopas in the fourth century bc at the time of Alexander, according to Arrian, probably

107-129 WaterUrban Pt1 Ch05 Pedersen.indd 114

12/05/2014 11:14

Waters at Babylon115

Figure 5.2.  Babylon and surroundings. Main cities in the area of Babylon. White: Reconstruction of Euphrates and Tigris with main canals around 550 bc essentially according to Gasche (2010), with some additions. Black: Remains of East–West wall between the rivers. Source: Google Earth Pro, with additions.

had approximately the same alignment as the modern Euphrates (Boiy and Verhoeven, 1998). In an eastern direction (Figure 5.4), the landscape with sites and levees has been included in the detailed Kish survey (Gibson, 1972). Only the areas next to Babylon have not been surveyed, leading to unsecure interpretations. The landscape outside Babylon was dominated by the main Banitu Canal, leading water from the Euphrates to the nearest city, Kish, some 15 km away from Babylon and further east. The exact point where the Euphrates water entered the canal is not known, but a reasonable position would be in the area where the modern Nil Canal goes in approximately the same direction. This was probably north of the northernmost corner of the outer town wall of Babylon. Further on, the Banitu Canal ran east of the moat outside the town wall. Alternatives like cutting through the city itself—for example, as a continuation of the Libil-kegalla Canal

107-129 WaterUrban Pt1 Ch05 Pedersen.indd 115

12/05/2014 11:14


A History of Water

Figure 5.3.  Babylon and surroundings in northern direction. Reconstructed waters and Babylon model using ArchiCAD placed on Google Earth Pro satellite image.

Figure 5.4.  Babylon and surroundings in eastern direction. Banitu Canal leading to Kish and further on. Reconstructed waters and Babylon model using ArchiCAD placed on Google Earth Pro satellite image.

107-129 WaterUrban Pt1 Ch05 Pedersen.indd 116

12/05/2014 11:14

Waters at Babylon117

discussed below, or using the moat—seem less likely, but only a detailed examination with borings could clarify the situation. However, modern main roads have been constructed in some of these places. As hinted above, the ancient royal road was situated on the south side of the Banitu Canal on the levee, and at a distance south thereof, probably the wall from Babylon to Kish. Other canals may have extended in an eastern direction from the Banitu Canal at the two northern town gates on the eastern outer wall of Babylon. These gates have ancient names, according to a cuneiform text referring to canals (Gate of the Šuki Canal and Gate of the Mada¯nu Canal; George, 1992: 140). There are levees on the appropriate places extending from outside the gates, but the situation may turn out to be more complicated than the immediate identification would suggest. In the south (Figure 5.5), there is a lack of any systematic survey of ancient sites and levees. The Euphrates runs through Babylon in a southern direction, but further to the south, it is unclear if during the Neo-Babylonian period it followed the modern line or any of the main levees that can be seen on DEMs. The Borsippa Canal deviated in a southern or southwestern direction towards the approximate 17 km distant city Borsippa, probably essentially following the line of the modern Al-Uhudia Canal. This once ran west of Hilla, but due to the west expansion of the city, the canal is now more in the middle of the modern city. More uncertain is the possible alignment of the Piqu ¯du Canal in the southeast. Due to the lack of appropriate surveys giving dates of levees, it has to be left as open questions as

Figure 5.5.  Babylon and surroundings in southern direction. Borsippa Canal leading to Borsippa. Reconstructed waters and Babylon model using ArchiCAD placed on Google Earth Pro satellite image.

107-129 WaterUrban Pt1 Ch05 Pedersen.indd 117

12/05/2014 11:14


A History of Water

to whether the city Dilbad, some 27 km south of Babylon, was situated at a main canal, and whether the city Marad, 60 km southeast of Babylon, was along the Euphrates, as the modern situation would suggest. Far away in a western direction is the large desert area. Between the desert and the Babylon area lie some 25 km of the western flood plains. These are not yet properly surveyed, except from an agricultural point of view. However, there are a number of references to ancient sites in the agriculture survey (Buringh, 1961). Somewhere in the middle of this area, the Pallukkatu Canal could have been running from north to south, like the modern Euphrates’ main arm. Near the western part of Babylon, the Akkp-šullim Canal and the New Canal were coming from an approximately northwestern direction somewhere in the area of the enormous modern fishponds, just some 1.5 km northwest of the northwestern corner of Babylon’s western town wall (Wunsch, 2000). In all directions, there were not only the main canals reported here, but also a large number of smaller ones, all of which bordered on characteristically agricultural units consisting of gardens, fields, and marginal lands; the latter were often in the form of a permanent or seasonal marsh in this area. City of Babylon The focus of the discussion of the waters in the city of Babylon will be during the reign of Nebuchadnezzar II (604–562 bc) and the following periods. Archaeological excavations have exposed large areas of this huge ancient city, which was possibly the largest in the world within town walls during its palmy days. At the beginning of the reign of Nebuchadnezzar II, Babylon measured about 4.5 km2 and consisted of the eastern Inner City and the Western City, both of which were surrounded by town walls and moats. Later, during Nebuchadnezzar II’s more than 40-year reign, new extended town walls and moats were built surrounding the eastern Outer City, resulting in a total city area of some 9 km2. Other town walls within the city divided it into what may have been districts, and massive walls surrounded the palace area. However, due to the city’s enormous size, the spacious areas excavated represent only some 3 percent of the Neo-Babylonian Inner and Western City (or 1.5 percent of the expanded city including the Outer City). For the lower levels of the Inner and Western City, dated to the Old and Middle Babylonian periods, much less has been excavated—only about 0.05 percent—due essentially to high modern groundwater and overlaying levels. The present high ground­water level in the Babylon area excludes normal excavations of lower archaeological levels. The limited examination of these lower levels during the German excavations was only possible due to the collapse of the Hindiya Dam, resulting in a much lower groundwater during part of the excavation (Koldewey, 1990; Pedersén, 2011).

107-129 WaterUrban Pt1 Ch05 Pedersen.indd 118

12/05/2014 11:14

Waters at Babylon119

Not only did Nebuchadnezzar II double the city area, but also during his reign the use of baked bricks increased in a remarkable way and in many buildings (especially public ones) replaced the traditional unbaked bricks. During the following almost 2,500 years, Babylon was used for brick mining. Solid walls were taken down and the bricks were reused elsewhere for construction purposes (e.g. in the modern provincial capital Hilla) and for rebuilding the Hindiya Dam on the Euphrates. Early excavations during the nineteenth century concentrated their efforts on finding cuneiform clay tablets; therefore, they were less focused on documenting the architectural remains of the city. With the German excavations directed by Robert Koldewey (1899–1917), the architectural examinations had a central position, establishing the layout of main town walls and of several important buildings like palaces, some temples, and a selection of private houses, streets, and canals. During these excavations, brick mining was terminated. Many brick walls had already disappeared, but could often be reconstructed from empty spaces left by the ancient walls. German, Italian, and especially Iraqi archaeologists have periodically continued excavation activities. Grand-scale, full-sized Iraqi reconstructions occurred especially in the 1980s (Koldewey, 1990, with additional references). There has been an important restudy of a group of cuneiform texts dealing with the topography of Babylon (George, 1992). There has so far been no detailed systematic archaeological survey of the vast city area of Babylon, and its immediate surroundings in addition to the excavated sections. Geomagnetic and ground-penetrating radar investigations could possibly in the future show not only where buildings and streets were situated, but also the placement and alignment of watercourses and canals inside the city. For the moment, the size of the built areas inside and outside the town walls are not really known, and much more important information could be supplied with more investigations (Baker, 2009). Only a limited number of ancient streets and even fewer canals are known. Babylon is situated at the northern outskirts of the provincial capital Hilla, and inside the remains of the town walls there are traditional agriculture areas. Several villages and suburbs of Hilla are in the process of expanding in the area, and main streets and fishponds are constructed. There was an allied military presence for some years following the 2003 invasion, and in future a lot of tourists can be expected. Since 2009, the World Monuments Fund has been working on a management plan for the future of Babylon (Peruzzetto et al., 2011). The study of Babylon’s water management in a historical perspective is only really at the beginning. There are several traces of often later water movement to be seen on the present surface in Babylon and on any satellite image, aerial photograph, or series of detailed maps. However, secure dating of these features would in most cases require excavations and borings with proper evaluations. The Euphrates River in the center of Babylon had been moving westwards already during the Neo-Babylonian

107-129 WaterUrban Pt1 Ch05 Pedersen.indd 119

12/05/2014 11:14


A History of Water

period in the direction of the Western City; at present, it flows approximately through the middle of the ancient Western City. In the north of the city, in ancient times, the river (as discussed above) came from the north, flowing immediately west of the northern part of the Outer City. Later on, it seems to have flowed more from northwest to the same northern part of the Outer City, whereas in modern times the river comes more to the south from a northwestern direction towards the middle section of the west side of the Outer City. These and other river movements can be seen when comparing maps from different times, the Corona, and later satellite images. Cuttings of a large section of the surface in the northwestern part of the Outer City for irrigation dams have disclosed the ancient river courses now to be seen on satellite images of that area, but which were hardly noticeable before. In connection with renewed Italian excavations, a reconsideration of the ancient water levels in Babylon has been presented, with several suggestions that should be tested (Bergamini, 1977). On the eastern side of the town wall around the Outer City, there were some interesting, preliminary results of a short geological investigation. The common inundations east of the eastern wall of the Outer City seemed to have been at least partly blocked by the remains of the town wall. The result was a higher level of the surface outside the town wall, and a lower level on the protected inner side of the wall, resulting in the modern lake in the southeast corner of the Outer City (Wahab, 1979). In several lower areas of both the Outer City and the eastern Inner City, there are traces of water streams cutting through the city area. The German archaeologists, influenced by Greek traditions, interpreted one series of such traces as a change of the Euphrates in the Achaemenid period running from the west through the middle of the northern part of the Outer City, cutting into the northeastern section of the Inner City, and returning to approximately the old direction south of the Southern Palace. In the Parthian Period, the river moved back to the old riverbed (Wetzel, 1930, 1957; Bergamini, 1977; Koldewey, 1990). Geological borings seem to have confirmed a watercourse at the place, but the dating of it has not been secured, or at least not published (Wahab, 1979). Some scholars have questioned the existence of the change, claiming that the archaeological evidence does not prove it and Greek traditions do not require it (e.g. Rollinger, 1993; Heinsch and Kuntner, 2011). In reality, several traces of watercourses to be seen on aerial photographs and satellite images may indicate an even more complex situation (cf. provisionally Lippolis et al., 2011, lacking ground truth), but only proper interpretations of borings, and a restudy of the excavation may provide the evidence for datings of all the changes and the complex water regime. In ancient times, there were several large-scale water-related projects registered in royal building inscriptions from Babylon. Many inscriptions found in relation to buildings mentioned in the texts refer to the construction of quay walls of baked brick and asphalt protecting the runs of the

107-129 WaterUrban Pt1 Ch05 Pedersen.indd 120

12/05/2014 11:14

Waters at Babylon121

Euphrates, canals, moats, and dams, as well as to the cleaning of the watercourses. Nebuchadnezzar reconstructed the Libil-kegalla Canal south of the South Palace with a bridge under the street of procession, but the continuation of the canal in an eastern direction came a few years after his reign, during the reign of his son-in-law, Neriglissar (559–556 bc). Several inscriptions have references to groundwater in the construction of main buildings. Large public buildings often had their foundations made of baked brick and asphalt down to groundwater. However, because of foundation problems due to flood water, the foundations of the palace of Nabopolassar (625–605 bc) had become weakened and were completely rebuilt by his son, Nebuchadnezzar II. Nebuchadnezzar had a large fortification (kalsu) constructed in the river as a western extension of the palace. This has been identified with the massive “westliche Vorwerk” recorded during excavations, and now partly covered by the hill with the modern presidential palace placed in the ancient Euphrates. After Nebuchadnezzar’s reign, part of his palace (possibly the mentioned fortification west of it) collapsed into the river and had to be rebuilt a few years later by Neriglissar (Langdon, 1912; Koldewey, 1990; George, 1992; George, 1993). The Euphrates was running in the middle of Babylon from the north to the south (Figure 5.6). The excavations have mostly been concentrated on

Figure 5.6.  Central Babylon. View in eastern direction. Inner City with palace and zikkurrat, behind the Outer City. Canal in the northern section of the palace and Libilkegalla Canal south of the palace. Reconstructed waters and Babylon model using ArchiCAD placed on Google Earth Pro satellite image.

107-129 WaterUrban Pt1 Ch05 Pedersen.indd 121

12/05/2014 11:14


A History of Water

the eastern Inner City surrounded by an 80 m wide moat and the double inner town wall. The large palace area, some main temples, the zikkurrat (i.e., the great temple tower), and a selection of private houses in a network of both main and smaller streets, as well as from a later period the Greek theater, have been exposed so far. In the northwest corner of the Inner City, next to the Euphrates and the moat, was the enormous palace, greatly expanded by Nebuchadnezzar. The Libil-kegalla Canal was on the south side of the palace; another canal led into the northern part of the palace. There were quays and harbors along the Euphrates, inside the city at the Libil-kegalla Canal, and along the canal into the palace. Other canals are known from texts, but have not been discovered. The quays and harbors were used not only for the abovementioned large-scale boat transport of goods like barley to huge silos next to the palace, but also for religious processions, with gods traveling on special boats at the New Year festivals to the (not yet archaeologically secured) New Year Temple north of the inner town wall (Koldewey, 1931, 1932, 1990; George, 1993; Van de Mieroop, 2003; Pedersén, 2005). The water system in central Babylon (Figure 5.7) consisted of more than a single solution both for fresh water and for sewers. The main slope of the flood plain and the main flow direction of the Euphrates are from the northwest to the southeast. This is also the main direction of the winds during some 300 days of the year. The best position for fresh water and fresh air was therefore in the northwest of the city. At this corner of the Inner City, the main royal palace was situated with no other known buildings in a northwest direction, but with built-up areas in all other directions. Fresh water for gardens, washing, and drinking was mostly taken from the Euphrates. As described above, the groundwater did not have the best quality on the flood plain, due to its salt content. However, in the Babylon area, especially at the riverbanks and in the area west of it, it now has better quality (salinity less than 1 g/l) than at most other places on the flood plain, due to infiltration from the river. The situation may have been the same in ancient times (Koldewey, 1990; Krásný et al., 2006; Hydrogeological Map of Karbala). The central palace area had a rather advanced, only partially understood fresh water system fed by the Euphrates. Not only was the already mentioned west–east canal in the northern part of the palace possible to use as a secure quay area, but also a series of tunnels diverting from the canal in a southern direction passed on Euphrates water under the palace area. Even if these constructions were not fully examined due to the wartime situation at the end of the German excavation, the archaeologists interpreted them as the water source for a series of fresh water wells in the palace area. Several wells have been excavated in the South Palace, two further wells at the nearby Ishtar Gate, and one well in each of the courtyards of the Ninmak Temple and the Ishtar Temple. There existed also one, before the time of Nebuchadnezzar, abandoned well in the area of private

107-129 WaterUrban Pt1 Ch05 Pedersen.indd 122

12/05/2014 11:14

Waters at Babylon123

Figure 5.7.  Central Babylon. Inner City with palace and zikkurrat. Canal in the northern section of the palace, Libil-kegalla Canal south of palace. Excavated fresh water wells marked with black dots. Reconstructed waters and Babylon model using ArchiCAD placed on Google Earth Pro satellite image.

houses in the Merkes central area with private houses. Ordinary people probably collected fresh water from the Euphrates. At the northwest section of the inner town wall, outside the northwest corner of the South Palace, was found what may have been an installation to collect fresh water from the river secretly, for use in the palace (Koldewey, 1931, 1932). Water from the river was used for irrigation purposes, not only outside the town walls, but also inside the city itself. Private gardens in Babylon are attested in legal documents (Wunsch, 2000). The Juniper Gardens at the Marduk temple in the southwestern part of the Inner City are attested in several texts from the Hellenistic period (Sachs and Hunger, 1988, 1989, 1996). The Greek traditions have the Hanging Gardens in the palace of Babylon as one of the seven architectural wonders of the world. Due to problems with secure archaeological attestations of gardens, and so far no clear reference in preserved cuneiform texts, there has been a sometimes lively discussion about different, not always well-founded alternatives of the possible placement inside and outside Babylon (e.g. Koldewey, 1931; Krischen, 1956; Demirji, 1981; Wiseman, 1985; Koldewey, 1990; Dalley, 2013). The evidence for water installations in connection with the excavator’s placement of the Hanging Gardens, so far hardly accounted for by anyone else except the excavators themselves, has to be restudied.

107-129 WaterUrban Pt1 Ch05 Pedersen.indd 123

12/05/2014 11:14


A History of Water

Even if yearly precipitation was too low to allow agriculture without irrigation, heavy rain could occasionally occur. This had been accounted for in different areas of the city, but especially in the palace area, which had an elaborate system for a rainwater outlet directing rainwater out from the palace. In the private houses in the city, a number of seep sink tubes have been found. These were in use as outlets for washing water and from toilets in the private houses. It was not established during the excavations how toilets could have worked in the palaces (Reuther, 1926; Koldewey, 1931, 1932, 1990). With such a small part of the huge Babylon city area properly excavated and analyzed, our knowledge of the water systems remains rather limited. Much more secure results would probably be possible to attain with an accurate restudy of the old excavations (including the cuneiform texts and other ancient documentation), application of modern techniques and interpretative methods, combined with ground truth on the spot in Babylon. Conclusions In this chapter, waters at Babylon have been considered from three points of view. In the section on the Mesopotamian flood plain, there was a short survey of both the modern and the ancient water conditions, essentially treating the river and canal systems, but also including quite short notes about precipitation and groundwater. In the two following sections, the focus was especially on the period during and immediately following the reign of Nebuchadnezzar II in the first half of the sixth century bc. In the section on the surroundings of Babylon, there were discussions in some more detail concerning canal and river systems and their reconstructions in the surroundings of Babylon. In the city of Babylon section, a somewhat more detailed treatment of selected aspects of water in the city of Babylon itself has been critically presented. Only a preliminary overview of available material, methods in use, and research progress has been given here. The historical water situation on the Mesopotamian flood plain is, despite a lot of research, only partially known today, and is essentially limited to central parts of the plain. The availability of new methods and materials can change the situation. Remote sensing with satellite images, DEM, geomagnetic, and ground-penetrating radar investigations have the potential to improve the knowledge considerably when combined with ground truth in the form of surveys, excavations, and borings properly interpreted. A new study of all the relevant cuneiform texts and other ancient writings has to be combined with the improved material evidence.

107-129 WaterUrban Pt1 Ch05 Pedersen.indd 124

12/05/2014 11:14

Waters at Babylon125

Acknowledgements The basic research was possible by means of support from the Swedish Research Council, the Mistra-supported Urban Mind project at Uppsala University, and the Excellence Cluster Topoi at Freie Universität Berlin. A longtime collaboration with J. Marzahn at Vorderasiatisches Museum Berlin and Deutsche Orient-Gesellschaft was fundamental for the Babylon research. Preliminary versions of this chapter were presented at the workshop on water systems and urbanization in Africa and beyond in Uppsala, and at the 58e Rencontre Assyriologique Internationale in Leiden, both in 2012. S. Alsam and J. Andersson have kindly commented on early versions of the chapter.

references Adams, R. McC. (1965). Land Behind Baghdad, Chicago, IL and London: University of Chicago Press. ——— (1972). “Appendix V. Settlement and irrigation patterns in ancient Akkad”, in M. Gibson, The City and Area of Kish, pp. 182–208, Maps 1–9, Coconut Grove, Miami, FL: Field Research Projects. ——— (1981). Heartland of Cities: Surveys of Ancient Settlement and Land Use on the Central Floodplain of the Euphrates, Chicago, IL and London: University of Chicago Press. Adams, R. McC. and H. J. Nissen (1972). The Uruk Countryside: The Natural Setting of Urban Societies, Chicago, IL and London: University of Chicago Press. Al-Jiburi, H. K. and N. H. Al-Basrawi (2011). “Hydrogeology of the Mesopotamia Plain”, Iraqi Bulletin of Geology and Mining, Special Issue No. 4: Geology of the Mesopotamia Plain, pp. 83–103. Aqrawi, A. A. M., J. Domas, and S. Z. Jassim (2006). “Quaternary deposits”, in S. Z. Jassim and J. C. Goff (eds), Geology of Iraq, pp. 185–98, Prague: Dolin & Brno: Moravian Museum. Bagg, A. M. (2003). “2000 Jahre Wasserbau im Alten Mesopotamien: Ein Überblick”, in C. Ohlig (ed.), Wasserhistorische Forschungen: Schwerpunkt Antike, pp. 107–17, Siegburg: Schriften der Deutschen Wasserhistorischen Gesellschaft 2, Books on Demand. ——— (2012). “Irrigation”, in D. T. Potts (ed.), A Companion to the Archaeology of the Ancient Near East, pp. 261–78, Oxford: Wiley-Blackwell. Baker, H. D. (2009). “A waste of space: unbuilt land in the Babylonian cities of the first millennium bc”, Iraq, 71, pp. 89–98. Bergamini, G. (1977). “Levels of Babylon reconsidered”, Mesopotamia, 12, pp. 111–72, figs 71–81. Boiy, T. and K. Verhoeven (1998). “Arrian, Anabasis VII 21.1–4 and the Pallukkatu Channel”, in M. Gasche and M. Tanret, Changing Watercourses in Babylonia: Towards a Reconstruction of the Ancient Environment in Lower Mesopotamia, Vol. 1, Mesopotamian History and Environment,

107-129 WaterUrban Pt1 Ch05 Pedersen.indd 125

12/05/2014 11:14


A History of Water

Series II, Memoirs V, pp. 147–58, Chicago, IL: Oriental Institute & Ghent: University of Ghent. Brown, A. G. (1997). Alluvial Geoarchaeology: Floodplain Archaeology and Environmental Change, Cambridge Manuals in Archaeology, Cambridge: Cambridge University Press. Buringh, P. (1960). Soils and Soil Conditions in Iraq, Republic of Iraq, Ministry of Agriculture, Directorate General of Agricultural Research and Projects, Wageningen: H. Veenman & Zonen N.V. ——— (1961). Soil Survey Report of the Hilla—Kifl Drainage Project (Hilla Section), Republic of Iraq, Ministry of Agriculture, Directorate General of Agricultural Research and Projects, Technical Bulletin 8, Baghdad: Government Press. Cadoux, H. W. (1906). “Recent changes in the course of the Lower Euphrates”, The Geographical Journal, 28(3), pp. 266–77. Cole, S. W. (1994). “Marsh formation in the Borsippa Region and the course of the Lower Euphrates”, Journal of Near Eastern Studies, 53(2), pp. 81–109. Dalley, S. (2013). The Mystery of the Hanging Garden of Babylon: An Elusive World Wonder Traced, Oxford: Oxford University Press. Demirji, M. S. (1981). “Where are the Hanging Gardens”, Sumer, 37, pp. 56–61 (Arabic section). Eger, A. A. (2011). “The swamps of home: marsh formation and settlement in the early medieval Near East”, Journal of Near Eastern Studies, 70(1), pp. 55–79. Gasche, H. (2010). “Les défenses avancées de Babylone à l’époque de Nabuchodonosor II”, Mesopotamia, 45, pp. 113–19, pls. I–VI. Gasche, H., I. Druc, C. Simon, A. Susini, C. A. Baud, M. Klohn, D. Lacotte, L. Chaix, and F. Bruschweiler (1989). “Habl as-Sakr 1986 nouvelles fouilles: L’ouvrage défensif de Nabuchodonosor II au nord de Sippar”, in Northern Akkad Project Reports, Vol. 2, Mesopotamian History and Environment, Series I, pp. 23–70, Ghent: Rijksuniversiteit Gent. Gasche, H., R. G. Killick, R. Nijs, G. Stoops, J. A. Black, and A. Gautier (1987). “Habl as-Sakr 1983–85: Nebuchadnezzar II’s cross-country wall north of Sippar”, in Northern Akkad Project Reports, Vol. 1, Mesopotamian History and Environment, Series I, pp. 3–46, Ghent: Rijksuniversiteit Gent. Gasche, H. and M. Tanret (1998). Changing Watercourses in Babylonia: Towards a Reconstruction of the Ancient Environment in Lower Mesopotamia, Vol. 1, Mesopotamian History and Environment, Series II, Memoirs V, Chicago, IL: Oriental Institute & Ghent: University of Ghent. Gasche, H., M. Tanret, S. W. Cole, and K. Verhoeven (2002). “Fleuves du temps et de la vie: Performance et instabilité du réseau fluviatile babylonien entre 2500 et 1500 avant notre ère”, Annales Histoire, Sciences Sociales, 57(3), pp. 531–44. George, A. R. (1992). Babylonian Topographical Texts, Orientalia Lovaniensia Analecta 40, Leuven: Peeters. ——— (1993). “Babylon revisited: archaeology and philology in harness”, Antiquity, 67(4), pp. 734–46. Gibson, M. (1972). The City and Area of Kish, Coconut Grove, Miami, FL: Field Research Projects. Heinsch, S. and W. Kuntner (2011). “Herodot und die Stadtmauer Babylons: Bemerkungen zur archäologischen Befundsituation der Landmauern”, in

107-129 WaterUrban Pt1 Ch05 Pedersen.indd 126

12/05/2014 11:14

Waters at Babylon127

R. Rollinger, B. Truschnegg, and R. Bichler (eds), Herodot und das Persische Weltreich, pp. 499–529, Wiesbaden: Harrassowitz. Hritz, C. (2010). “Tracing settlement patterns and channel systems in southern Mesopotamia using remote sensing”, Journal of Field Archaeology, 35(2), pp. 184–203. Janssen, C. (1995). Babil, the City of Witchcraft and Wine: The Name and Fame of Babylon in Medieval Arabic Geographical Texts, Mesopotamian History and Environment, Series I, Memoirs II, Chicago, IL: Oriental Institute & Ghent: University of Ghent. Koldewey, R. (1931). Die Königsburgen von Babylon I: Die Südburg, Wissenschaftliche Veröffentlichung der Deutschen Orient-Gesellschaft 55, Leipzig: J. C. Hinrichs. ——— (1932). Die Königsburgen von Babylon II: Die Nordburg und der Sommerpalast Nebukadnezars im Hügel Babil, Wissenschaftliche Veröffentlichung der Deutschen Orient-Gesellschaft 55, Leipzig: J. C. Hinrichs. ——— (1990). Das wieder erstandene Babylon, Neu herausgegeben von Barthel Hrouda, München: C. H. Beck’sche Verlagsbuchhandlung; and Berlin: Akademie-Verlag. Krásný, J., S. Alsam, and S. Z. Jassim (2006). “Hydrogeology”, in S. Z. Jassim and J. C. Goff (eds), Geology of Iraq, pp. 251–87, Prague: Dolin & Brno: Moravian Museum. Krischen, F. (1956). Weltwunder der Baukunst in Babylonien und Jonien, Tübingen: Wasmut. Langdon, S. (1912). Die neubabylonischen Königsinschriften (transl. by R. Zehnpfund), Vorderasiatische Bibliothek 4, Leipzig: J. C. Hinrichs’sche Buchhandlung. Lebon, J. H. G. (1964). “Recent research on the land potential of Iraq”, Geographical Review, 54(1), pp. 104–9. Le Strange, G. (1895). “Description of Mesopotamia and Baghdad, written about the year 900 A.D. by Ibn Serapion. Part 1 and 2”, Journal of the Royal Asiatic Society, 1895(1, 2), pp. 1–76, 255–315. Lippolis, L., B. Monopoli, and P. Baggio (2011). “Babylon’s urban layout and territory from above”, Mesopotamia, 46, pp. 1–8, pls. 1–5. Marzahn, J. and G. Schauerte (eds) (2008). Babylon: Mythos und Wahrheit, München: Hirmer. Partow, H. (2001). The Mesopotamian Marshlands: Demise of an Ecosystem, United Nations Environment Programme, Division of Early Warning and Assessment. Technical Report 01-3, Nairobi, Kenya. Pedersén, O. (2005). Archive und Bibliotheken in Babylon: Die Tontafeln der Grabung Robert Koldeweys 1899–1917, Abhandlungen der Deutschen Orient-Gesellschaft 25, Wiesbaden: Harrassowitz (sdv). ——— (2011). “Excavated and unexcavated libraries in Babylon”, in E. Cancik-Kirschbaum, M. van Ess, and J. Marzahn (eds), Babylon: Wissenskultur in Orient und Okzident, Topoi: Berlin Studies of the Ancient World 1, pp. 47–67, Berlin: De Gruyter. Peruzzetto, A., J. Allen, and G. Haney (2011). “The future of Babylon: management, conservation planning and cultural landscape at Babylon”, Mesopotamia, 46, pp. 53–7.

107-129 WaterUrban Pt1 Ch05 Pedersen.indd 127

12/05/2014 11:14


A History of Water

Pongratz-Leisten, B. (1994). Ina šulmi ¯rub: ı Die kulttopographische und ideologische Programmatik der akı¯tu-Prozession in Babylonien und Assyrien im 1. Jahrtausend v. Chr. Baghdader Forschungen, Band 16, Mainz: Philipp von Zabern. Poyck, A. P. G. (1962). Farm Studies in Iraq: An Agro-Economic Study of the Agriculture of the Hilla-Diwaniya Area in Iraq, Wageningen: H. Veenman en Zonen N.V. Reade, J. (2010). “How many miles to Babylon?”, in H. D. Baker, E. Robson, and G. Zólyomi, Your Praise is Sweet: A Memorial Volume for Jeremy Black from Students, Colleagues and Friends, pp. 281–90, London: British Institute for the Study of Iraq. Reuther, O. (1926). Die Innenstadt von Babylon (Merkes), Wissenschaftliche Veröffentlichung der Deutschen Orient-Gesellschaft 47, Leipzig: J. C. Hinrichs. Rollinger, R. (1993). Herodots babylonischer Logos. Eine kritische Untersuchung der Glaubwürdigkeitsdiskussion an Hand ausgewählter Beispiele: Historische Parallelüberlieferung—Argumentationen—Archäologischer Befund— Konsequenzen für eine Geschichte Babylons in persischer Zeit, Innsbrucker Beiträge zur Kulturwissenschaft, Sonderheft 84, Innsbruck: Institut für Sprachwissenschaft der Universität Innsbruck. Rzóska, J., J. F. Talling, and K. E. Banister (1980). Euphrates and Tigris: Mesopotamian Ecology and Destiny, Monographiae biologicae 38, The Hague: Dr. W. Junk. Sachs, A. J. and H. Hunger (1988, 1989, 1996). Astronomical Diaries and Related Texts from Babylonia, Vols 1–3, Österreichische Akademie der Wissenschaften, Philosophisch-Historische Klasse, Denkschriften 195, 210, 247, Wien: Verlag der Österreichischen Akademie der Wissenschaften. Steinkeller, P. (2001). “New light on the hydrology and topography of southern Babylonia in the third millennium”, Zeitschrift für Assyriologie und Vorderasiatische Archäologie, 91(1), pp. 22–84. Van de Mieroop, M. (2003). “Reading Babylon”, American Journal of Archaeology, 107(2), pp. 257–75. Verhoeven, K. (1998). “Geomorphological research in the Mesopotamian flood plain”, in M. Gasche and M. Tanret, Changing Watercourses in Babylonia: Towards a Reconstruction of the Ancient Environment in Lower Mesopotamia, Vol. 1, Mesopotamian History and Environment, Series II, Memoirs V, pp. 159–245, Chicago, IL: Oriental Institute & Ghent: University of Ghent. Wahab, S. A. (1979). “Some preliminary results of the geological survey in Babylon”, Sumer, 35, pp. 160–3. Wetzel, F. (1930). Die Stadtmauern von Babylon, Wissenschaftliche Veröffentlichung der Deutschen Orient-Gesellschaft 48, Leipzig: J. C. Hinrichs. ——— (1957). Das Babylon der Spätzeit, Wissenschaftliche Veröffentlichung der Deutschen Orient-Gesellschaft 62, Berlin: Gebr. Mann Verlag. Wilkinson, T. J. (1990). “Early channels and landscape development around Abu Salabikh, a preliminary report”, Iraq, 52, pp. 75–83. ——— (2003). Archaeological Landscapes of the Near East, Tucson, AZ: University of Arizona Press. ——— (2012). “Introduction to geography, climate, topography, and hydrology”, in D. T. Potts (ed.), A Companion to the Archaeology of the Ancient Near East, pp. 3–26, Oxford: Wiley-Blackwell.

107-129 WaterUrban Pt1 Ch05 Pedersen.indd 128

12/05/2014 11:14

Waters at Babylon129

Wirth, E. (1962). Agrargeographie des Irak, Hamburger Geographische Studien 13, Hamburg: Selbstverlag des Instituts für Geographie und Wirtschaftsgeographie der Universität Hamburg. Wiseman, D. J. (1985). Nebuchadrezzar and Babylon, The Schweich lectures of the British Academy 1983, Oxford: Oxford University Press. Wunsch, C. (2000). Das Egibi-Archiv I. Die Felder und Gärten. Band I–II, Cuneiform Monographs 20A-B, Groningen: Styx. Yacoub, S. Y. (2011). “Stratigraphy of the Mesopotamia plain”, Iraqi Bulletin of Geology and Mining, Special Issue No. 4: Geology of the Mesopotamia Plain, pp. 47–82.

Maps Geological Hazards Map of Karbala Quadrangle, Scale 1:250,000, Sheet NI-38-14, Baghdad: GEOSURV. Geological Map of Karbala Quadrangle, Scale 1:250,000, Sheet NI-38-14, Baghdad: GEOSURV. Geomorphological Map of Iraq, Scale 1:1,000,000, Baghdad: GEOSURV. Hydrogeological Map of Iraq, Scale 1:1,000,000, Baghdad: GEOSURV. Hydrogeological Map of Karbala, Scale 1:250,000, Sheet NI-38-14, Baghdad: GEOSURV. Iraq Department of Antiquities (1976). Archaeological Atlas of Iraq (in Arabic), Baghdad: Department of Antiquities. Julius (1862). “Originalkarte vom Thal des Euphrat und einem Theile des Alten Mesopotamien bei mittlerem Wasserstande”, in Petermanns Geographische Mittheilungen, Tafel 16. Land Use Map of Iraq, Scale 1:1,000,000, Baghdad: GEOSURV. Quaternary Sediment Map of Iraq, Scale 1:1,000,000, Baghdad: GEOSURV. Selby, W. B. and W. Collingwood (1859). “Plan of the supposed ruins of Babylon”. Selby, W. B., W. Collingwood, and J. B. Bewsher (1885). “Surveys of Ancient Babylon and the surrounding ruins with part of the rivers Tigris and Euphrates, the Hindiyeh Canal, the Sea of Nejf & the Shat Atshar made by order of the Government of India in 1860 to 1865.”

Digital references ANE.kmz Placemarks for Ancient Near Eastern sites on Google Earth: www.lingfil. uu.se/staff/olof_pedersen/Google_Earth Corona Images for the Ancient Near East shown in Web GIS over Google Earth (in BETA): corona.cast.uark.edu/index.html Digital Model of Babylon project with better-scaled color illustrations to the present article: www.lingfil.uu.se/staff/olof_pedersen/Babylon_Model Google Earth download: www.google.com/earth/index.html

107-129 WaterUrban Pt1 Ch05 Pedersen.indd 129

12/05/2014 11:14


Water Control in Ancient Greek Cities

Demetris Koutsoyiannis and Anna Patrikiou Context The control of water resources in ancient Greece, as well as in modern Greece, is affected by its geophysical characteristics and climate. Earlier civilizations bloomed in large river valleys, which had water in abundance (Mesopotamia near the Tigris and the Euphrates, Egypt near the Nile, India near the Indus). However, Greece does not have large rivers, and it is divided by mountains into small plains. It is in these plains where the major part of land cultivation takes place (about 20 percent of the Greek peninsula and islands). Historically, the gathering of people and activities in these plains led to the development of urban centers. The physical boundaries of small plains form the boundaries of areas where the so-called poleis or city-states, entities with self-governance, autonomy, and independence, were developed. These range from fairly small states, with an area of 100 km2, to fairly large states, which spread over an area of about 5,000 km2. Diagrammatically, we can thus visualize the ancient Greek city-states from squares of 10 × 10 km, which could be crossed from end to end in two hours or so, to squares of 70 × 70 km, which need no more than 14 hours to cross on foot (Doxiadis, 1964). Observing carefully the locations of those city-states, we notice that most ancient Greek important centers were built in the driest areas (Figure 6.1). We don’t know the exact reasons for this, but we may assume that ancient Greeks considered a dry climate as more convenient or healthier. Certainly, that dry climate and the implied water scarcity had consequences and impacts on the heart of civilization and social organization, most of them positive. Maybe it was exactly that scarcity that triggered the progress in philosophy and technology, and, furthermore, conditioned the character and behavior of Greeks. The prehistory of Greece starts perhaps in the seventh millennium bc, when the Greek territory had been inhabited by indigenous pre-Greek people, who at some stage practiced agriculture. The first Greek-speaking tribes are generally thought to have arrived in the Greek mainland between the late third millennium and the early second millennium bc. Among the

130-148 WaterUrban Pt1 Ch06 Koutsoyiannis.indd 130

12/05/2014 11:20

Water Control in Ancient Greek Cities131

Figure 6.1.  Map of climatic regimes in Greece, based on the distribution of mean annual rainfall, indicating the location of major ancient Greek centers.

civilizations that have bloomed in this territory through prehistory, we can distinguish the Minoan (with Knossos, shown in Figure 6.1, being a characteristic site), Cycladic (characteristic sites: Thera, Delos), and Mycenaean civilizations (characteristic site: Mycenae). The most famous sites during the historical era are Athens and Sparta (also shown in Figure 6.1). A typical classification of the different periods of Greek antiquity is given in Table 6.1. Table 6.2 provides a more detailed timeline of the historic period as reflected in the history of the Athenian city-state. From mythology to philosophy and science The earliest attempts by humankind to explain nature have been reflected in mythologies. Greek mythology offers a great deal of stories related to

130-148 WaterUrban Pt1 Ch06 Koutsoyiannis.indd 131

12/05/2014 11:20


A History of Water

Table 6.1.  Main periods of Greek antiquity. Period name

Years bc


Minoan (prehistory)


Cycladic (prehistory) Mycenaean (prehistory) Greek Dark Age (misnomer) Archaic Classical Hellenistic

c.3100–1600 c.1550–1150 c.1200–800 c.800–500 336–323 323–146

Island of Crete (mostly known from large houses and luxurious palaces, e.g. Knossos, Zakros, Mallia, Gortys, Phaestos) Islands of the Aegean (e.g. Thera, Delos) Mainland Greece Also known as Geometric or Homeric Age 323: death of Alexander the Great; 146: annexation by Rome

Note: chronologies not generally agreed.

Table 6.2.  Timeline of Greek states’ organization as reflected in the history of the Athenian city-state. Period name

Years bc

Eupatrid Oligarchy Solonian Tyranny of Peisistratids Foundation of democracy (Democracy of Cleisthenes, Persian Wars,   Delian League and postwar rebuilding) Radical Democracy of Pericles Peloponnesian War and Oligarchy Post-Peloponnesian War Radical Democracy Macedonian and Roman Domination

700–600 600–561 561–510 510–462 462–431 431–403 403–322 322–146

water, many of which have a strong metaphorical meaning and symbolism. The myth of the competition of Athena and Poseidon reveals that scarcity may become richness and it underlines the value of wisdom. Athenians, in order to choose their patron god, organized a competition for the two prevalent candidates: Athena (goddess of wisdom, as well as of household arts and crafts) and Poseidon (god of waters). The one who offered the best gift to the city would be the winner. Poseidon offered abundant water by creating a well in Acropolis. Athena offered the olive tree and an explanation why it would be wiser to choose her gift. Athenians opted for wisdom, which means that scarcity may not be a punishment but a choice, and a powerful one. Interestingly, according to one version of the story (Thomson, 1949), it is the female population who voted for Athena, while men preferred Poseidon, but the election was won by women, with a majority of just one vote (and in revenge, men suspended women’s votes in subsequent decisions). Among other Greek myths related to the element of water (one of the so-called “four elements”), a popular one is the battle of Hercules against

130-148 WaterUrban Pt1 Ch06 Koutsoyiannis.indd 132

12/05/2014 11:20

Water Control in Ancient Greek Cities133

Acheloos, the largest in the discharge river of Greece, which Herodotus (c.484–425 bc) compares with the Nile. The battle was won by Hercules and the literal meaning of that victory was explained later by the historian Diodorus Siculus (c.90–30 bc) and the geographer Strabo (c.64 bc–ad 24): it was related to the channel excavation and the construction of dikes to confine the shifting bed of Acheloos (see also Koutsoyiannis et al., 2007). Towards the end of the archaic era, around 600 bc and mainly in Ionia, technological needs, combined with a new awareness about nature, triggered physical explanations of natural phenomena. They were no longer believed to happen due to the intervention of supernatural powers. That was a new approach, leading to the foundation of philosophy and science. The study of hydrometeorological phenomena (evaporation, cloud formation, rain, hail and snow, river flow, and, more generally, the hydrological cycle) had a major role in the birth of science. During the fifth and fourth centuries bc, philosophy and science were further developed in classical Athens, forming a body of knowledge that would be dominant for about 2,000 years. During the Hellenistic period (fourth–first centuries bc), there was significant progress in mathematics, physics, and technology; scientific views at that time were advanced and often consistent with contemporary scientific views (see also Koutsoyiannis et al., 2007). Greek philosophers’ thoughts and theories were influenced by different facets of problems related to the nature and the dynamics of water. Many of these theories now seem erroneous to us (as happens with most theories, as time goes by), but there are many impressive elements in Greek exegeses of hydrometeorological processes. Below we refer to three of the most important philosophers belonging to three different eras: Thales for the archaic period; Aristotle for the classical period; and Hero for the Hellenistic period. Thales of Miletus (c.640–546 bc), founder of the Ionic philosophy (and, according to many, father of philosophy and of science), proclaimed water as the fundamental substance of the world. He proposed a physical exegesis for the “Nile puzzle” (the fact that the Nile floods occur during summertime when rainfall in Egypt is minimal), thus emphasizing the importance of hydrology in science. This explanation was based on the regime of winds and was blatantly wrong, but the important thing is that a natural phenomenon was described and studied on physical grounds (see also Koutsoyiannis et al., 2007). A correct exegesis had to wait until the Hellenistic period—it was given by Eratosthenes (c.276–195 bc). Interestingly, Thales accomplished the diversion of the Halys River, thus emphasizing the link of technology and philosophy, at the dawn of the latter. Aristotle (384–322 bc) was a student of Plato, but his theories were influenced by Ionic philosophers. His treatise Meteorologica offers a great contribution to the explanation of hydrometeorogical phenomena (see also Koutsoyiannis et al., 2007):

130-148 WaterUrban Pt1 Ch06 Koutsoyiannis.indd 133

12/05/2014 11:20


A History of Water

ἔτι δ’ ἡ ὑπὸ τοῦ ἡλίου ἀναγωγὴ τοῦ ὑγροῦ ὁμοία τοῖς θερμαινομένοις ἐστὶν ὕδασιν ὑπὸ πυρός. The sun causes the moisture to rise; this is similar to what happens when water is heated by fire. (Meteorologica II.2, 355a 15) συνίσταται πάλιν ἡ ἀτμὶς ψυχομένη διά τε τὴν ἀπόλειψιν τοῦ θερμοῦ καὶ τὸν τόπον, καὶ γίγνεται ὕδωρ ἐξ ἀέρος· γενόμενον δὲ πάλιν φέρεται πρὸς τὴν γῆν. ἔστι δ’ ἡ μὲν ἐξ ὕδατος ἀναθυμίασις ἀτμίς, ἡ δ’ ἐξ ἀέρος εἰς ὕδωρ νέφος. The vapour that is cooled, for lack of heat in the area where it lies, condenses and turns from air into water; and after the water has formed in this way it falls down again to the earth; the exhalation of water is vapour; air condensing into water is cloud. (Meteorologica I.9, 346b 30)

Aristotle also recognized the principle of mass conservation within the hydrological cycle: ὥστε οὐδέποτε ξηρανεῖται· πάλιν γὰρ ἐκεῖνο φθήσεται καταβὰν εἰς τὴν αὐτὴν τὸ προανελθόν. Thus, [the sea] will never dry up; for [the water] that has gone up beforehand will return to it. (Meteorologica II.3, 356b 26) κἂν μὴ κατ’ ἐνιαυτὸν ἀποδιδῷ καὶ καθ’ ἑκάστην ὁμοίως χώραν, ἀλλ’ ἔν γέ τισιν τεταγμένοις χρόνοις ἀποδίδωσι πᾶν τὸ ληφθέν. Even if the same amount does not come back every year or in a given place, yet in a certain period all quantity that has been abstracted is returned. (Meteorologica II.2, 355a 26)

He understood deeply the concept of “change”, perhaps better than we do today. He was fully aware that the landscape changes through the ages and that rivers are formed and disappear in the course of time (see also Koutsoyiannis et al., 2007): ἀλλὰ μὴν εἴπερ καὶ οἱ ποταμοὶ γίγνονται καὶ φθείρονται καὶ μὴ ἀεὶ οἱ αὐτοὶ τόποι τῆς γῆς ἔνυδροι, καὶ τὴν θάλατταν ἀνάγκη μεταβάλλειν ὁμοίως. τῆς δὲ θαλάττης τὰ μὲν ἀπολειπούσης τὰ δ’ ἐπιούσης ἀεὶ φανερὸν ὅτι τῆς πάσης γῆς οὐκ ἀεὶ τὰ αὐτὰ τὰ μέν ἐστιν θάλαττα τὰ δ’ ἤπειρος, ἀλλὰ μεταβάλλει τῷ χρόνῳ πάντα.

130-148 WaterUrban Pt1 Ch06 Koutsoyiannis.indd 134

12/05/2014 11:20

Water Control in Ancient Greek Cities135

But if rivers are formed and disappear and the same places were not always covered by water, the sea must change correspondingly… And if the sea is receding in one place and advancing in another it is clear that the same parts of the whole earth are not always either sea or land, but that all changes in course of time. (Meteorologica II.14, 353a 16)

He also understood by experiment that salt contained in water is not evaporated (Meteorologica II.3): ὅτι δὲ γίγνεται ἀτμίζουσα πότιμος καὶ οὐκ εἰς θάλατταν συγκρίνεται τὸ ἀτμίζον, ὅταν συνιστῆται πάλιν, πεπειραμένοι λέγωμεν. Salt water when it turns into vapour becomes drinkable [fresh water] and the vapour does not form salt water when it condenses again; this I know by experiment. (Meteorologica II.3, 358b)

This has certainly found technological application in desalination (removal of salt from sea water), useful in a country with a scarcity of fresh water but with many shores and islands. Thus, we learn from a commentary on Aristotle’s Meteorologica II, written by Olympiodorus (the peripatetic philosopher who lived in the fifth century ad), that: Sailors, when they labour under a scarcity of fresh water at sea, boil the sea-water, and suspend large sponges from the mouth of a brazen vessel, to imbibe what is evaporated, and in drawing this off from the sponges, they find it to be sweet [fresh] water. (Morewood, 1838; see also quotation by Alexander of Aphrodisias in Forbes, 1970)

Hero (Heron) of Alexandria (b. c.150 bc) is mostly known as an engineer, but his comprehension of physics is very advanced, as evidenced from his treatise Pneumatica. The following extract is characteristic: Vessels which seem to most men empty are not empty, as they suppose, but full of air. Now the air, as those who have treated of physics are agreed, is composed of particles minute and light, and for the most part invisible. If, then, we pour water into an apparently empty vessel, air will leave the vessel proportioned in quantity to the water which enters it. This may be seen from the following experiment. Let the vessel which seems to be empty be inverted, and, being carefully kept upright, pressed down into water; the water will not enter it even though it be entirely immersed: so that it is manifest that the air, being matter, and having itself filled all the space in the vessel, does not allow the water to enter. Now, if we bore the bottom of the vessel, the water

130-148 WaterUrban Pt1 Ch06 Koutsoyiannis.indd 135

12/05/2014 11:20


A History of Water

will enter through the mouth, but the air will escape through the hole […] Hence it must be assumed that the air is matter. The air when set in motion becomes wind (for wind is nothing else but air in motion), and if, when the bottom of the vessel has been pierced and the water is entering, we place the hand over the hole, we shall feel the wind escaping from the vessel; and this is nothing else but the air which is being driven out by the water. It is not then to be supposed that there exists in nature a distinct and continuous vacuum, but that it is distributed in small measures through air and liquid and all other bodies. […] Winds are produced from excessive exhalation, whereby the air is disturbed and rarefied, and sets in motion the air in immediate contact with it. This movement of the air, however, is not everywhere of uniform velocity: it is more violent in the neighbourhood of the exhalation, where the motion began. (English translation by Bennet Woodcroft; see also Koutsoyiannis et al., 2007)

Technology Technological applications to solve practical problems related to water storage, transfer, and management anticipated the development of scientific knowledge. They are seen in prehistory, in several civilizations in Mesopotamia, Egypt, India, and Greece (Mays et al., 2007). A few examples of hydraulic engineering achievements running through the different stages of Greek antiquity are discussed below. Some of the earliest and most impressive hydraulic constructions of the prehistoric era are met in Minoan cities in Crete. The findings include aqueducts made of clay pipes, sewer facilities, and street drains (Figure 6.2). Water was collected either from groundwater exploitation (Knossos, Zakros, Palekastro), or from springs combined with aqueducts and/or cisterns. Wells and cisterns for rainwater collection were also in use (Phaestos, Chamaizi). In the Palace of Knossos, water was conveyed from springs at distances of 700 m to 5 km, using terracotta pipes. The conic shape of the pipes is hydraulically interesting and the reasons justifying it have become the subject of speculations (Buffet and Evrard, 1950; Angelakis et al., 2012). Equally impressive were the rainwater drains. The nineteenthcentury Italian physiologist Angelo Mosso (1907), during a visit in Aghia Triadha, wrote: One day, after a heavy downpour of rain, I was interested to find that all the drains acted perfectly, and I saw the water flow from the sewers through which a man could walk upright. I doubt if there is any other instance of a drainage system acting after 4000 years.

Similar urban sewer systems have also been found in the island of Thera (Santorini) and other prehistoric sites of the Aegean civilization

130-148 WaterUrban Pt1 Ch06 Koutsoyiannis.indd 136

12/05/2014 11:20

Water Control in Ancient Greek Cities137

Figure 6.2.  Parts of the sanitary and storm sewer systems in Agia Triadha (Angelakis et al., 2005).

(c.3200–1100 bc). At least five terracotta bathtubs were found in excavations, and these must have been in use until the great eruption of the Thera volcano around 1600 bc (chronology by Friedrich et al., 2006). One was in a room that must have been a bathroom, equipped with an advanced sewage system (Marinatos, 1999). In Delos, another island of the Cyclades, where important remains of that period have been found, the water supply largely depended on rainwater collected and stored in cisterns. Most houses on that island had underground cisterns in their yards for stormwater storage. In mainland Greece, the Mycenaean civilization depended largely on agricultural production. In order to cover the increased water needs for agriculture (even in modern Greece, about 85 percent of total water consumption is used for irrigation), Mycenaeans chose closed river basins for their settlements, and developed flood control and drainage infrastructures on an amazingly large scale (Koutsoyiannis et al., 2012). The prosperity of areas like the Arcadian and Boeotian Orchomenos is directly connected to the successful operation of these projects. Coming to the archaic and classical era, the most famous hydraulic work of ancient Greece was the aqueduct of ancient Samos, which was admired

130-148 WaterUrban Pt1 Ch06 Koutsoyiannis.indd 137

12/05/2014 11:20


A History of Water

in antiquity (e.g. by Herodotus) as well as in modern times. The most amazing part of the aqueduct is the Εὐπαλίνειον ὄρυγμα (Eupalineion orygma), or “Eupalinean digging” (after Eupalinos, an engineer from Megara), a tunnel 1,036 m long dug from two openings. Its construction started in 530 bc, during the tyranny of Polycrates, and it took ten years to complete. It is the first known deep tunnel in history (shallow tunnels are much easier to construct—using qanat technology). Like in modern construction practice, Eupalinos started from two openings (N and S) and the two construction lines met in point E (Figure 6.3). Eupalinos certainly had a good working knowledge of geometry and geodesy to carve segments of the same straight line from two openings in a mountainous terrain. There is evidence that he solved the problem with simple means and in an accurate manner by putting poles up over the mountain along the path in a straight line (Figure 6.3). Then he lined up the workers in the tunnel segments with these poles (Koutsoyiannis et al., 2008). At places, Eupalinos abandoned the straight-line course. At point A he left the straight line NA and followed the direction AB; a plausible explanation for this is that he found a natural fracture or rift and, broadening this, he was able to proceed faster. He found a clever geometrical way to eliminate the impact of uncertainty in position and direction (magnified due to already having abandoned the straight-line route) and to ensure the intersection of the two construction lines: by deliberately abandoning the straight line routes at points D and F, and changing direction toward left

Figure 6.3.  Section and plan of the Eupalinean tunnel (sketch adapted from Koutsoyiannis et al., 2008).

130-148 WaterUrban Pt1 Ch06 Koutsoyiannis.indd 138

12/05/2014 11:20

Water Control in Ancient Greek Cities139

and right, respectively, he made it mathematically certain that the two lines would intersect. Eupalinos devised an especially smart engineering solution to balance the construction needs with the physical properties of water flow. The choice of a horizontal main tunnel was dictated by the technological means of the time (a sloping one would be impossible to construct from two sides). Evidently, he was aware of the hydraulic principle that water needs a gradient to flow. Thus, starting from the horizontal tunnel, he achieved the necessary gradient by excavating a sloping channel along one side of the floor. In places where, due to the slope, the channel would be very deep, a second small tunnel below the main tunnel was built. In Athens, the first major hydraulic project was constructed under the tyrant Peisistratos (in power between 546 and 527 bc) and his sons, and has thus been known as the Peisistratean aqueduct. The largest part of the aqueduct was carved as a tunnel at depths reaching 14 m. In the bottom of the tunnel (or channel), a pipe made of ceramic sections was placed. Unlike the Minoan pipes, which had a conic shape, the pipe sections of the Peisistratean aqueduct were cylindrical and had elliptic openings in their upper part, covered by ceramic covers, for their cleaning and maintenance. Other aqueducts were also constructed with similar technologies in several phases, forming a network of pipelines. One of them, the Hymettus aqueduct, follows a route parallel to the Peisistratean (Chiotis and Chioti, 2012). The last and longest one (25 km), the Hadrianean aqueduct, was constructed in Roman times. One of the oldest aqueducts (it is not clearly identified whether it is the Peisistratean or the Hymettus) is still in operation, providing irrigation water to the National Garden, in the center of Athens. The Hadrianean aqueduct provided drinking water up to the mid-twentieth century. In addition to large-scale aqueducts, Athens had numerous small-scale constructions, such as wells for groundwater exploitation and cisterns receiving rainwater from roofs. In several cases, such small-scale constructions were interconnected, forming complex systems that stored groundwater and rainwater. Those small-scale technologies have survived up to the present date in small villages and cottages around Greece. Hydraulic systems for flood control and energy utilization, such as dams, draining systems, and watermills, were also built in the same period. An example is the Great Drain of the Athenian Agora (Figure 6.4) built in the early fifth century bc, and which still drains the Agora today. Another impressive example is the Alyzia Dam (Figure 6.5), most likely built in the classical period to protect the downstream plain from floods and sediments (Zarkadoulas, 2005; Koutsoyiannis et al., 2008). It is in perfect condition 2,500 years after its construction, thanks to its carvedin-stone spillway, which protected the dam from overtopping, and is still in operation.

130-148 WaterUrban Pt1 Ch06 Koutsoyiannis.indd 139

12/05/2014 11:20


A History of Water

Figure 6.4.  The Great Drain of the Athenian Agora (Chiotis and Chioti, 2012).

The hygienic systems in ancient Greece were simple but functional. The lavatories were supplied with natural running water by large conduits and were connected to sewers. Unlike toilets found in Minoan palaces, which resembled modern ones in several aspects, including privacy of use (Angelakis et al., 2005), Greek lavatories in the classical era and later used to be public—discouraging privacy. Figure 6.6 depicts a small-scale example, but the seating capacity could be much larger: in the Athenian Roman agora, it reached 65 (Lang, 1968). The reason for this is not obvious, but perhaps it is related to the pursuit of humility as implied by Antiphanes (writer of Attic comedy, c.408–334 bc), who suggests: Whoever thinks he’s more than human, going to the public latrine, will see himself just like everyone else. (Translation by Lang, 1968: 27)

The Hellenistic era was a long-lasting period of peace in which destroyed cities and infrastructures were rebuilt, usually on the same site as their predecessors. This period is characterized by significant scientific progress, which triggered technological innovations including the

130-148 WaterUrban Pt1 Ch06 Koutsoyiannis.indd 140

12/05/2014 11:20

Water Control in Ancient Greek Cities141

Figure 6.5.  The Alyzia Dam in western Greece. On the right is the spillway, with its irregular shape formed by erosion through centuries; the photo was taken in summer, when the stream has no water (courtesy of N. Zarkadoulas).

construction of pipelines under pressure. For the first time in history, the pressurized flow was applied on a large technological scale for water conveyance. An excellent example of this is an aqueduct (Madradag) in the city of Pergamon, located on top of a hill, 30 km inland from the Aegean Sea, in western Anatolia (now Turkey). The aqueduct includes an inverted siphon made of metal (lead) and is anchored with big stone constructions (Koutsoyiannis et al., 2008). Another important technological development of this period was the invention of various mechanisms, machines, and devices, including the windmill and the first steam engine in history (by Hero of Alexandria). A prominent example that is in use even today (Figure 6.7) is Archimedes’ helix or water-screw, the first pump with the modern meaning of the term. Archimedes was a Syracusian mathematician and engineer (c.287–212 bc), considered by many to be the greatest mathematician of antiquity or even of all time. The invention of the water-screw is tied to the study of the spiral, for which Archimedes wrote a treatise entitled On Spirals in 225 bc. This pump is an ingenious device functioning in a simple and elegant manner by rotating an inclined cylinder bearing helical blades around its axis, whose bottom is immersed in the water to be pumped. As the screw

130-148 WaterUrban Pt1 Ch06 Koutsoyiannis.indd 141

12/05/2014 11:20


A History of Water

Figure 6.6.  Sketch of a small lavatory in the Gymnasium of Minoa on Amorgos; it was built contemporarily with the Gymnasium during the mid-fourth century bc (Antoniou, 2007).

turns, water is trapped between the helical blades and the walls, thus it rises up the length of the screw and drains out at the top. Legislation and public institutions for water Apart from technological solutions for water issues, Greek societies developed a framework of laws and institutions for water management. The first known water control regulations of the Athenian city-state were made by Solon, the Athenian statesman and poet of the late seventh and early sixth century bc, who was elected archon in 594 bc and shaped a legal system by which he reformed the economy and politics of Athens. Most of his laws were later described by Plutarch (c.ad 47–127), from whom we learn about the following regulations related to water supply:

130-148 WaterUrban Pt1 Ch06 Koutsoyiannis.indd 142

12/05/2014 11:20

Water Control in Ancient Greek Cities143

ἐπεὶ δὲ πρὸς ὕδωρ οὔτε ποταμοῖς ἐστιν ἀενάοις οὔτε λίμναις τισὶν οὔτ᾽ ἀφθόνοις πηγαῖς ἡ χώρα διαρκής, ἀλλ᾽ οἱ πλεῖστοι φρέασι ποιητοῖς ἐχρῶντο, νόμον ἔγραψεν, ὅπου μέν ἐστι δημόσιον φρέαρ ἐντὸς ἱππικοῦ, χρῆσθαι τούτῳ: τὸ δ᾽ ἱππικὸν διάστημα τεσσάρων ἦν σταδίων: ὅπου δὲ πλεῖον ἀπέχει, ζητεῖν ὕδωρ ἴδιον: ἐὰν δὲ ὀρύξαντες ὀργυιῶν δέκα βάθος παρ᾽ ἑαυτοῖς μὴ εὕρωσι, τότε λαμβάνειν παρὰ τοῦ γείτονος ἑξάχουν ὑδρίαν δὶς ἑκάστης ἡμέρας πληροῦντας: ἀπορίᾳ γὰρ ᾤετο δεῖν βοηθεῖν, οὐκ ἀργίαν ἐφοδιάζειν. Since the area is not sufficiently supplied with water, either from continuous flow rivers, or lakes or rich springs, but most people used artificial wells, [Solon] made a law, that, where there was a public well within a hippicon, that is, four stadia [4 furlongs, 710 m], all should use that; but when it was farther off, they should try and procure water of their own; and if they had dug ten fathoms [18.3 m] deep and could find no water, they had liberty to fetch a hydria [pitcher] of six choae [20 litres] twice a day from their neighbours; for he thought it prudent to make provision against need, but not to supply laziness. (Plutarch, Solon, 23)

This law, among other targets, aimed to balance the public and private interests for the construction and operation of wells, and seems to have been kept during the whole history of the Athenian city-state.

Figure 6.7.  A series of Archimedes’ water-screws in their modern form (in which the walls are not attached to the screw), as implemented in the waste water treatment plant of Athens, which pumps 1 million m3 of waste water per day.

130-148 WaterUrban Pt1 Ch06 Koutsoyiannis.indd 143

12/05/2014 11:20


A History of Water

Consequently, in addition to large-scale public works (mainly aqueducts as described above, fountains, etc.), smaller-scale private installations like wells and cisterns were necessary particularly in times of war and crisis. It may thus be hypothesized that there were regulations forcing people to maintain the wells at a good condition and ready to use. This hypothesis is supported (cf. Koutsoyiannis et al., 2008) by the following passage from Aristotle: ὑδάτων τε καὶ ναμάτων μάλιστα μὲν ὑπάρχειν πλῆθος οἰκεῖον, εἰ δὲ μή, τοῦτό γε εὕρηται διὰ τοῦ κατασκευάζειν ὑποδοχὰς ὀμβρίοις ὕδασιν ἀφθόνους καὶ μεγάλας, ὥστε μηδέποτε ὑπολείπειν εἰργομένους τῆς χώρας διὰ πόλεμον. … and [the city] must possess if possible a plentiful natural supply of pools and springs, but failing this, a mode has been invented of supplying water by means of constructing an abundance of large reservoirs for rainwater, so that a supply may never fail the citizens when they are debarred from their territory by war. (Aristotle, Politics, 7, 1330b; translation from http://hydra.perseus.tufts.edu)

Regulations also dealt with cases of flood damages: From Demosthenes’ speech, Against Kallikles, which refers to property damage after heavy rain and flooding, we can infer that there was a law penalizing anyone respon­ sible for a manmade obstruction to natural flow of water that caused damage to someone else’s property (penalty 1,000 drachmas or else forfeit of the land on which the obstruction stood; MacDowell, 1986; Krasilnikoff, 2002). Protection from pollution was another issue regulated by law, as evidenced from an epigraph of about 440 or 420 bc. The epigraph contains the “law for tanners”, according to which: [No one] was to soak skins in the Ilissos [a river that ran outside the defensive walls of Athens] above the precinct of Herakles, nor to dress hides, nor to [throw rubbish?] into the river. (MacDowell, 1986)

An impressive example of regulation for sustaining a minimum river flow, most probably for environmental reasons, survives owing to fifth-century bc epigraphic evidence (Davies, 1996) from the city of Gortyn, Crete. As the city is crossed by the Lithaios River (today called Mitropolianos), which dominates the valley of Messara, the law prohibited too-large water abstraction from the river, by imposing a minimum width of the river flow: Θιοί· τô ποταμô αἴ κα κατὰ τὸ μέττον τὰν ῥοὰν θιθῆι ῥῆν κατὰ το Ϝὸν αυτô, θιθεμένōι ἄπατον ἤμην. Τὰν δὲ ῥοὰν λείπεν ὄττον κατέκει ἀ ἐπ’ ἀγορᾶι δέπυρα ἤ πλίον, μεῖον δὲ μὴ. Gods. If anyone makes the flow of the river run from the middle of the river towards his own [property], it is without penalty for the person so doing.

130-148 WaterUrban Pt1 Ch06 Koutsoyiannis.indd 144

12/05/2014 11:20

Water Control in Ancient Greek Cities145

[He is] to leave the flow as wide as the bridge that the agora holds, or more, but not less. (Davies, 1996)

Appropriate institutions completed the legislation framework and facilitated its implementation. In Athens a distinguished public administrator, called κρουνῶν ἐπιµελητής (krouno ¯n epimele¯te¯s) or “Superintendent of Fountains”, was appointed to operate and maintain the city’s water system, to monitor enforcement of the regulations, and to ensure the fair distribution of water. From Aristotle (Athenaion Politeia, 43.1), we learn that this officer was one of the few that were elected by voting, whereas most other officers were chosen by lot; an interpretation is that this position was particularly important within the governance system of Athens. Themistocles himself (“the man most instrumental in achieving the salvation of Hellas” from the Persian threat, as Plutarch (The Life of Themistocles, 7.3) describes him; cf. the Battle of Salamis) had served in this position. In 333 bc, the Athenians awarded a gold wreath to the Superintendent of Fountains, Pytheus, because he had restored and maintained several fountains and aqueducts. Generally, private sponsoring of public hydraulic systems was encouraged. The entire regulatory and management system of water in Athens must have worked very well and approached what today we call sustainable water management. Present-day water resource policymakers and hydraulic engineers have emphasized the usefulness of non-structural measures in urban water management and the importance of small-scale structural measures like domestic cisterns (Koutsoyiannis et al., 2008). A proper legal and institutional framework is also necessary for the structural part of water management, and in particular for regulating the construction process of public works. It was a common practice in ancient Greece that competition announcements, project specifications, and project contracts were written on marble stelae erected in public sites, so that everyone could see and know all project details and, simultaneously, make breach of contract difficult. An interesting example (see Koutsoyiannis and Angelakis, 2007) is the contract for draining and exploitation of Ptechae Lake (probably identified with Dystos Lake in Southern Euboea). The stele (second half of the fourth century bc) was revealed in Chalkis (1860) and is kept in the Archaeological Museum of Athens. The project is what we call today BOT: Build, Operate, Transfer. The rather wordy contract (like those of the present day) is written on a Pentelian marble stele. On the surface, relief sculptures show the gods that were worshipped in the region: Apollo, Artemis, and Leto. A carved scripture in 66 verses, signed by more than 150 people, contains the construction contract. The main contract is contained in the first 35 verses (see Koutsoyiannis and Angelakis, 2007) and is made between the city of the Eretrians,

130-148 WaterUrban Pt1 Ch06 Koutsoyiannis.indd 145

12/05/2014 11:20


A History of Water

representing the 31 municipalities of the Eretrian region, and the contractor, Chairephanes. According to the contract, the draining works include the construction of drainage canals, sewers, and wells for the drainage of water to natural underground holes or cracks, and miscellaneous protection works including wooden or metallic railings. In addition, irrigation works, such as the construction of a reservoir with side lengths of up to two stadia (360 m), for storing irrigation water, and sluice gates, are included. A four-year construction period is agreed, which could be extended in the case of war. The contractor is granted the right to exploit the dried fields for ten years (extended in the case of war), which starts when the drying has finished. He is also granted the privilege of customs-free import of materials (stones and wood). On the other hand, he is obliged to: pay all labor costs without any charge for the people Eretria; pay the amount of 30 talents in monthly installments as a rental for the permission to exploit the lake for ten years; maintain all works for the exploitation period, in order that the lake is in good condition after the finishing of the contract; compensate the landowners by one drachma per foot of land area that is to be expropriated for the construction of works; and avoid harm on private property as much as possible by locating the works in non-cultivated areas. In the case of death of the contractor, his heirs and collaborators would replace him in transactions with the city. Penalties are enforced against any person trying to annul the contract. Finally, the contractor is obliged to submit a good construction guarantee up to the amount of 30 talents. The contract is followed by two resolutions of the parliament, by which asylum is granted to Chairephanes and his collaborators for the whole duration of the contract, and the keeping of the contract is confirmed by oath to Apollo and Artemis. In light of the above evidence about the legal and institutional framework of ancient Greece, and given the problems in similar issues in modern Greek society (e.g. Koutsoyiannis, 2011), a citizen of the present day can hardly be proud of the contemporary achievements. Neither the balance of private and public sector, of small- and large-scale structures, of structural and non-structural measures, nor transparency and decisionmaking procedures are currently as advanced as they were more than two thousand years ago in democratically organized Greek cities. Conclusion The wisdom and the technical achievements of the ancient Greeks to handle the power of, and the need for, water offer us a useful, generalpurpose lesson: abundance may make us feel happy and leisured, but scarcity, and problems springing from it, trigger knowledge discovery, progress, and evolution. The importance of seeking harmony among

130-148 WaterUrban Pt1 Ch06 Koutsoyiannis.indd 146

12/05/2014 11:20

Water Control in Ancient Greek Cities147

different aspects and parts, including in water management, was a highlight in ancient Greece. Thus the use of small-scale infrastructures, in parallel to the large-scale ones, was a big step towards sustainability and resilience. Harmonization of public and private interests and functions may perhaps represent an optimal policy for water control. In retrospect, the prehistory and history of Greece provides a clear indication that scientific progress was an important agent for water control, but it seems to be neither a necessary nor a sufficient condition. Perhaps common sense and a functional social organization are more important than scientific knowledge. Apparently, ideas, technologies, and practices developed during all periods of Greek civilization greatly influenced our contemporary body of knowledge; as Will Durant (1939: vii–viii) put it: “Excepting machinery, there is hardly anything secular in our culture that does not come from Greece.” References Angelakis, A. N., D. Koutsoyiannis, and P. Papanicolaou (2012). “On the Geometry of the Minoan Water Conduits”, 3rd IWA Specialized Conference on Water & Wastewater Technologies in Ancient Civilizations, pp. 172–7, Istanbul, Turkey: International Water Association. Angelakis, A. N., D. Koutsoyiannis, and G. Tchobanoglous (2005). “Urban wastewater and stormwater technologies in ancient Greece”, Water Research, 39(1), pp. 210–20. Antoniou, G. P. (2007). “Lavatories in ancient Greece”, Water Science and Technology: Water Supply, 7(1), pp. 155–64. Buffet, B. and R. Evrard (1950). L’Eau Potable à Travers Les Âges, Liege, Belgium: Editions Soledi. Chiotis, E. D. and L. E. Chioti (2012). “Water supply of Athens in the antiquity”, in A. N. Angelakis, L. W. Mays, D. Koutsoyiannis, and N. Mamassis (eds), Evolution of Water Supply Throughout the Millennia, London: IWA Publishing. Davies, J. K. (1996). “Deconstructing Gortyn: when is a code a code?”, in L. Foxhall and A. D. E. Lewis (eds), Greek Law in its Political Setting: Justifications Not Justice, pp. 33–56, Oxford: Oxford University Press. Doxiadis, C. (1964). “The ancient Greek city and the city of the present”, Ekistics, 18(108), pp. 346–64. Durant, W. (1939). The Life of Greece (The Story of Civilization, Part II), New York: Simon & Schuster. Forbes, R. J. (1970). A Short History of the Art of Distillation, Leiden, Netherlands: Brill. Friedrich, W. L., B. Kromer, M. Friedrich, J. Heinemeier, T. Pfeiffer, and S. Talamo (2006). “Santorini eruption radiocarbon dated to 1627–1600 B.C.”, Science, 312(5773), p. 548. Koutsoyiannis, D. (2011). “Scale of water resources development and sustainability: small is beautiful, large is great”, Hydrological Sciences Journal, 56(4), pp. 553–75.

130-148 WaterUrban Pt1 Ch06 Koutsoyiannis.indd 147

12/05/2014 11:20


A History of Water

Koutsoyiannis, D. and A. N. Angelakis (2007). “Agricultural hydraulic works in ancient Greece”, in S. W. Trimble (ed.), Encyclopedia of Water Science, 2nd edn, pp. 24–7, CRC Press. Koutsoyiannis, D., N. Mamassis, A. Efstratiadis, N. Zarkadoulas, and Y. Markonis (2012). “Floods in Greece”, in Z. W. Kundzewicz (ed.), Changes of Flood Risk in Europe, pp. 238–56, Wallingford: IAHS Press. Koutsoyiannis, D., N. Mamassis, and A. Tegos (2007). “Logical and illogical exegeses of hydrometeorological phenomena in ancient Greece”, Water Science and Technology: Water Supply, 7(1), pp. 13–22. Koutsoyiannis, D., N. Zarkadoulas, A. N. Angelakis, and G. Tchobanoglous (2008). “Urban water management in Ancient Greece: legacies and lessons”, Journal of Water Resources Planning and Management—ASCE, 134(1), pp. 45–54. Krasilnikoff, J. A. (2002). “Water and farming in classical Greece: evidence, method and perspectives”, in J. E. Skydsgaard and K. Ascani (eds), Ancient History Matters: Studies Presented to Jens Erik Skydsgaard on his Seventieh Birthday, L’Erma Di Bretschneider, Rome. Lang, M. (1968). Waterworks in the Athenian Agora, Princeton, NJ: American School of Classical Studies at Athens. MacDowell, D. M. (1986). The Law in Classical Athens, Cornell University Press, New York. Marinatos, M. (1999). “Excavations at Thera VI–VII, 1972–73 Seasons”, The Archaeological Society at Athens Library, No. 180, 2nd edn, Athens. Mays, L. W., D. Koutsoyiannis, and A. N. Angelakis (2007). “A brief history of urban water supply in antiquity”, Water Science and Technology: Water Supply, 7(1), pp. 1–12. Morewood, S. (1838). A philosophical and statistical history of the inventions and customs of ancient and modern nations in the manufacture and use of inebriating liquors: with the present practice of distillation in all its varieties: together with an extensive illustration of the consumption and effects of opium, and other stimulants used in the East, as substitutes for wine and spirits year, W. Curry, jun. and company, and W. Carson, London. Mosso, A. (1907). Escursioni nel mediterraneo e gli scavi di Creta, Milan, Italy: Treves (English translation, The Palaces of Crete and their Builders, New York, 1908). Thomson, G. D. (1949). The Prehistoric Aegean, London: Lawrence & Wishart. Zarkadoulas, N. (2005). “The Dam of Ancient Alyzia (Το φράγμα της αρχαίας Αλυζίας)”, postgraduate thesis, Dept. of Water Resources, Hydraulic and Maritime Engineering, National Technical University of Athens, Athens, Greece (in Greek).

130-148 WaterUrban Pt1 Ch06 Koutsoyiannis.indd 148

12/05/2014 11:20


Plumbing Ancient Rome

Katherine Rinne Introduction If you had stood in the Roman Forum with Emperor Constantine (ad 303–37), surrounded by gleaming marble buildings, you might not have known that the beautifully paved and level public square beneath your feet concealed a complex history of environmental transformation. You, like many fourth-century Romans, might have been unfamiliar with the subterranean landscape where several small streams had once merged together in a bowl-shaped valley—a wild and water-rich terrain that migratory people had already inhabited on a seasonal basis for millennia. Instead of monumental buildings and smooth paving stones in every direction, those early inhabitants would have been surrounded on nearly all sides by rugged plateaus and ravines, deeply carved by perennial streams that flowed from the northeast and east. Situated in the flood plain of the Tiber River, their valley, which we now call the Forum Valley, stood only slightly higher than the river itself. This chapter will review what we know of Rome’s environmental setting before its mythic founding in 753 bc, and summarize its physical transformation over 1,000 years through massive engineering projects that first drained the land in order to create a city; only to then rehydrate the same area with aqueduct water so that the city could expand. We will then briefly examine some of the ramifications that these strategies had for Rome’s urban development during the medieval and Renaissance periods. A place rich in springs During the Pliocene (between 5 million and 2.5 million years ago), the Mediterranean Sea submerged the entire area of what would become Rome, including the plateaus surrounding the Forum Valley and much of the surrounding countryside now known as the Campagna. It was only during the last glacial age when sea levels subsided dramatically, about 10,000 years ago, that the topography of marine, volcanic, and alluvial

149-172 WaterUrban Pt1 Ch07 Rinne.indd 149

12/05/2014 11:21


A History of Water

Figure 7.1.  A diagram showing the major streams that flowed in prehistoric Rome and its immediate region. The dashed lines indicate flowing water and the double dashed line is the Tiber River flowing from the north (top). The Tiber Island is completely submerged; the Capitoline is shown as two small hills; and the Palatine is an island. As the water receded, the Tiber Island emerged and land bridges appeared: to connect the two Capitoline islands with each other and also to the Quirinal Hill; and to connect the Palatine to the Esquiline. Source: “Stato delle Acque nell’Epoca Diluviana”, Giuseppe Ponzi, Sulla Storia Fisica del Bacino di Roma (1867), pl. 1.

149-172 WaterUrban Pt1 Ch07 Rinne.indd 150

12/05/2014 11:21

Plumbing Ancient Rome151

soils that we now recognize as the stratigraphy of Rome began to emerge (Facenna et al., 1995: 45–6) (Figure 7.1). By the time of Rome’s mythic foundation in 753 bc, the topography and hydrography of the area was distinguished by a series of plateaus separated by steep ravines, small densely vegetated valleys, and a broad plain through which the Tiber, “the river closest to god”, had relentlessly carved its channel, depositing a bed of alluvial soil 60 m deep. The Tiber is often swift flowing, even in the summer, and it carries a heavy sediment load. Although it is a relatively short river by global standards—only 343 km long—there are more than 40 tributaries above Rome. Originat­ ing at Monte Fumaiolo in the Apennines, the Tiber’s watershed covers 17,156 km2. Nearly half of this area is characterized by relatively impermeable rock including marls, serpentinite, and shale. This increases the amount of runoff flowing into the river during major storms, and amplifies flooding events, which have been frequent and fierce (Frosini, 1977: 26–7). Rome’s foundation myth begins with a flood, with Romulus and Remus washed ashore at the foot of the Palatine Hill, where Romulus later established his new city. According to Cicero, Romulus chose this place because it was rich in springs—at least 23 springs of varying capacity are known to have originated within the area now contained by Rome’s historic walls.1 In turn, the springs fed streams—most of these were small, but a few had a sizable flow (Corazza and Lombardi, 1995: 179–211). The streams acted as territorial boundaries, creating defensible locations atop each ridge that connected to the plateau by a land spur. One of these ridges, the Palatine—almost completely girded by streams at its base—connected to the Esquiline plateau. One small stream, now called the Forum Brook, originated at a spring on the slope of the Esquiline Hill. Relatively shallow, the brook had, over millennia, carved a deep channel through the tufa deposits that formed the plateau and then flowed across the Forum Valley to the Tiber. Other nearby springs, including the Juturna, Lautole, and Tullianum, emerged from the surrounding Quirinal, Viminal, Velia, Palatine, and Capitoline hills, feeding their own smaller streams that joined the Brook in the valley. Once united, they flowed west through the Velabrum, a nearly flat alluvial plain that connected the Forum Valley to the Tiber River. The elevation of the entire Velabrum was very low—barely rising above the level of the Tiber, which flooded often and created seasonal fresh water marshes in this area (Holland, 1961: 4–7, 21–2) (Figure 7.2). Originally settled during the Iron Age (long before Romulus and Remus), shepherds and their flocks visited the valleys, streams, and Roman hills on a seasonal basis. Each winter they migrated down from the Alban Hills and Apennine Mountains to the relatively warm and lush meadows, marshes, and alluvial plains of central Italy, including into the area that would become Rome. Simple huts of reeds and mud, duck blinds for fowling and hunting, and other temporary (or perhaps portable)

149-172 WaterUrban Pt1 Ch07 Rinne.indd 151

12/05/2014 11:21


A History of Water

Figure 7.2.  A hypothetical reconstruction of the topography on ancient Rome (detail, Lanciani, Forma Urbis Romae, 1893–1901), shown with an overlay of springs and seasonally marshy areas. The black dots represent documented ancient springs. The seasonal marshes were: 1) Palus Caprae in the Campus Martius; 2) Forum Valley; 3) Velabrum; 4) Colosseum Valley; 5) Circus Maximus Valley; and 6) the Codetta in the Transtiberim.

platforms for storing food were among the few structures to be seen. In the harsh hot summers, the shepherds returned with their animals to the cool mountains where an abundant supply of water and fodder awaited. This kind of cyclic transhumance passage had probably continued uninterrupted for millennia (Bjur and Santillo Frizell, 2009). By the time Romulus appeared, the shepherds and their flocks had beaten paths through the willows and osiers down to the banks of the Forum Brook, where they caught fishes, trapped birds, foraged nettles and ferns, and collected basketry materials in a seasonally marshy, muddy, and dynamic landscape.

149-172 WaterUrban Pt1 Ch07 Rinne.indd 152

12/05/2014 11:21

Plumbing Ancient Rome153

It is also clear from archaeological evidence that modest permanent settlements already existed on the hilltops and that there were altars and shrines already built along the slopes of the hills (Gjerstad, 1953). Louise Holland was one of the first archaeologists to suggest that springs, streams, and marshes were essential components of archaic Roman city building. She drew her conclusions about the Forum Brook, in particular, by looking at similar-sized brooks that still flowed in similar geological settings in central Italy during the mid-twentieth century, when she conducted her research. She determined that perennial rivers and streams acted as natural territorial boundaries—like the Tiber flowing between the Etruscans and Latins—and that they were also defensive barriers for scores of hill towns throughout the Italic peninsula, like Veii (Figure 7.3). Living water—that is, perennial rivers and streams—was considered sacred across cultures throughout archaic Italy. Perennial water boundaries could not be crossed without first conducting priestly rituals known as auguries; this was true even during warfare. Thus, the Forum Brook, though not a particularly imposing physical obstacle, still held psychological importance. Rivers and streams were the domains of gods, while nymphs, like Egeria and Juturna, presided over springs. The gods and more important nymphs each had his or her particular devotions. Father Tiber was the most important water god in the construction of Rome’s mythical history, and the nymphs Egeria and Juturna were also part of the foundation narrative (Ovid 1.707). According to tradition, it was Egeria, the wife of Numa Pompilius (Rome’s second “king”), who brought religious practice and law to the newly founded city (Livy 1.19–21). Romulus avoided this saturated valley and founded his Latin city on the Palatine Hill overlooking the Tiber River and its island. But the Latins were not alone in the area. Etruscans controlled the Janiculum Hill across the Tiber, and the Sabines controlled the Quirinal Hill just across the valley. While it is well known that relations between the Latins, Sabines, and Etruscans were often volatile, by the early sixth century (i.e., during the so-called “Age of the Kings”, 715–509 bc) they were so politically and culturally entwined—two Etruscans and one Sabine actually served as “kings” of Rome during the early seventh to the mid-sixth century—that it became incumbent to fashion a single shared public space. They chose the Forum Valley, an inauspicious site since experience should have made evident its inherent difficulties; its low elevation, close proximity to the Tiber, and history of repeated flooding—on average, once a year (Gregori et al., 1988). But even though this was a muddy and sometimes stagnant drainage basin, it must have provided a convenient “neutral ground” to create what became Rome’s symbolic, political, and economic center— the Roman Forum. In what had once been a vital productive area with ecological complexity, its natural qualities had become nuisances and impediments to political unity. Only by obliterating this watery identity could the new Rome find a center around which it could grow and flourish.

149-172 WaterUrban Pt1 Ch07 Rinne.indd 153

12/05/2014 11:21

149-172 WaterUrban Pt1 Ch07 Rinne.indd 154

12/05/2014 11:21

Source: George Dennis (1848). Cities and Cemeteries of Etruria, Vol. 1, Map 1.

Figure 7.3.  “Map of Veii, adapted from Gell.” Veii was founded atop an irregular, easily defended plateau that was almost completely girded by streams in the valleys below.

Plumbing Ancient Rome155

Rome’s “big dig” Before the Forum Valley could be used as a civic space and outfitted with monumental buildings, three things had to happen. The ground level of the entire area had to be raised by several meters to mitigate Tiber flooding; the streams flowing through the valley needed to be channeled in order to dry out the saturated land; and bridges were needed to connect streets and pathways across the now channeled streams. These public works required a centralized administration to coordinate and control a large workforce. According to Livy and other historians, Tarquinius Priscus (614–576 bc), one of the Etruscan “kings”, sponsored construction of a drain through the Forum Valley around the year 600. Fortunately the literary tradition meshes well with archaeological evidence and this date is fairly secure. The complex details of the drainage project are now being revealed and correlated through archaeological research, providing us with fresh insights into the foundation—literally down to the native soil—of the Roman Forum. We now have a much clearer sense of the technologies and expertise that ancient Romans actively employed to drain valleys and level hills as a foundation for a complex urban environment. Taking core samples from across the Forum and the Velabrum, Albert Ammerman (1990) has proposed a compelling scenario. He describes a multi-year construction project to fill the valley with imported debris, tufa stone fragments, and soil to guard against flooding. This was done in stages until the ground level was about 3 m higher (about 9 m above sea level). This was still only a few meters above the Tiber’s normal height. Paved with gravel, the level area on both sides of the brook stretched from the foot of the Capitoline to the foot of the Palatine, and was stable enough for permanent buildings. As this work progressed, construction began on a giant drain that employed a new strategy previously unknown on the Italian peninsula (Hopkins, 2007). It was probably laid out as a temporary wooden structure, following the Forum Brook’s crooked course, which reflects the general reluctance to confront “a potentially hostile and dangerous power” (Richardson, 1992: 91) that resides within sacred flowing water. During the leveling stage, a permanent monumental stone-lined drain, the Cloaca Maxima, was built. It acted as a retaining wall to hold back the soil in the newly raised forum, channeled the brook as it flowed through the valley, and kept its water from saturating the newly created level land (Figure 7.4): Other, less demanding hydraulic techniques were in use near Rome at the time; yet the kings chose a new, complex masonry technique for the Cloaca. It remains to determine why they chose such a radically different and perhaps unnecessarily difficult engineering for their canal. (Hopkins, 2007: 8, fn. 46)

149-172 WaterUrban Pt1 Ch07 Rinne.indd 155

12/05/2014 11:21


A History of Water

Fashioned without mortar from huge blocks of pietra gabina (2.50 m long, 0.80 m high, and 1 m wide), its size was remarkable (4.50 m wide by 3.30 m high at its mouth), and its strength was noteworthy (Narducci, 1889: 40). With any body of moving water like the Tiber River or the Forum Brook, a natural ford, a narrow chasm, or a proximate path or road can over time become a designated crossing. Holland points out that as Rome’s population and influence grew, the Forum Brook was traversed more frequently, and permanently inaugurated bridges were needed at these important points. This was long before the Cloaca Maxima was built. Holland identifies four small permanent crossings, known as Janus bridges, along the brook. Rebuilt many times, the crossings continued to span the open

Figure 7.4.  Reconstruction of the early-sixth-century Cloaca Maxima. Source: Courtesy of John N. N. Hopkins (2007).

149-172 WaterUrban Pt1 Ch07 Rinne.indd 156

12/05/2014 11:21

Plumbing Ancient Rome157

channel until the late second century bc, when the Cloaca Maxima was finally vaulted and paved over. This erased the brook from Rome’s public memory and hastened the transformation of the Cloaca Maxima from a drain into a sewer (Holland, 1961: 5–6 and 45, fn. 68). Horace and other writers from the first century bc through the second century ad were unanimous in their disdain for the archaic marsh in the Forum Valley (Horace, Epistles, 1.8.8–9). To them it was “dank with waters that flooded back from the river” (Ovid, Fasti 6.401–2). Because there was so little direct evidence from which they could draw, ancient authors often projected back from their contemporary conditions in order to reconstruct an archaic topographic history for Rome. Neither Horace nor other writers of his time really knew the Forum marsh. As we have seen, it had been drained 500 years earlier, the ground leveled and raised several meters, and much of the area paved. Regardless, the ancient authors portrayed the marsh as a pestilential and malarial landscape—one without redeeming qualities and a problem to be eliminated. While the Forum area (drained, raised, leveled, and paved) could be malarial when they wrote, it is likely that the Forum marsh, at least in the early days of the settlement, was more an asset to the population than a public health liability. The native reeds, willows, and mud probably provided building materials for slaves and poorer members of the community. Ducks, kingfishers, gulls, herons, and perhaps bitterns would have been plentiful, and were probably an important component of people’s diet, along with native Allium (wild onions), nettles, ferns, and other wild greens, as well as the frogs, snails, or other small animals that called the wetland their home (Smith, 1877: 130–42). Some Tiber fish would have swum into the shallow, warm marsh to lay their eggs (Hehn, 1885: 228–31). Altogether, this was an ecologically rich and productive landscape—one that was probably vital to the lower classes. The Forum Valley was not the only area of the future city that was subject to regular flooding and seasonally marshy conditions. Nor was channeling the Forum Brook a singular situation. There were many springs feeding several streams that flowed to the Tiber River. Their names remain largely unknown, but ancient drains in the valleys of the Circus Maximus and the Colosseum, in the Campus Martius (Field of Mars), the Ager Vaticanus (Vatican and Borgo), and in the Transtiberim (Trastevere) confirm their existence (see Figure 7.1 and Figure 7.2). Bit by bit, Rome’s original watery landscape was replaced by a “densely artificial urban landscape” (Aldrete, 2007: 12) of monumental temples, baths, palaces, porticoes, and fora, all built of stone, with domestic and commercial building of brick and timber scattered in between. A brook once flowed through the long, fairly straight valley running between the Palatine and Aventine hills. It was channeled and the valley drained in the sixth century bc (shortly after the Forum Brook) in order to build the Circus Maximus, where chariot races and other spectacles

149-172 WaterUrban Pt1 Ch07 Rinne.indd 157

12/05/2014 11:21


A History of Water

were staged (Livy 1.35). Its drain, the Cloaca Circus Maximus, which ran down the middle of the valley, must have been covered over in part to facilitate racing, but most of it was left open and served as the median strip down the center of the racetrack for many centuries. It bears emphasizing that the scale of these public works projects all required a well-developed civic administration that had at its command material resources including stone; the technology to quarry it into blocks and move the blocks to the job site; and vast labor resources (in this case, slaves). This implies a level of control over peoples, landscapes, and resources that already hints at Rome’s territorial reach even before the Imperial Period. In fact, Livy remarks that Tarquin “sent for workmen from all parts of Etruria” for construction projects and also “compelled the plebeians to take their share of the work” (Livy 1.56). This burden was in addition to their military service, and ancient historians suggest that they felt it less of a hardship to build temples than to build drains. Pliny the Elder relates that: Tarquinius Priscus having commenced the sewers, and set the lower classes to work upon them, the laboriousness and prolonged duration of the employment became equally an object of dread to them; and the consequence was, that suicide was a thing of common occurrence, the citizens adopting this method of escaping their troubles. (Pliny the Elder, 36, 24)

Like Horace, Pliny had no way to understand fully Rome’s pre-urban ecology and the complexity of that landscape, and how the plebeians may have related to it (Purcell, 1996). Rather than see the displeasure of the lower classes solely as a result of working underground or being more susceptible to malaria, the work may have challenged religious scruples (Holland, 1961: 34). Additionally, I think it possible that persons from Rome’s lower classes (no doubt including slaves) were angered because they were forced to destroy close-at-hand productive landscapes that may still have provided them with food and other resources. The pattern of exploiting flood-prone areas like the Circus Valley for recreation or spectacles survived well into the Imperial Period. Augustus (27 bc–ad 14) turned his attention to draining a low-lying marshy area known as the Codeta, located across the Tiber in the Transtiberim. This area was suitable for his Naumachia Augusti—an outdoor theater, for which he built a wholly new aqueduct, the Aqua Alsietina (2 bc), in order to present mock naval battles (Richardson, 1992: 92). In another example, the emperors Caligula (ad 37–41) and Nero (ad 54–68) drained and channeled another stream to build the Circus Gaius et Neronis in the valley that ran west/east from the Mons Vaticanus to the Tiber. Whether to create a dry area for spectacles and sport, or to fill and empty the enormous Naumachia, massive drains were necessary.

149-172 WaterUrban Pt1 Ch07 Rinne.indd 158

12/05/2014 11:21

Plumbing Ancient Rome159

The Colosseum Valley was also drained. Holocene alluvial deposits brought its elevation to about 14 m above sea level, significantly higher than the Forum Valley, but still low enough for occasional flooding. Fed by spring water that flowed through the Esquiline Valley in the Labicano Brook, it was first fitted with an artificial channel in the Republican period (509–31 bc) (Funiciello et al., 1995: 928 and 933–5). Nero built a vast artificial lake in the valley known as the Stagnum of Nero. The Aqua Claudia (ad 38–52) and perhaps the Labicano Brook supplied the water. As part of an initiative to erase Nero’s memory, Emperor Vespasian drained Nero’s Stagnum and built the Flavian Amphitheater (the Colosseum). Some of the ancient drains still function, as modern drains were linked to them in the nineteenth century (Narducci, 1889: 66–7). The Campus Martius (now called the Campo Marzio) occupies the broad arc of the alluvial flood plain where the Tiber bends. Two streams, the Aqua Sallustiana and the Petronia Amnis, flowed through the Campus. The Sallustiana, fed by several abundant springs, carved the valley between the Pincian and Quirinal hills. It flowed first west, then south (although this is disputed by some). It passed through an exceptionally low area where its waters contributed to an expansive marshy area known as the Palus Caprae, which included the area where the Pantheon now stands. The Petronia Amnis originated at the Fons Cati spring (now called Acqua di San Felice), on the western face of the Quirinal Hill. It still flows from its source inside the grounds of the Quirinal Palace. The Fons Cati probably augmented the Palus Caprae, which extended all the way south to the Tiber, where the water finally debouched into the river just north of the area now occupied by the Jewish ghetto. Rodolfo Lanciani (1897: fig. 1) and Corazza and Lombardi (1995: 182, fig. 2) see both streams emptying into a permanently marshy Paulus Caprae, but never meeting. Richardson (1992: 289–90) supports two independent streams, with the Sallustiana heading west toward the Mausoleum of Augustus. Coarelli (1997: 16, fig. 2) sees the Petronia Amnis passing through a smaller Palus Caprae that runs east–west rather than north–south and then flowing southwest to the Tiber near the Chiavia di Santa Lucia. The muddy quagmire of the Campus Martius, which was public land, remained largely empty of buildings for centuries. A part of it was used as a military training ground (hence its name, the Field of Mars), its saturated land simulating extreme conditions that soldiers would encounter during campaigns. Especially in the Paulus Caprae, the Campus must have provided a fruitful environment for fishing, birding, and other activities like those that had once taken place in the Forum Valley. During the early Republic it seems that it was “regularly used as a public pasture” (Narducci, 1889: 66–7). As the city expanded, the Campus Martius became prime real estate for large private and civic structures. As in the Forum and Circus valleys, the land flooded often and needed to be drained, leveled, and in some

149-172 WaterUrban Pt1 Ch07 Rinne.indd 159

12/05/2014 11:21


A History of Water

places raised. The least saturated areas were built up first, most notably the Circus Flaminius of 221 bc, and the small Republican temples (some ruins still stand) in the Largo Argentina area that date from the second and first centuries bc. But the most impressive complex was the Porticus and Theater of Pompey of 55 bc, a large stone theater and public outdoor sculpture garden with fountains, pools, and trees. Two sets of drains were needed: one set to dry out the land and another to carry away any runoff water from the fountains and pools. The few drains that have been uncovered head to the east—that is, toward the more saturated and hence lower elevation of the Palus Caprae, which flowed south to the Tiber. Regardless of which view you adopt regarding the location and size of the Palus Caprae, Marcus Vipsanius Agrippa, son-in-law of the emperor Augustus, would have faced a major land reclamation project when he built the Pantheon, Baths of Agrippa, and an enormous outdoor pool, called the Stagnum, between 25 and 19 bc (Figure 7.5). The Aqua Virgo (19 bc), his new aqueduct, supplied the baths and pool. It is instructive to think of the aqueduct as a channeled brook that emptied directly into an area already oversupplied with water. Therefore, it was necessary not only to evacuate standing and flowing water from the Aqua Sallustiana and/or the Petronia Amnis, but also to plan how water from the Aqua Virgo would move into and through the area. The baths and the pool water would have been kept separate, each with its own drain. The overflow from the baths headed south in a closed drain, perhaps parallel to that built for the Portico of Pompey. The overflow from the Stagnum flowed in the Eripus, an open-air ornamental canal, its course heading first west and then northwest to empty into the Tiber near the Pons Neronianus. Once these public buildings were in place, multi-storey residential buildings begin to populate the eastern Campus Martius. Flooding Rome with water and waste As the number of aqueducts and the amount of water flowing into Rome increased, so too did its population and the amount of waste and debris generated. This called for dedicated sewer drains, so that those drains carrying runoff water intended for a second life in industry, laundries, irrigation, or animal fountains might remain free of human and animal waste. The Baths of Caracalla, for example, had drains for runoff water that could be reused (in one case, for grain mills located below the public areas) and others for the black water that flowed through the latrines and then directly into sewers. Purpose-built drains collected rainwater from the streets and roofs. Of necessity they were open at street level and because debris easily fell into the openings, drains transformed into sewers. In any case, runoff water, whether from streams or aqueducts, entered the drains and ultimately

149-172 WaterUrban Pt1 Ch07 Rinne.indd 160

12/05/2014 11:21

Plumbing Ancient Rome161

Figure 7.5.  Major building projects of the first century Rodolfo Lanciani (1893–1901).


in the Campus Martius.

Source: Forma Urbis Romae, Milan: Hoepli, pls. 15 and 22, with notations by the author.

flowed into the Tiber, carrying waste and debris with it. In a sense, Rome’s Tiber became an enormous sewer that had an impact on everyone living and working along its banks. This created a serious public health problem that was exacerbated in the lowest parts of the city. Throughout Rome’s intensive, centuries-long land reclamation effort, the Tiber continued to flood—reaching even into the Forum, in spite of the elevated terrace sponsored by the Tarquins. The relatively low and flat Campus Martius and the Ager Vaticanus received the first impact of the floods, and the long straight Via Lata (now the Corso) provided an effective chute to send water directly to the foot of the Capitoline Hill. The land spur that once connected that hill to the Quirinal would have diverted flood­water to the southwest, where it scoured the foot of the Capitoline’s north face

149-172 WaterUrban Pt1 Ch07 Rinne.indd 161

12/05/2014 11:21


A History of Water

then swung back around to the south. It was at the Forum Boarium (Cattle Market) located at the mouth of the Velabrum Valley just opposite the Tiber Island that flood water from the built-up areas of the Campus Martius met the swollen river. The market would have received the brunt of the turbulence before the water rushed into the Forum Valley and the Circus Valley. Rome’s floods are well documented. Cicero (3: 7) relates that in 54 bc: At Rome, and especially on the Appian Way, as far as the temple of Mars, there is a flood of remarkable size. The promenade of Crassipes has been swept away, along with pleasure grounds, and many shops. There is a great extent of water as far as the public fish-pond.

Cassius Dio (53: 33) relates that in 23 bc a Tiber flood carried away the “wooden bridge”—that is, the Pons Sublicius—and that in 13 bc, Cornelius Balbus, the patron for the newly dedicated Theater of Balbus, located in the southern end of the Campus Martius, was “unable even to enter his theatre except by boat on account of the floods caused by the Tiber” (54: 25). Not only did the floods damage buildings, but also they carried alluvium and debris that clogged the drains, causing them to back up. This allowed flood water to emerge in inland areas that might not otherwise have been affected. Alluvial deposits accumulated and ground levels rose. Streets, squares, and buildings would have been cleaned as quickly as possible, but alluvium might have been allowed to accumulate in unoccupied parts of the city. The difficulty of maintaining the drains and sewers is made explicit by Dionysus of Halicarnassus (II: 67) who noted “that once, when the sewers had been neglected and were no longer passable for the water, the censors let out the cleaning and repairing of them at a thousand talents”, not an insignificant sum.2 The rise and fall of Rome’s water supply Drains expelled extraneous water from the city as early as 600 bc, but less than 300 years later it was necessary to divert fresh water from the Roman Campagna to supplement the urban supply. The numerous fresh water springs in Rome were no longer sufficient for a growing city with expanding political and economic importance. Using technology similar to that used for the drains, Rome’s first aqueduct, the Aqua Appia of 312 bc, ran almost entirely underground and brought water to the busy Porta Trigemina on the Tiber (Rodgers, 2004: 1, 5; Evans, 2002: 65–74). The history of Rome’s fabled ancient aqueducts, their construction methods, the amount and quality of the waters that they carried, their source springs, routes across the Campagna, points of entry into the city, and their impact on population growth and urban form are well known, and therefore will not be discussed in any detail here.3

149-172 WaterUrban Pt1 Ch07 Rinne.indd 162

12/05/2014 11:21

Plumbing Ancient Rome163

It is widely recognized that the abundant supply of aqueduct water provided the impetus for Rome’s phenomenal expansion beyond the hills and valleys near to the Tiber Island and its river. As Rome increased its control over Latium, there was greater access to surrounding springs, which meant more water for the city. As Rome expanded its reach beyond the Italic peninsula and the Mediterranean, it became a magnet for immigrants, all of whom needed access to water. Between 312 bc and ad 226, 11 aqueducts were constructed to serve Rome. The source springs for the earliest of these, the Appia, were located underground at a low elevation. This meant that the aqueduct was also subterranean and served only the lowest areas of the city, which at that time consisted of the area near to the port, the Circus Maximus, and the Velabrum. Subsequent aqueducts using higher source springs that were located farther away served Rome’s hills. Aqueduct arcades were often used to cross the longer distances, and some of their ruins still stand in the Campagna (Figure 7.6). The majority of the water came from springs in the Alban Hills located to the southeast of Rome, and this in turn dictated in large part where the aqueducts entered the city—that is, from the southeast where the Porta Maggiore now stands. This had an impact on urban development and water distribution.

Figure 7.6.  Ruins of the Aqua Claudia crossing the Roman Campagna outside Rome. Source: Author’s photograph.

149-172 WaterUrban Pt1 Ch07 Rinne.indd 163

12/05/2014 11:21


A History of Water

Rome’s earliest aqueducts—the Appia (312 bc), Anio Vetus (272), Marcia (144), Tepula (125), and Julia (33)—served large public buildings and facilities, and their water also flowed to numerous public fountains that were available to all persons—at the port, in the fora, adjacent to temples, in the markets, and at street corners. The Aqua Virgo (19 bc) was a specialized line—sponsored by Agrippa to support the urban vision of the emperor Augustus (Evans, 2002: 137). It was the first aqueduct built specifically to serve an imperial bath. After the Virgo came five more aqueducts: the inferior Aqua Alsietina (2 bc), which served Augustus’ Naumachia in the Transtiberim; the Claudia and Anio Novus (both ad 38–52); the Traiana (109); and the Alexandrina (226). Of 11 aqueducts, only the Virgo, Alsietina, and Traiana entered Rome from the north. Agrippa’s Baths were part of a large imperial complex of public buildings in the Campus Martius: the Pantheon (27–25 bc); the Saepta Julia (a voting place planned by Caesar and completed by Agrippa in 26 bc); the Basilica Neptuni (its precise use unknown, but connected to the Baths and perhaps also dating from 19 bc); and beautiful and spacious gardens in which the Stagnum was located (see Figure 7.5). Because a dedicated aqueduct served Agrippa’s Baths, there was an enormous amount of water, which meant that the complex could achieve a grand scale not seen before in Rome. It also meant that new functions—specifically those associated with the Greek gymnasium—could be added to the bathing establishment. The gymnasium was “an institution for the military and athletic training of young citizens as well as for their intellectual and artistic development”, so “by introducing extensive gardens, outdoor athletic facilities, and pools”, the baths “embraced the full scope of the gymnasium idea” (Yegül, 1992: 7, 137). Its site in the Field of Mars, where war exercises had taken place for centuries, was particularly apt. Although it is unclear how Agrippa’s Baths were administered after his death in 12 bc—when he left them and the adjoining gardens to the Roman people to be enjoyed free of charge—it is clear that even poorer people had the opportunity to bathe and then stroll in a verdant open space filled with trees and statues. The setting must have been in stark contrast to the increasingly noisy city, filled with hard reverberating stone surfaces. Agrippa’s Baths ushered in a new bathing model, integrating baths with active exercise and intellectual pursuits that was copied and expanded for more than 300 years. Nero’s Baths (dedicated in ad 62) were built close to those of Agrippa and took advantage of the Stagnum. The Baths of Titus (dedicated in ad 80) stood atop the Oppian Hill. While larger than those of Agrippa, they were still modest when compared with the better-known, fully developed, later monumental Baths of Trajan (ad 109), Caracalla (completed in 216/17), and Diocletian (completed in 305/6). These last three were so extravagant that large purpose-built cisterns were needed to collect water during the night when the baths were closed to ensure that the daily supply would not run out. Constantine also built a bath

149-172 WaterUrban Pt1 Ch07 Rinne.indd 164

12/05/2014 11:21

Plumbing Ancient Rome165

complex. It, and a few other imperial baths, still functioned after he moved the capital of the Roman Empire to Constantinople in 330, but by the early sixth century they were moribund. Rome’s arcaded aqueduct structures heading across the Campagna were vulnerable to attack when Rome was under siege. Procopius, historian and advisor to Belisarius (Emperor Justinian’s chief military commander), relates that in ad 537, Visigoth armies commanded by Vitiges surrounded and attacked Rome as part of an effort to subdue the entire Italic peninsula. Having established camps at various locations outside the walls, the soldiers “tore open all the aqueducts, so that no water at all might enter the city from them” (Procopius V: 83–7). Recalling a strategy that he had used successfully to enter Naples during a siege, Belisarius had the now dry aqueducts blocked up for “a considerable distance” inside the city so that Vitiges’ soldiers could not enter into Rome through them. The Imperial Baths were rendered useless “because of the scarcity of water” (Procopius V: 190–1), and starvation loomed because the grain mills, once driven by motive force of the Aqua Traiana on the Janiculum Hill, had stopped. As a temporary measure, Belisarius had the mill wheels moved to floating barges on the Tiber River. The Goth armies were routed and Rome was saved—at least temporarily. Belisarius quickly repaired the Aqua Virgo to the Campus Martius, and also the Aqua Traiana, in order to revive and maintain the grain mills. Even so, some of the aqueducts remained useless for many years, and some were abandoned completely. Aqueducts are susceptible to structural damage and they also must be regularly cleaned, because calcium deposits build up within the channels and pipes, constricting the flow. The restoration record is incomplete, but we know that by the sixth century, the papacy had assumed control of Rome’s water infrastructure. Pope Gregory I (590–604) restored the Virgo (and perhaps the Claudia, Marcia, and Traiana) by 602, and Honorius (625–38) restored the Traiana (Duchense, 1886 I: 324). Hadrian I (772–95) restored the Virgo as far as the area now occupied by the Trevi Fountain, the Traiana to the Vatican, and the Claudia and Marcia to the Caelian and Esquiline hills. Sergius II (844–7) also restored the Marcia, and Gregory IV (827–44) restored the Traiana as far as the Janiculum Hill, again with an interest in maintaining the mills. After the Saracens plundered the Campagna and destroyed the aqueducts in 846, repairs were not carried out until the pontificate of Nicolas I (858–67) (Duchense, 1886 I: 503, 504, 510, 519 fn. 64, 522 fn. 111; and II: 91, 154, 164). The restoration record ends here (Coates-Stephens, 1998). The assumption has been that aqueduct water must have ceased to flow shortly after the last recorded papal project. The Virgo remains the exception. It continued to sputter along over the next several centuries, under papal control. Then, during the period when the popes abandoned Rome for Avignon between 1309 and 1377, civic administrators established their right to maintain it, as well as the fountains, roads, and bridges. The

149-172 WaterUrban Pt1 Ch07 Rinne.indd 165

12/05/2014 11:21


A History of Water

aqueduct was considered such an important civic monument that the 1363 Statutes of Rome devote six paragraphs to detailing how it was to be maintained (Karmon, 2005: 4, citing Re, 1880 III: 26, 264). The history of Rome’s aqueduct restorations intertwined with that of the drains, in that as water continued to flow (even in smaller quantities), drains were still needed to carry away runoff water. Therefore, even without confirmation from written documents, some of the more important drains must also have been restored and continued to function until the ninth century. After that, the seemingly unending strife between popes, nobles, and emperors, and an almost continual threat of invasion between the tenth and thirteenth centuries, left little political will, money, or labor force to carry out aqueduct or drain construction that would benefit Rome. Records for the construction or restoration of drains made during this time are, as the archaeologist Narducci (1889: 8) pointed out at the end of the nineteenth century, concealed by “a dense veil” that falls after the fifth century. A few facts are known; Leo IV (847–55) had a ditch built to divert stream water from the Monti della Creta behind the Vatican so that it would flow along the outside base of the north side of the new defensive walls that he ordered built around the Borgo Vecchio and the Vatican after the Saracen invasion (Narducci, 1889: 52). In 1230, Gregory IX restored Rome’s walls and constructed “many sections of drains” following Frederick II’s raid of the Papal States (Narducci, 1889: 8). But, because the old collector drains were in ruins, filled with rubble, or buried under meters of alluvium and debris, the new drains were built above the old ones and may not have connected with them at any point. Any private drains might flow into the larger collectors (Verdi, 1997: 155). It was easier to dig open collector drains that ran down the middle of the street, like the Chiavica di San Silvestro, which ran from the Fountain of Trevi to the Mausoleum of Augustus. Their drain mouths were therefore above the level of the Tiber rather than partially submerged, as in antiquity. This increased the amount of standing garbage, trash, and fecal matter along the riverbanks. During the medieval period, Rome’s civic administration and the papal government continued to be pawns in late medieval European power struggles. This left the city without a clear political leadership that felt compelled to care for its citizens and its public works. Parts of the Roman Forum were abandoned and, over time, the ground level rose by several meters with accumulated alluvial soils and collapsed buildings. Even in the Campus Martius, by now heavily populated, many buildings were abandoned due to structural damage. In cases like this, there would be no need to clean away debris from the flood. Also, drains continued to clog and were rendered useless over time, as they too were buried further underground. Rome’s infrastructure no longer operated at an integrated urban scale, but at fragmented and disconnected local scales, perhaps as small as a

149-172 WaterUrban Pt1 Ch07 Rinne.indd 166

12/05/2014 11:21

Plumbing Ancient Rome167

single piazza or neighborhood. A more complete discussion of Rome’s medieval and Renaissance drains has yet to be written, in part because the archival material is either very specific (about a particular drain), or very general and lacking topographic information. New information is emerging slowly through ongoing archaeological investigations, particularly at the Crypta Balbi, and in the area of the Forum Transitorium and Forum of Trajan to the east of the Roman Forum (Manacorda, 2001, 2005). By the late tenth century, when it is assumed that most of the aqueducts had ceased to function, there was very little water to control at the civic scale. This caused spatial patterns to change and Rome experienced profound urban contraction. Many people left Rome entirely and many of those who stayed moved down to the Campo Marzio to be close to the Trevi Fountain and the Tiber River, while those in the Borgo may have derived some benefit from a small fourth-century aqueduct, the Acqua Damasiana, which fed the Vatican immediately to the west. Remarkably, new monasteries were founded on the now-dehydrated hills; Santa Croce in Gerusalemme (Benedict VII: 975), Sant’Alessio al Aventino (Benedict VII: 977), SS Quattro Coronati (Pascal II: 1116), Santa Bibiana (Onorio III: 1220), and San Pancrazio (Alexander IV: 1255) among others. Small private baths were also built, at least in the Campo Marzio. For example, an eleventh-century bath associated with the Church of San Salvatore in Pensilis (found beautifully preserved) operated until the early fourteenth century (Ministero per i Beni Culturali e le Attivita Culturali, 2000: 25). It is uncertain how these monastic communities and private baths obtained water, other than by constructing cisterns and digging wells. The existing monasteries that continued to function must have utilized these same strategies and in a few cases they might have had access to one of the small springs that still flowed in the city. Other Roman springs, like the Juturna, had been buried underground for centuries, but even so, they continued to flow and to saturate the land, creating an unhealthy malarial environment that further weakened an already difficult urban situation. Without aqueduct water the situation became desperate. So, in 1122, Pope Calixtus II (1119–24) sponsored construction of a new watercourse called the Acqua Marrana (also Mariana). It tapped springs that had once supplied the ancient Aqua Julia. The water was redirected to the subterranean channel of another ancient aqueduct, the Claudia. As the Marrana approached Rome it was let into a new ground-level ditch that headed toward Porta Asinaria and the papal residence at the Lateran Palace. But rather than enter the city at this point the ditch ran outside the walls until it reached Porta Metronia, where it passed through the gate into the city through the Valley of Egeria. There it joined an existing brook that still ran through the Circus Maximus and flowed into the Tiber just south of the Cloaca Maxima. The Marrana was specifically intended to operate a series of grain mills owned by the Church that were constructed along the route of the stream and to water orchards, vines, and gardens. There was also an

149-172 WaterUrban Pt1 Ch07 Rinne.indd 167

12/05/2014 11:21


A History of Water

artificial lake for watering animals just outside Porta Metronia (Becchetti, 1999). The renaissance of urban order Rome’s centuries-long physical decline was exacerbated when the popes abandoned the city for Avignon in 1309, where they remained until 1376. Then, beginning in 1378, two competing popes—one nominally in Rome (but usually resident elsewhere) and the other in Avignon—ushered in the Western Schism, which lasted until 1417. Finally, Martin V returned the papacy to Rome in 1420, essentially slowing urban decline. Treatises by Francesco Filarete, Francesco di Giorgio, Leon Battista Alberti, and other architectural theorists offered detailed plans to organize cities around integrated systematic sanitary measures, including sewers and fresh water supplies, and some of these principles were put into practice. New straight streets were cut through the contortions of Rome’s medieval neighborhoods, and Martin V, Nicholas V (1447–55), Sixtus IV (1471–84), and Clement VII (1523–34) introduced legislation intended to regulate the disposal of human waste, debris, and rubbish. At the same time, their attention turned to restoring the Aqua Virgo (by now called the Acqua Vergine), which, although still operating, was in a degraded state. The first clear evidence for the Vergine’s restoration—this time by Pope Nicholas V—dates from 1452. Over the next 100 years, there were other piecemeal restoration efforts, and popes Pius IV and Pius V successfully restored it all the way to the source springs by 1570.4 Soon thereafter, Sixtus V built the Acqua Felice (1585–7), which used some of the Aqua Marcia source springs, and Paul V built the Acqua Paolo (1607– 12), which used some of the Aqua Traiana source springs. Not only were many of the same springs appropriated, but also the new aqueducts essentially followed the ancient courses, colonized some of the ancient ruins, served the same parts of the city as had their ancient counterparts, and flowed through gravity from higher to lower elevations. As in antiquity, the new water supply spurred urban growth and facilitated changing spatial patterns. Now there was water for all types of public fountains, private palaces, and also for new villas and gardens on the hills that might compete in an artistic display with those from antiquity. An enormous amount of water now entered Rome—the Vergine alone delivered 16,560 liters per minute—and so new drains were also needed to carry away increased runoff water (Rinne, 2010: 55). Three areas received special attention: the Vatican and Borgo; the Campo Marzio; and an area that included the Forums of Augustus and Trajan, known at the Pantano (“the swamp”, because of the subterranean water that collected there). Open drains like the Chiavica di San Silvestro were lowered and covered over, and a few ancient drains—like those in the Pantano that linked to the

149-172 WaterUrban Pt1 Ch07 Rinne.indd 168

12/05/2014 11:21

Plumbing Ancient Rome169

Cloaca Maxima—were restored in a process known as risanimento. Once the drainage work was complete, the ground was leveled, sometimes raised, and new streets might be built, like Via Alessandrina and Via Bonella in the Pantano, or old ones repaved or paved for the first time (Lanciani, 1992, IV: 29–30; Rinne, 2010: 198–218). The most important innovation involved building new drains that connected to ancient drains that were, in turn, also restored (Rinne, 2012). Begun during the pontificate of Pius V, this was difficult work, as alluvial deposits had raised the ground level so that ancient drains were buried far below the streets (Narducci, 1889). The chief advantage was that the hybrid drains were built many meters below the new fresh water conduits, which prevented contaminated water from leaching into the drinking supply. Wherever possible, this strategy was employed, especially in the relatively flat Campo Marzio. Like the fresh water that was distributed in conduits, the water in the drains flowed through gravity. Rome’s water legacy Maintaining urban order requires an organized and sophisticated water supply and drainage system controlled by a central authority, whether this is an emperor, pope, dictator, or city council. Thus the system of linking together the new aqueducts, fountains, and drains was followed by succeeding generations of papal and civil administrators—and, of course, this is the strategy that is employed today. The differences are few, but they are critical. First, Rome has three new aqueducts (all built since 1870) that use mechanical pumps to deliver water from distant springs. This means that private residences—even the top storeys at the highest elevations— could receive a private water supply. As a result, public fountains began to disappear. Second, the once-pristine source springs have in some cases (most notably the Virgo/Vergine) been compromised by suburban and rural development. Thus, part of Rome’s water supply is treated in order to achieve European Union standards, while others (again, the Virgo/Vergine) are no longer part of the drinking supply. Third, Rome’s population has increased dramatically, especially in the periphery. Now 8 million individuals claim the water supply and rely on the sanitary sewers, both of which are now administered by the Azienda Comunale Energia ed Ambiente (ACEA).5 Modern treatment plants have been built to process sewage generated in the Greater Metropolitan region of Rome, but as recently as 2000, there were still areas without sewers and “pollution of the river Tiber has remained at extremely serious levels” (Lobina and Iacovitti, 2000: 5). Conservation efforts are in place. For example, less water now flows to ornamental fountains like the Trevi (and the water recirculates); many formerly free-flowing drinking fountains are today manually operated; and watershed environments are more closely monitored. But there is great

149-172 WaterUrban Pt1 Ch07 Rinne.indd 169

12/05/2014 11:21


A History of Water

stress on the system, which still relies in part on its ancient and renaissance roots and restorations. There are no simple solutions, but the secret to establishing and maintaining a secure future for Rome’s water supply may very well reside in an understanding of its ancient framework. Notes 1 The exact number of springs flowing in ancient Rome is unknown. One stillflowing spring carries 10–15 liters per second. Corazza and Lombardi (1995: 196) describe the lens of water for that spring as “abbastanza vasto”. 2 Floods continued to inundate Rome’s lower elevations until the end of the nineteenth century, when large collector drains were built and the Tiber embanked. These drains diverted all the water away from the urbanized areas and sent it far downstream (Narducci, 1889: 53–7). 3 For the aqueducts, see Ashby (1930), Cassio (1756), Evans (2004), Fea (1832), Hodge (1992), de Kleijn (2001), Lanciani (1880), Pace (1983), Parker (1876), Rodgers (2004), and Taylor (2000). 4 For recent discussions of the Virgo/Vergine restoration, see Karmon (2005), Long (2008), and Rinne (2010: chapter 2). 5 See Lobina and Iacovitti (2000) for an analysis of the water and sanitation crisis that is unfolding in Rome.

references Aldrete, Gregory (2007). Floods of the Tiber in Ancient Rome, Baltimore, MD and London: Johns Hopkins University Press. Ammerman, Albert (1990). “On the origins of the Forum Romanum”, American Journal of Archaeology, 194(4), pp. 627–45. Ashby, Thomas (1930). The Aqueducts of Rome, Oxford: Oxford University Press. Bjur, Hans and Barbro Santillo Frizell (2009). Via Tiburtina: Space, Movement and Artefacts in the Urban Landscapes, Stockholm: Svenska Institutet i Rom. Cassio, Alberto (1756). Corso dell’acque antiche portate sopra XIV. aquidotti da lontane contrade nelle XIV. regioni dentro Roma; delle moderne, e di altre in essa nascenti; con l’illustrazione di molte antichità da scrittori, e antiquari non conosciute, ne nominate, Rome: Stamperia Giannini. Cassius Dio (1987). The Roman History, the Reign of Augustus, London: Penguin Books. Cicero (1972). The Letters to His Brother Quintus, Cambridge, MA and London: Harvard University Press and William Heinemann, Ltd. Coarelli, Filippo (1997). Il Campo Marzio dalle Origini alla Fine della Repubblica, Rome: Quasar. Coates-Stephens, R. (1998). “The walls and aqueducts of Rome in the early Middle Ages, A.D. 500–1000”, Journal of Roman Studies, 88, pp. 166–78. Corazza, A. and L. Lombardi (1995). “Idrogeologia dell’area del centro storico di Roma”, in Memorie descrittive della Carta Geologica d’Italia, pp. 179–211, Rome: Istituto Poligrafico e Zecca dello Stato.

149-172 WaterUrban Pt1 Ch07 Rinne.indd 170

12/05/2014 11:21

Plumbing Ancient Rome171

Dennis, George (1848). Cities and Cemeteries of Etruria, Vol. 1, London: John Murray. Dionysius of Halicarnassus (1961). The Roman Antiquities of Dionysius of Halicarnassus, Boston, MA and London: Harvard University Press and William Heinemann, Ltd. Duchense, Louis M. O. (1886). Le Liber pontificalis, Paris: E. Thorin. Evans, Harry (2002). Water Distribution in Ancient Rome: The Evidence of Frontinus, Ann Arbor, MI: University of Michigan Press. Facenna, C., R. Funiciello, and F. Marra (1995). “Inquadramento Geologico Strutturale dell’Area Romana”, in Memorie descrittive della Carta Geologica d’Italia, pp. 31–47, Rome: Istituto Poligrafico e Zecca dello Stato. Fea, Carlo (1832). Storia delle acque antiche sorgenti in Roma, perdute, e modo di ristabilire. 2: Dei condotti antico-moderni delle acque, Vergine, Felice e Paola, e loro autori, Rome: Stamperia della R.C.A. Frosini, Pietro (1977). Il Tevere. Le Inondazioni di Roma e I Provvedimenti presi dal Governo Italiano per Evitarle, Rome: Accademia Nazionale dei Lincei. Funiciello, Renato, L. Lombardi, F. Marra, and M. Parotto (1995). “Seismic damage and geological heterogeneity in Rome’s Colosseum area: are they related?”, Annali di Geofisica, 38(5–6), pp. 927–37. Gjerstad, Einar (1953). Early Rome, Lund: C. W. K. Gleerup. Gregori, G. P., R. Santoleri, M. P. Pavesi, M. Colacino, E. Fiorentino, and D. de Francheschi (1988). “The analysis of point-like historical data series”, in W. Schroder (ed.), Past, Present and Future Trends in Geophysical Research, pp. 146–211, Bremen-Roennebeck: Interdivisional Commission on History of IAGA. Hehn, Victor (1885). The Wanderings of Plants and Animals from their First Homes, London: Swan Sonnenschein and Co. Hodge, Trevor (1992). Roman Aqueducts and Water Supply, London: Duckworth. Holland, Louise (1961). Janus and the Bridge, Rome: American Academy in Rome. Hopkins, John N. N. (2007). “The Cloaca Maxima and the monumental manipulation of water in archaic Rome”, The Waters of Rome, 4, IATH, University of Virginia, www3.iath.virginia.edu/waters/Journal4Hopkins.pdf. Horace (1978). Satires, Epistles, and Ars Poetica, with an English translation by H. Rushton Fairclough, Cambridge, Mass: Harvard University Press. Karmon, David (2005). “Restoring the ancient water supply system in Renaissance Rome: the popes, the civic administration, and the Acqua Vergine”, The Waters of Rome, 3, IATH, University of Virginia, www3.iath.virginia.edu/waters/ Journal3KarmonNew.pdf. Kleijn, Gerda de (2001). The Water Supply of Ancient Rome: City Area, Water and Population, Amsterdam: J. C. Gieben. Lanciani, Rodolfo (1880). Topografia di Roma antica: I commentarii di Frontino intorno le acque e gli acquedotti, Rome: Salviucci. ——— (1897). The Ruins and Excavations of Ancient Rome, Boston, MA and New York: Houghton, Mifflin & Company. ——— (1992). Storia degli scavi di Roma e notizie intorno le collezioni romane di antichità, Rome: Edizione Quasar. Livy (1979). The Early History of Rome, Book 4, Hammondsworth, Middlesex, England: Penguin Books. Lobina, Emanuele and Daniele Iacovitti (2000). “D36: Watertime Case Study-Rome”. Available at www.watertime.net/wt_cs_cit_ncr.asp.

149-172 WaterUrban Pt1 Ch07 Rinne.indd 171

12/05/2014 11:21


A History of Water

Long, Pamela O. (2008). “Hydraulic engineering and the study of antiquity: Rome, 1557–1570”, Renaissance Quarterly, 61(4), pp. 1098–1138. Manacorda, Daniele (2001). Crypta Balbi: archeolgia e storia di un paesaggio urbano, Milan: Electa. Ministero per i Beni Culturali e le Attivita Culturali (2000). Museo Nazionale Romano Crypta Balbi, Rome: Electa. Narducci, Pietro (1889). Sulla fognatura della città di Roma, Rome: Forzani. Ovid, (1989). Fasti; with an English translation by Sir James George Frazer, Cambridge, Mass: Harvard University Press. Pace, Pietrantonio (1983). Gli acquedotti di Roma e il “De Aquaeductu” di Frontino. Contesto critico, versione, e commento, Rome: San Eligio. Parker, John Henry (1876). The Aqueducts of Ancient Rome, Oxford: Oxford University Press. Pliny the Elder (1861). The Natural History, London: Henry G. Bohn. Ponzi, Giuseppe (1867). Sulla storia fisica del Bacino di Roma, Rome: Tipografia delle Belle Arti. Procopius of Cesarea (1919). De Belli Gotti, Boston, MA and London: Harvard University Press and William Heinemann, Ltd. Purcell, Nicolas (1996). “Rome and the management of water: environment, culture and power”, in Graham Shipley and John Salmon (eds), Human Landscapes in Classical Antiquity; Environment and Culture, pp. 180–212, London: Routledge. Re, Camillo (1880). Statuti della città di Roma del secolo XIV, Rome: Tipografia della Pace. Richardson, Lawrence (1992). A New Topographical Dictionary of Ancient Rome, Baltimore, MD and London: Johns Hopkins University Press. Rinne, Katherine (2010). The Waters of Rome: Aqueducts, Fountains, and the Birth of the Baroque City, New Haven, CT and London: Yale University Press. ——— (2012). “Urban ablutions: cleansing Counter-Reformation Rome”, in Mark Wrigley (ed.), Rome, Pollution and Propriety; Dirt, Disease and Hygiene in the Eternal City from Antiquity to Modernity, pp. 182–201, Cambridge: Cambridge University Press and The British School at Rome. Rodgers, Robert H. (2004). Frontinus: De Aquaeductu urbis Romae, Cambridge: Cambridge University Press. Smith, Strother (1877). The Tiber and Its Tributaries, Their Natural History and Classical Associations, London: Longmans Green. Taylor, Rabun (2000). Public Needs and Private Pleasures. Water Distribution, the Tiber River and the Urban Development of Rome, Rome: “L’Erma” di Bretschneider. Verdi, Orietta (1997). Maestri di edifici e di strade a Roma nel secolo XV: fonti e problemi, Rome: Roma nel Rinascimento. Yegül, Fikret (1992). Baths and Bathing in Classical Antiquity, Cambridge, MA and London: MIT Press.

149-172 WaterUrban Pt1 Ch07 Rinne.indd 172

12/05/2014 11:21


Aksum: Water and Urbanization in Northern Ethiopia

Federica Sulas “Complain when you see the king; sow when you see the water” (Tegreña proverb; Conti Rossini, 1942: 57).

Introduction1 The region between modern northern Ethiopia and central Eritrea offers an example of the intricate links between water and urbanization. This history has received little scholarly attention, even though the region has long been recognized as a center of plant domestication (Harlan, 1971) and host of some of the earliest state societies in sub-Saharan Africa (Fattovich, 2008, 2010; Phillipson, 2012). The highlands of Tigray between northern Ethiopia and Eritrea (Figure 8.1) saw the emergence of complex societies from the second millennium bc (Phillipson, 2009, 2012; Fattovich, 2010), leading to the rise of the kingdom of Aksum2 (50 bc– ad 700/800). The development of urbanism at Aksum was strongly influenced by its strategic geographical and topographic position, and the availability of water and land resources. The new kingdom thrived for a thousand years, engaging in long-distance trade and commerce, developing a written literature and coinage, and acting as a gateway for the introduction and spread of Christianity into Africa (Phillipson, 2012). Aksum’s position, together with its water and land resources, sustained local communities and, subsequently, their engagement in, and intermittent control over, regional and interregional trade. A somewhat independent subsistence system was already in place by the first millennium bc and survived the collapse of temporal power towards the late first millennium ad. These developments were followed by the southward expansion of the state in the late first–early second millennium ad. A new polity emerged further south in central Ethiopia (Zägwe Dynasty, c.ad 1137–1270), and was soon overtaken by the Solomonic Dynasty (ad 1270–1974) and the establishment of the Ethiopian Empire. By eliciting environmental, archaeological, and historical sources, this chapter discusses how the early societies of northern Ethiopia and Eritrea

173-197 WaterUrban Pt2 Ch08 Sulas.indd 175

12/05/2014 11:22


A History of Water

Figure 8.1.  Map of northern Ethiopia and Eritrea.

interacted with a highly diversified environment, and the types of strategies they have developed to manage water and land resources. The emergence of the Aksumite kingdom By the early first millennium bc, several large-scale settlements had arisen across the region spanning the Tigrean highlands to the coastal plains of the Red Sea. A main polity, known as D≤MT (c.900/800–400 bc), was centered at Yeha (Figure 8.1) and associated with the Pre-Aksumite culture. The Pre-Aksumite culture was characterized by a strong influence from southern Arabia and was only part of a mosaic of different cultures present in Tigray and Eritrea during the first millennium bc (Schmidt et al., 2008; Phillipson, 2009, 2012; Fattovich, 2010). By the mid-first millennium bc, permanent settlement is amply documented in the Aksum area with a new polity centered on Betä Giyorgis hill (Proto-Aksumite, 400–50 bc; Figure 8.2). The main residential area (‘Ona Nägäst) and an elite cemetery (‘Ona Enda Abboy Zägwe) were located on the hilltop, and small settlements and rural homesteads were scattered across the greater Aksum area (Fattovich, 2008; Sernicola, 2008). Subsequently, the residential area increased and, from the mid-first millennium ad, the focus of political control shifted from Betä Giyorgis hilltop to its southern pediment and the plain, marking the rise of the kingdom of Aksum (50 bc–ad 700/800). Aksum soon consolidated its influence on a regional and interregional scale, and extended its political and economic control towards the Red Sea, the Eastern Desert, and possibly the upper Nile Valley. A centralized state emerged by the

173-197 WaterUrban Pt2 Ch08 Sulas.indd 176

12/05/2014 11:22

Aksum: Water and Urbanization in Northern Ethiopia177

second century ad and the Aksumite court officially adopted Christianity by the early third century ad. The spread of the new religion among the sub-strata of the society is likely to have taken place within the following two centuries (Conti Rossini, 1928: 141–95; Phillipson, 2012: 91–106). By the late third century ad, the kingdom began minting coins in gold, silver, and copper, which rapidly spread towards commercial partner regions as far as India (Munro-Hay, 1991; Phillipson, 2012: 181–93). At this time, Aksumite influence can be detected as far afield as the Red Sea coast, the western Sudanese lowlands, the regions to the west of the Tekeze River (Ethiopia), and part of southwestern Arabia (Phillipson, 2012: 200–3). The results of intensive archaeological surveying and excavations show changing patterns of settlement throughout the development of Aksum (Sernicola, 2008). Even though the number of people present in a given time is not known, it is commonly accepted that the ancient capital hosted a population of several thousands at the peak of its expansion (Michels, 2005; Fattovich, 2008). Aksumite settlements included towns, villages, isolated hamlets and, following the introduction of Christianity, churches and monasteries (see Sernicola, 2008). By the mid-first millennium ad, the

Figure 8.2.  Map of Aksum showing the location of water cisterns and springs (circles) mentioned in the text.

173-197 WaterUrban Pt2 Ch08 Sulas.indd 177

12/05/2014 11:22


A History of Water

kingdom had reached its maximum expansion in terms of both terri­ torial control and urban development (Fattovich, 2008). In the capital itself, the urban quarters were located in the flood plain: a large residential area, including monumental buildings (commonly referred to as palaces) and (royal) funerary complexes, stretched along Betä Giyorgis pediment to the north and northwest of the modern town. Lower-class residential structures were also present in the plain and on the hilltop of Betä Giyorgis, and at small rural sites in the surrounding countryside. In the sixth and seventh centuries ad, the settlement at Aksum occupied a much smaller area, concentrated around the Cathedral of Maryam Seyon. The archaeological evidence indicates that a social and economic crisis occurred around this time. The triggers of change are still uncertain, but great emphasis has been placed upon the combined impact of demographic pressure, overwhelming exploitation of the land, and climate deteri­ oration, which eventually would have led to the decline of Aksum (Butzer, 1981, 2012). From the late sixth century, Aksum’s commerce in the Red Sea and Indian Ocean was gradually eroded by external powers: the Persians’ expansion in southern Arabia and, shortly afterward, the Arab conquests and control of the Red Sea trade route. At home, tribal movements into the Eritrean highlands threatened Aksumite activities away from the capital and toward the coast (see Fattovich, 2010). By the late seventh/early eighth century ad, Aksum was no longer the capital city. While the decline of the elite occupation is well expressed in the archaeological record, rural compounds persisted in relatively significant numbers in the countryside (Sernicola, 2008; Sernicola and Sulas, 2012). Very little is known about the following few centuries, but archaeological survey data indicate that between the tenth and fifteenth centuries, settlement was concentrated mainly along the southern pediment and on the hilltop of Betä Giyorgis. Moreover, buried soil records from the hills and river valleys north of the town reflect the persistence of settlement and arable land use in relatively stable landscape conditions until about ad 1600 (French et al., 2009; Sulas, 2010). Water in the modern Ethiopian highlands Discussions on the development of ancient societies, whether farming communities or complex states, in northern Ethiopia and Eritrea often concentrate on rainfall, and largely exclude other forms of water and the impact of these on society. In a region where water waste is a critical factor due to prolonged dry periods and droughts, the management of soil moisture has always been vital for productive farming and domestic activities. Yet catchment hydrology finds no consideration in literature, even though its dynamics are fundamental in understanding past human decisions (see Pikirayi, in press). Moreover, an important caveat against addressing water history in the region concerns the types of records and

173-197 WaterUrban Pt2 Ch08 Sulas.indd 178

12/05/2014 11:22

Aksum: Water and Urbanization in Northern Ethiopia179

processes associated with water. While built infrastructures such as dams can survive for millennia and provide information about the types and uses of water resources, research on supply provided by rainfall and groundwater is complicated, as these forms of water operate at multiple scales through time, and are influenced by several factors. In northern Ethiopia and Eritrea, built evidence of water management is limited, but archaeology, environmental sciences, and historical sources provide relevant data. The landscape and its farming communities offer further insights into aspects of land and water management that have characterized this region for millennia. Ethiopia is a land of rivers and, as the highlands slope northwestwards, nearly all the large rivers flow in that direction to the Nile. Aksum lies on a water-divide between two main drainage systems (Figure 8.1): the Tekeze/ Atbara River (a major Nile tributary) and the Märäb/Gaš River (which flows into the Sudanese lowlands). Linked to these two main rivers is a network of smaller tributaries and streams, most of which afford high and often erratic flows during the rainy season, but rapidly become dry during the following months. The undulating topography of the Aksum area is characterized by flatlands and hills with gently sloping to steep sides, covered by scattered bushes and shrubs, and sparse patches of trees. The town lies on a large plain at an average elevation of about 2,200 m above sea level and is surrounded by a number of hills. The local geology is dominated by trachytic and phonolitic domes, stratified flood basalts with reddish sandstone, laminated siltstones, and conglomerates. Quaternary deposits of alluvium, colluvium, and tufa are found along river valleys, foot-slopes, and in depressions (Assefa and Russo, 1997). At the contact between hard and soft rocks, flat-topped hills, known as ambat (sing. amba), provide good land for settlement and cultivation over the tops, and gently sloping to steep sides (Coltorti et al., 2007). Foot-slopes and piedmonts are made of hillwash, and open onto fertile flood plains. The regional climate is complex and varies significantly within short distances, shifting from the extreme low temperatures of the mountain tops to the tropical desert of the lowlands. Most of Tigray falls within the main climatic belt known locally as wäyna däga (“land of vineyards”), a temperate zone between 1,500 m and 2,500 m above sea level with annual temperature means of 15–20ºC, and moderate rainfall of 200–700 mm per year. However, rainfall and temperature vary significantly across the region, and there have been few attempts to regionalize rain patterns (Nyssen et al., 2005). The highly seasonal rainfall pattern is determined by the monsoon, which is largely controlled by the movement of the Inter-Tropical Convergence Zone. Most of the rain falls between June and August, and less reliable rains may occur between March and May. In the decade 1995–2005, Aksum received less than 650 mm rainfall per year, most of which was concentrated in July and August. Its temperature

173-197 WaterUrban Pt2 Ch08 Sulas.indd 179

12/05/2014 11:22


A History of Water

reached the annual mean of 19.7ºC, and evapotranspiration shifted between 613 mm and 569 mm per year (Tadesse et al., 2010).3 If rainfall has always been a vital, yet uncontrollable resource, the underground reserves significant amounts of water: the volcanic geology (trachyte, flood basalt, sandstone, and alluvial deposits) provides optimal rock bodies for aquifers, and parent material for medium- to light-textured soils. Shallow aquifers are confined to sediments of low productivity, but the deep ones found in fractured basaltic rocks afford good permeability (Alemayehu et al., 2011).4 Groundwaters are recharged by direct infiltration of rainfall and through lateral flow. Springs are thus relatively common in Aksum’s landscape (Figure 8.2), and the ones associated with flood basalts, sedimentary rocks, and lava flows afford high yields. The flood plains, where most farming takes place, are recharged every year by mountain streams—many of which have shallow water tables filled with coarse-grained alluvium with good hydraulic conductivity. As such, flood plains maintain shallow water tables during the dry season. Rainfall, surface water, and groundwater underpin the rural farming population, to which land is not only the basic means of livelihood and a resource for income, but also a source of shared history and culture. Traditional farming combines the use of maräša (ard-plough), cattle, and annual crops (Simoons, 1960). In the highlands, altitude and temperature offer favorable conditions for growing a variety of cereals (t·ef, barley, wheat, sorghum, finger millet), legumes (chickpeas, lentils, beans, broad beans), and oilseed plants (flax, noug). Soil management, particularly the preservation of moisture, has always been a fundamental task for the highlanders, who employ crop rotation and short-period fallows to allow land recovery.5 The crop repertoire, together with the use of plough, has been historically stable (McCann, 1995), and is still preferred in the countryside against state-sponsored mechanized and monoculture agriculture. Landscape history and urban development Environmental proxies indicate that present climatic conditions began sometime in the second millennium bc (Bard et al., 2000), and the record from Tigray is broadly consistent with the sequencing of the northeastern African palaeoclimate during the late Holocene, which is characterized by increasing aridity interspersed by wet pulses (Dramis et al., 2003; Marshall et al., 2009). In the Tigrean highlands, wet conditions seem to have prevailed during three main periods: c.2050–1550 bc, c.550 bc–ad 450, and around ad 1000 (Machado et al., 1998; Gebru et al., 2009; Terwilliger et al., 2011). A rapid increase in aridity occurred after the mid-first millennium ad (c.ad 500–50; see Machado et al., 1998; Terwilliger et al., 2011). However, alternating intervals of minor high and minor low Nile flood discharge between the seventh and the early tenth centuries (Hassan, 2007) suggest

173-197 WaterUrban Pt2 Ch08 Sulas.indd 180

12/05/2014 11:22

Aksum: Water and Urbanization in Northern Ethiopia181

that rainfall on the Ethiopian highlands provided a relatively consistent input to the Nile. The following centuries until the mid-second millennium ad saw the recurrence of abnormal floods, either excessively high or low, which triggered famines both in Egypt (Hassan, 2007) and in northern Ethiopia (Pankhurst, 1985). The rise and decline of Aksum have been linked to environmental issues for a long time. In the late 1930s, Monneret de Villard (1938) first identified two main phases of sediment depositions next to the Cathedral of Maryam Seyon (Figure 8.3): the earliest phase would have preceded the rise of Aksum and a second phase would have taken place after the decline of the kingdom. These first environmental data found later support in the geoarchaeological investigations by Butzer (1981), who recorded four aggradation phases and linked them to climatic changes and cultural developments. In particular, the first aggradation phase was associated with a period of increased precipitation and the growth of Aksum as a political center; a second aggradation phase was linked to the erosion of degraded agricultural lands upslope and heavy rains, following a period of settlement and demographic increase. The last two phases of aggradation occurred several centuries after the decline of the Aksumite kingdom. Combined, these data provided the basis for a model of urban

Figure 8.3.  Figurative map of Aksum, based on Littmann et al. (1913) and Monneret de Villard (1938). Squares: Aksumite structures (buildings, churches and funerary complexes), inscriptions (I), and throne-bases (T); dots: stelae; circles: cisterns and springs; dark-grey outlines: compounds and homesteads as in the 1930s.

173-197 WaterUrban Pt2 Ch08 Sulas.indd 181

12/05/2014 11:22


A History of Water

growth linked to climate change and environmental degradation. During the last two decades, intense research in and around Aksum has provided new local data that suggest a more complex scenario (see Sernicola and Sulas, 2012). On the one hand, studies of plant remains from archaeological and landscape contexts (seeds, charred wood, pollen, and phytoliths; see Boardman, 1999; Bard et al., 2000; D’Andrea, 2008; French et al., 2009; Sulas, 2010) indicate that a woody savannah vegetation cover was in place at Aksum before, during, and after the kingdom, whereas there is no evidence for the presence of woodland. On the other hand, new sedimentary records from the hilltop of Betä Giyorgis, hillsides, and river valleys north of the town reflect prolonged landscape stability associated with settlement and land uses from the mid-fourth millennium bc until about ad 1600 (French et al., 2009; Sulas et al., 2009; Sulas, 2010). On hilltops and uplands, buried thick soil horizons were associated with settlement and farming, while buried soil records from hillside deposits may have been linked to pastures. The botanical remains (phytoliths and charred wood) from the buried soils reflect a substantially stable woody savannah vegetation cover with tree patches along watercourses and settlements, of which palm trees were a significant component. In addition, there is some indication of changes in soil moisture conditions that seem to have been concomitant to both phases of landscape stability and localized soil erosion, suggesting that changing rainfall was not the main trigger of soil moisture change (Sulas, 2010).6 Water, farmers, and kings The subsistence of the ancient societies in the Tigrean highlands was based on farming and husbandry. During the early first millennium bc, non-elite farming communities cultivated grains such as barley and emmer (Boardman, 1999; D’Andrea, 2008; D’Andrea et al., 2008, 2011). Near Eastern crops7 were most frequently grown by these early communities, who largely depended upon a vegetarian diet (D’Andrea et al., 2011). As indicated by Boardman (1999: 144), the archaeobotanical evidence implies “the existence of dry-land agriculture based on ox-plough and mixed annual crops by the mid-first millennium bc and possibly earlier”. A more varied suite of crops were grown at Aksum from the mid-first millennium ad. To support such a farming-based subsistence and urban development, former research has emphasized the use of stream-fed ditch irrigation as it was observed at Yeha in the mid-1970s (Michels, 2005). However, this is only sparsely practiced in northern Ethiopia (Simoons, 1960; McCann, 1995), and, today, rainwater drainage is mainly controlled by ploughing along field boundaries and digging shallow channels (depths of 15–20 cm), a practice occasionally observed in historical periods (Salt, 1814: 19; McCann, 1995: 59). Since the onset of the present climate

173-197 WaterUrban Pt2 Ch08 Sulas.indd 182

12/05/2014 11:22

Aksum: Water and Urbanization in Northern Ethiopia183

(second millennium bc), rainfall, cool temperatures, and low evaporation have supplied enough water and moisture for ensuring yields in the highlands.8 However, some form of irrigation may have been practiced, though it is unlikely that this was on a wide scale. In Akkälä Guzay (Figure 8.1), for example, horizontal fields and remnants of old dams and canals may be linked to ancient irrigation (Brunner, 2005). Similarly, the remains of massive walls together with a terrace on Betä Giyorgis hilltop have been associated with water management (Fattovich, 2008). The buried soil records associated with the Pre-Aksumite and Aksumite occupations from Betä Giyorgis and other upland and river valley sites north of Aksum yield no indication of irrigation, but there is evidence of moisture changes, most likely in relation to changing water-table levels (Sulas, 2010). More substantial evidence for ancient irrigation comes from the lowlands and coastal plains of Eritrea. Here, much drier conditions and higher evaporation rates would have impinged upon crop growing that was solely dependent on rainfall for its water supply. The vast plain of Adulis, for example, borders a large river and may well have provided a “man-made oasis where flood irrigation accumulated fertile silt” (Brunner, 2005: 36). Further inland, the area of Kärän preserves silty terraces and relict canals that Brunner (2005: 33) links to early irrigation, possibly dating to the second millennium bc. Indeed, there may have been no need for irrigation before the onset of present climatic conditions. Most of the highland farming may have thrived without the support of irrigation, but significant water supply was needed to sustain a multifaceted population of farmers, craftsmen, traders, priests, and rulers. Water-storing structures have been recorded in the Aksum area and on the Qohayto plateau (Eritrea), but none has been fully investigated. Many of these structures are located in between or in close proximity to Aksumite settlement sites. On the hilltop of Betä Giyorgis, for example, the two modern cisterns of May ≤Agam (Figure 8.2) cut into alluvial (probably crystalline) rock that afford a high water yield, and a series of Aksumite sites are located in the immediate vicinity.9 By far the largest and most important cistern on the hilltop is ≤Ela Nägäst,10 located next to the early residential area of ≤Ona Nägäst (Ricci, 1974, 1990: 139; Fattovich, 2008). The cistern cuts across stratified sedimentary rocks and is elliptical in shape (c.25 m by 20 m long, originally 7 m deep). It is recharged by groundwater, direct precipitation, and overland flow during periods of rain (Fattovich et al., 2000: 42). The mixed potsherds recovered from the fill date to all phases of the Aksumite periods. At Aksum itself, several springs and water cisterns are present, but none firmly predate the fifteenth century. The most celebrated water source is May Šum (Figure 8.2); the structure was originally about 65 m in diameter and 5 m deep, and has been enlarged at various stages (Littmann et al., 1913, Vol. 2: 70–3; Monneret de Villard, 1938: 8–11; Phillipson, 1997: 156–60).11 In addition to rainfall, the cistern receives substantial runoff

173-197 WaterUrban Pt2 Ch08 Sulas.indd 183

12/05/2014 11:22


A History of Water

from the surrounding hillsides and, presumably, from natural springs along the slope and hilltop of May Qoho. Local traditions place the building of the structure in the fifteenth century (Phillipson, 1997: 156–60), though a much earlier date is very likely (see also Michels, 2005: 142). In Eritrea, a series of ancient cisterns have been recorded within major Aksumite settlement areas on the Qohayto plateau (c.2,600 m above sea level, Figure 8.1), and provisionally ascribed to the Aksumite Period (Wenig, 1997; Wenig and Curtis, 2008). Among them, the so-called “Safira Dam” is located at the beginning of a shallow valley that falls down into a gorge. The basin is square (c.67 m long and approx. 3 m high at its center) and has natural borders on three sides, whereas the downstream side is made up of a built wall constructed from regular courses of fitted blocks with rough stone steps. Early descriptions emphasize the resemblance with the famous dam of Marib in highland Yemen (e.g. Bent, 1893: 219–20), but others recognize it as a cistern for rain-capturing, similar to the reservoir of May Šum at Aksum (e.g. Littmann et al., 1913, Vol. 2: 148–52; Conti Rossini, 1928, Vol. 2: 243). The structure awaits in-depth study, but a recent assessment by Brunner (2005: 34) interprets it as a rain-harvesting cistern,12 arguing that “the knowledge for the arrangement of the cistern came from Yemen, but the architecture of the wall is definitely Aksumite”. The location on top of a plateau, where rainwater is abundant, and the lack of waterproof revetment, favored water draining into the valley floor downstream. While archaeological and environmental records illuminate some of the ways water was used, local traditions and micro-toponyms emphasize a connection between Aksum and water. The very origin of the name “Aksum” may illustrate this bond: the syllable “ak-” may derive from the Cushitic root for “water”, and šum is the Semitic term for “chief ” (Munro-Hay, 1991: 96; Finneran, 2007a: 152). Other hypotheses favor a western Agaw etymology (akuesem), meaning “water reservoir”.13 According to the Liber Axumae,14 the “capital” was moved three times, and a swamp covered most of the present site (Monneret de Villard, 1938; Munro-Hay, 1991). The latter was chosen by the kings Abreha and Asbeha, who miraculously drained the area of a great lake.15 Monneret de Villard (1938: 19) argues that the plain was very low and covered by pools in ancient times; this hypothesis is supported by his investigations of alluvial deposits next to the Cathedral of Maryam Seyon. Indeed, part of the marshland persisted at least until recently, as indicated by the “marshy ground” marked in the first topographic map of Aksum, drawn by Salt in 1805 (Mountnorris, 1809, Vol. 3: 100). It is noteworthy that most of the major Aksumite monuments and buildings are located on a strip of land running from the pediment of May Qoho with no extension westwards (Figure 8.3). Aerial photo interpretation (Fattovich et al., 2000: 42–3) suggests the presence of a cross-pattern dividing the town into four sectors (main axes N–S and E–W), which are consistent with the spatial distribution of the archaeology: the elite residential palaces are located

173-197 WaterUrban Pt2 Ch08 Sulas.indd 184

12/05/2014 11:22

Aksum: Water and Urbanization in Northern Ethiopia185

mainly in the southwestern sector, and the ceremonial area (churches and stelae fields) occupies the northwest sector. Another local tradition from the Liber Axumae mentions the presence of 72 water sources at Aksum (Conti Rossini, 1909: 5; cf. Beckingham and Huntingford, 1961: 522–3). Further on, the text refers again to Aksum: the Mother of God spoke to Abba Heryakos… and to Yared the priest of Aksum, and brought them together in one place which is called May Kerwah… and they… came to a district of Aksum, the name of which is May Kerwah… (Beckingham and Huntingford, 1961: 523–4; see also Conti Rossini, 1909: 5)

May Kerwah is also mentioned in the hagiography of Yared, deacon and musician from near Aksum, who lived sometime in the sixth century and became saint of the Ethiopian Orthodox Church. This tradition places May Kerwah near the “house of Gäbrä Mäsqäl” (Conti Rossini, 1904: 27; see also Finneran, 2007b). The name is used today with reference to a well dedicated to Saint Yared and located to the north of the town (Figure 8.2), on the way to the tombs of the Aksumite kings Kaleb and his son Gäbrä Mäsqäl (c.sixth century ad). The very mention of May Kerwah by two distinctive sources is significant if we consider that the main reservoir of May Šum is not attested as such until the late nineteenth century ad (e.g. Vigoni, 1881: 138), though its presence has been recorded since the early sixteenth century. For example, a Portuguese account from the 1520s (Beckingham and Huntingford, 1961: 38) describes it as a “very handsome tank of (or lake of) spring water”, next to where the market was held; perhaps an indication that the center of urban life was toward the north of the town (cf. Monneret de Villard, 1938: 14).16 Travelers’ accounts and, later, colonial reports indicate that May Šum was Aksum’s main supply of water from the early seventeenth century until at least the early twentieth century (e.g. Barradas, 1634: 12, 119; Bruce, 1790, Vol. 3: 132; Mountnorris, 1809, Vol. 3: 100; Parkyns, 1853: 208; Bent, 1893: 125; Monneret de Villard, 1938). The written sources also provide ample references to the rural landscape with cropping fields and rich orchards. Summer rains, streams, springs, and the dew provided abundant “good” water for plough-farming a variety of grains and pulses (e.g. Barradas, 1634: 12–15, 119–20; Ludolf, 1684: 26–7; Bruce, 1813, Vol. 5: 233–4; Parkyns, 1853: 206–10); the majority of which was already cultivated in Aksumite times. Irrigation ditches were only observed near Yeha in the early sixteenth century (Beckingham and Huntingford, 1961: 141) as there was no particular need for it: moisture continually distilling from the mountains: for the solid stones not admitting the rain, the water falls off them, and spreading under the fertile turf, wonderfully recreates and enlivens the growing plants. (Ludolf, 1684, Vol. 9: 48)

173-197 WaterUrban Pt2 Ch08 Sulas.indd 185

12/05/2014 11:22


A History of Water

Water and society in the rise and decline of the Aksumite kingdom African countries are among the most afflicted by water scarcity17 and, yet, the continent has a substantial amount of groundwater, which is estimated to provide fresh water more than a hundred times the annual renewable fresh water resources (MacDonald et al., 2012). Scaling the picture down, 75 percent of about 4 million people of Tigray are food-insecure and plagued by recurring droughts. Is this a result of climate change? We know that present climatic conditions were established sometime in the second millennium bc. In this region of northeastern Africa, the water-landscape– people relationship shaped the development of urbanism, a process that lasted for a thousand years. Following the decline of the Aksumite kingdom in the late first millennium ad, urbanism disappeared from the Ethiopian highlands, and several centuries passed before new urban centers emerged. Rather than discussing the main agents of such changes, the following section outlines the interactions of water and societies, and the processes that led to urban development and, subsequently, ruralization at Aksum. The available information, drawn from environmental, archaeological, and historical records, indicates the persistence of substantially stable climate and soil conditions, a local knowledge of water resources and specific management choices, and a tailored and flexible subsistence system. What, then, was required for a successful water management at Aksum? Under monsoonal climate and at high altitude, wet and dry seasons have a significant impact on land productivity and water quality. A main challenge was, and still is, associated with predicting the end of the dry season and the beginning of the rainy one. This is a critical factor and, today, farmers sow their main staples immediately before the rains begin. At a household level, there was the need to have access to clean water, particularly during long dry periods when water from rivers and streams may have been scarce. Sizable cisterns are likely to have been the main device for harvesting and storing rainwater. To ensure adequate and safe supply, water harvesting and storing technologies had to meet the following minimum requirements: to impound enough water to satisfy the demand; to store the maximum quantity of water possible; to limit any leakage; and to be located where stored water could be put to the best use. From a topographic point of view, broad valleys where streams run slowly into steep-sided cross-sections offer suitable conditions for reservoirs. Where rainfall patterns and runoff are the main concern, catchments with steep, rocky slopes are favored, as the rain falling runs downhill into a well-defined stream-bed, in which absorption and evaporation are at a minimum. Geological settings and building technology influence the possibility and amount of leakage and the infiltration potential of reservoirs. The cistern of ≤Ela Nägäst on Betä Giyorgis (Figure 8.2), for example, was

173-197 WaterUrban Pt2 Ch08 Sulas.indd 186

12/05/2014 11:22

Aksum: Water and Urbanization in Northern Ethiopia187

cut into sedimentary rocks and recharged by groundwater, direct precipitation, and overland flow during periods of rain. ≤Ela Nägäst is located on higher ground on the southwestern sector of the hilltop, facing gently sloping lands to the southeast, and bordering a crest to the southwest. The sloping lands opposite to the cistern form a shallow depression where rainwater collects and channels into a small seasonal stream (May Lahlah), which drains along the southwestern slope until it reaches the old quarter of Malake Aksum (Figure 8.3). In the southwestern sector of Betä Giyorgis hilltop, a seasonal stream (May Lahlah?) and possibly springs (May ≤Agam?) would have supplied water during the rainy season, while the cistern of ≤Ela Nägäst ensured supply for the following dry months. Similarly, May Šum is located at the foot of a steep hill and to the side of a seasonal stream, called the May Heg ˇˇga18 (Figures 8.2 and 8.3)—a position that guaranteed direct capture of rainfall from the hillsides and mutual groundwater recharging via the seasonal stream. Stored water needed to be kept clean from toxins and waterborne diseases. In this respect, classic Maya farmers succeeded in keeping clean the water stored in large reservoirs by “transforming artificial reservoirs into wetland biospheres” (Scarborough, 2009; Lucero et al., 2011). The understanding of wetland biosphere principles would have allowed classic Maya farmers to control the spread of water-loving plants and other organisms. While there is no physical evidence to support such claims, the cisterns associated with Aksumite settlements, such as ≤Ela Nägäst on Betä Giyorgis and Safra in Qohayto, would have required control of water quality throughout the year. Indeed, these two examples still provide the main water supply for the local farming communities. Compared to anywhere else in Africa, the early and medieval history of Ethiopia is exceptionally well documented. However, we can say very little for the earlier periods, including the Aksumite ones, as regards provincial organization, inheritance laws, or settlement of an army. Archaeological evidence for the early period of urban development at Aksum provides some indications of a social and economic structure. Fattovich (2008), for example, argues for the presence of three main groups within the Aksumite society: upper elite (kings and nobles), lower elite (lower-status nobles and/or wealthy farmers), and farmers. There is also evidence for an increasing social hierarchy that, from a division between elite and commoners in the last few centuries bc, culminated in further differentiation and the emergence of an intermediate group (lower elite or rich farmers) in the mid-first millennium ad. Fattovich (2008) also draws evidence for land tenure from landmarks and inscriptions. Isolated, rough monoliths on Betä Giyorgis hilltop, dated to the last few centuries bc, may have been erected to mark control of the territory. A series of short inscriptions with family names or lineages (c.third century ad) from Betä Giyorgis and the early to mid-fourth-century royal inscriptions along roads at the entry of Aksum may also be considered landmarks. Whether these features

173-197 WaterUrban Pt2 Ch08 Sulas.indd 187

12/05/2014 11:22


A History of Water

marked territorial boundaries or not, they can only hint at some form of land holding, but they bear no indication of how or by whom land and water resources were managed. It is relevant to consider the situation in place in the following centuries, for which a rich manuscript literature elucidates the social structure and land tenure of historical Ethiopia from the thirteenth century ad (Crummey, 1980, 2000). Strongly hierarchical, the society of historical Ethiopia was not rigidly stratified, and the nobility and the peasantry both believed themselves linked by common ancestry. Social differences were related to two main property rights: gwelt, the right to surpluses through tribute; and rest, the right to direct access to land. The ruling class had only a modest degree of direct access to land, as most of it was under peasant control, and so were the associated means of production, mainly oxen. As Crummey (1980: 131) outlined it: “ploughs were simple, irrigation elementary, the exploitation of water-power non-existent and of animal-power low…”. Farmland was, thus, held on an individual basis by peasants, while pastures were common land. Most peasants had direct access to the land through hereditary, normally inalienable rights (rest), for which they paid a royal tithe; in addition to this, they might pay a tribute (a portion of their production). All forms of land tenure were conditioned by rights that frequently overlapped and, in practice, they might be held by farmers, holders of overrights (gwelt), and the court (Crummey, 2007). The farmers of medieval Ethiopia, thus, enjoyed a certain degree of independence from the state and their lords. The control of agricultural production, including the decisions about what to plant, where and when, was in the hands of farming households. Ethiopian nobles, churches, and endowed monasteries, even though they acquired titles to extensive lands, never converted these nominal rights into estates united by a centrally directed economy (Crummey, 2000). If ancient Aksumite farmers enjoyed a certain degree of independence, similar to the farmers of medieval Ethiopia since the thirteenth century, it would be expected that changes of ruling groups (kings, chiefs, nobles) would not have affected land and water management practices signifi­ cantly.19 The available documentation points to an efficient management of water resulting from an ad hoc balance of soil/landscape factors and land-use strategies. As argued by D’Andrea et al. (2008), dry-farming fields may have been strategically placed in areas where runoff is naturally concentrated, and limited small-scale water diversion and terracing may have been used to cultivate lands in less favorable topographic positions. Aksum’s countryside appears to have been farmed continuously since at least the late first millennium bc. Although direct information is lacking, it has been suggested that Aksum hosted a population of several thousands in its heyday.20 Firm data on population density are not available until the early twentieth century, when historical sources record a figure of 10,000 for the inhabitants of Aksum (e.g. Consociazione Turistica Italiana, 1938:

173-197 WaterUrban Pt2 Ch08 Sulas.indd 188

12/05/2014 11:22

Aksum: Water and Urbanization in Northern Ethiopia189

259). At this time, the quarters of the town had different names and, among those listed in the Liber Axumae, six were still in use (see Monneret de Villard, 1938: 13; Consociazione Turistica Italiana, 1938: 261): Malake Aksum, Kuduku, Bagi≤o, Nefas, Farheba, and ≤Ahorò (Figure 8.3). Most of these quarters unfolded along the banks of the ˘May Heg ˇˇga, the pediment of Betä Giyorgis, and the little watercourse of May Lahlah. During the rainy season, the plain was drained by the May Heg ˇˇga to the east and the May ≤Abagät (May ≤Ayni) to the west, and several wells and the large cistern of May Šum supplied water for domestic use. Throughout the twentieth century, the understanding of the relationship between the development of subsistence systems and environmental change at Aksum was based on a combination of multi-scale and multidisciplinary data. In particular, the narrative about the rise and decline of state societies in the region has emphasized the role of rainfall changes: increased rainfall would have boosted agricultural productivity to sustain state formation first; and, subsequently, it had a detrimental impact on an overexploited landscape, which eventually led to the demise of the kingdom in the later first millennium ad. This model heavily draws from regional and continental environmental proxies and context-specific archaeological records. On the one hand, Aksumite urban development has long been linked to the idea of an agricultural substratum dependent on irrigation. On the other hand, regional and continental records of rainfall changes have been directly applied to explain cultural and economic developments in the Aksum area. This type of argument restrains societal responses to environmental change and, significantly, fails to account for other factors, including hydrological processes. It is now clear that neither woodland nor irrigation were a common feature of the local landscape before and during the Aksumite period. A greater amount of environmental and archaeological records are needed to contextualize non-environmental factors affecting societal decision-making over time. Indeed, the paucity of local landscape records implies that any attempt at modeling synchronous histories of forest expansion/clearance or the impact of changing rainfall, to mention just two important topics, has to rely on assumptions of environmental and cultural uniformity across vast regions. In the Zambezi region of southern Africa, for example, negative environmental consequences such as droughts, excess rainfall, and flooding pushed ancient societies to converge into social formation beyond the village and, in some cases, to develop state organizations (Pikirayi, in press). Some of these early states depended on groundwater resources, rather than bounding themselves to the major rivers present in the region. The Zambezi region, thus, provides an example of how people responded to unfavorable environmental conditions beyond the commonly cited scenario of environmental degradation, followed by a “collapse”. Indeed, the literature on past societies often emphasizes direct connections

173-197 WaterUrban Pt2 Ch08 Sulas.indd 189

12/05/2014 11:22


A History of Water

between population increase and intensification of resource use, resulting in environmental degradation and collapse. Abandonment of the main urban centers is irredeemably taken as an expression of collapse, but history shows how such direct connections can be misleading (see, for example, Tainter, 2006). Furthermore, the abandonment of urban quarters does not necessarily imply the disappearance of a culture (Fletcher, 2011). At Aksum, the decline in occupation recorded for the late first millennium ad has been described in terms of “abandonment”. However, recent research has shed new light on the trajectories of post-Aksumite and medieval Ethiopia (see Phillipson, 2012), and there is now evidence that the disappearance of elite settlement did not mark the “disappearance of a tradition of social life” (cf. Fletcher, 2011: 311). At Aksum, the “traditional” ox-plough farming persisted despite significant changes in power and authority over the last 3,000 years. Beginning with the shift of political focus from Betä Giyorgis hilltop onto the Aksum plain in the late first millennium bc, a change in power is likely to have played a significant role in the decline of the permanently based Aksumite civilization in the late first millennium ad. The subsequent state society was mobile and emerged in the central highlands of Ethiopia with the establishment of the Solomonic Dynasty (ad 1270). More recently, a shift occurred in resource management and power associated with the modernization policy of Hayla Sellase (r. 1930–74), which since the 1950s has sought to reform agriculture by implementing land fragmentation, damming watercourses, and large-scale irrigation (Kloos, 1991; Kiros, 1993; Abegaz, 2005). For the first time in Ethiopia’s history, decisions on water and land use were taken in the capital (Addis Ababa), rather than by local farmers, and the control of water became a symbol of political and socioeconomic power. Conclusion Water management was, and still is, an important aspect of the subsistence of northern Ethiopian and Eritrean farmers. In the highlands, abundant rains, suitable temperature, good soil moisture retention, and ground­water supplied water for ancient farming and households. The arid lowlands and coastal plains of modern Eritrea, instead, needed substantial watering to grow cereals and pulses; this may have been achieved by flood irrigation. The apparent lack of centralized control of water resources may be an indication that the Aksumite kings exerted power through means other than controlling access and management of water and land. But water, as is often the case, has been an “agent in the constitution and continuity of societies” (Tvedt and Oestigaard, 2010: 15) of northern Ethiopia and Eritrea. In spite of sociopolitical shifts that punctuated the history of this region, “traditional” plough farming has continued for almost two millennia.

173-197 WaterUrban Pt2 Ch08 Sulas.indd 190

12/05/2014 11:22

Aksum: Water and Urbanization in Northern Ethiopia191

Small-scale land and water management was sustainable in times of both population expansion and contractions. Moreover, a farming cycle governed by a bi-modal climate with long dry seasons and short rainy periods allowed several months for farmers to engage in other activities, which might have been controlled by the central state (McCann, 2001; Sulas et al., 2009). The findings of recent research open up new avenues for reframing the settlement trajectories in the Ethiopian and Eritrean highlands from the late Aksumite onwards. While the Aksumite legacy was feeding into the emergence of the Ethiopian Empire, settlement continued at Aksum. The example from Aksumite and, later, historical Ethiopia is of immense importance for framing discourses on water history and land tenure in subSaharan Africa. Besides dynastic Egypt, archaeological and documentary records for early forms of land and water management are only available for the Ethiopian and Eritrean highlands. AcknowledgEments I wish to thank Terje Oestigaard, Terje Tvedt and Paul Sinclair for stimulating discussions on water and society in Ethiopia. My warmest gratitude goes to Laqe Giorgis and Gosh Assefa, who taught me about land and water resources while surveying Aksum’s countryside. I am also in debt with Innocent Pikirayi, Donald Crummey and Alex Metcalfe for comments on an early draft that have helped in clarifying some of the arguments put forward in this chapter. Any mistake or inaccuracy is, by all means, the full responsibility of the author. Notes   1 The transliteration of Ethiopian names follows, wherever possible, the system used in the Encyclopaedia Aethiopica (2003–10), Wiesbaden: Harrassowitz.   2 The term “Aksumite” embraces several meanings, being used to refer to the cultural traits of northern Ethiopian and Eritrean archaeology, and their chronology. Here, I follow the cultural sequence developed by Fattovich (2008, 2010) and Phillipson (2000, 2009, 2012). “Pre-Aksumite” is used for the cultural developments of the first millennium bc (c.1000/900–400/300 bc) in western and central Tigray; “Aksumite” refers to the kingdom of Aksum, dating between c.50 bc and ad 700/800. The term “Proto-Aksumite” refers to the intermediate period (c.400–50 bc) between the decline of the Pre-Aksumite culture and the rise of the Aksumite kingdom.  3 The rainfall mean is, thus, nearly level with the evapotranspiration mean, suggesting that in the context of rainfed farming, yields are secured by balancing water gaining (rainfall) and losing (evapotranspiration).

173-197 WaterUrban Pt2 Ch08 Sulas.indd 191

12/05/2014 11:22


A History of Water

 4 Groundwater recharge preferentially occurs at the unconfined area of elevated topography, which consists of the hills made of fractured phonolite and trachyte; see Alemayehu et al. (2011).   5 Cereals tend to be the most important crop within the rotation, providing a highly nutritious food source. Legumes are also an essential part of the rotation, as they fix nitrogen, allowing a build-up of soil fertility; see McCann (1995: 50–2).   6 Changing soil moisture levels have been poorly understood in the past as well as in the present day, although they are clearly a key factor in active landscape modification and resource management.   7 The Near Eastern crops identified in Pre-Aksumite and Aksumite contexts are free-threshing wheat, emmer, barley, lentil, linseed/flax, and grape. The African crops attested in Aksumite contexts are t·ef and finger millet; see D’Andrea (2008).  8 At present, there is no archaeological record of cultivation of species requiring more water than those attested in Aksumite times, and that are still cultivated today.   9 These include the remains of a substantial building at Da≤aro, two churches, and a lithic workshop; see Sernicola (2008). 10 ≤Ela Nägäst means “well of the kings” and, according to local traditions, this name was used to refer to ≤Ona Nägäst; see Fattovich et al. (2000: 26). 11 In particular, the structure was enlarged and reinforced in 1927 under the supervision of A. Pollera; see Consociazione Turistica Italiana (1938: 264). 12 The cistern was rebuilt about 30 years ago, with no respect for its authentic appearance. Today, it supplies water for about 100 households, though the Safra administrative region has a population of just about 2,500 people; see Raadvad (2007: 96). 13 For a review, see Schneider (1996), who also mentions a further link with water inferred from the regalia on two coins of the Aksumite king Afilas (c.300 ad). The latter depict the king’s bust framed by semicircles that are interpreted as water gushes, indicating that Aksum was located near a river or remarkable water source. 14 The Liber Axumae (the original name is Mäshafä Aksum, or “Book of Aksum”) is a composite work incorporating information on the historical topography of Aksum, a detailed description of its Cathedral of Maryam Seyon, and the church cartulary record of endowments dating from the fifteenth to the nineteenth centuries; see critical edition by Conti Rossini (1909) and discussion by Lusini (2003). 15 Conti Rossini’s (1909: 3) translation of the Ge≤ez reads: ‘Pour la troisième fois, la ville fut édifiée par Abreha¯ et Asbeha… mais sa construction s’accomplit à l’aide d’un miracle et d’un prodige. En effet, jadis il y avait là une grande étendue d’eau…’; see also Beckingham and Huntingford (1961: 521–5). The tradition is preserved in Ms d’Abb. 97, the source of Conti Rossini, which possibly dates to the fifteenth century; see Monneret de Villard (1938: 50–2) and Munro-Hay (2003: 183). 16 The marked place was moved further south sometime before the 1860s, when it was first recorded in its current location. In the late 1890s, Wylde (1901: 146) noted how the May Hegˇgˇa, after running across the “sacred grove” (next to Cathedral’s precinct), cuts through the “market green”, before dying into the valley. Monneret de Villard (1938) suggests that the moving of the market

173-197 WaterUrban Pt2 Ch08 Sulas.indd 192

12/05/2014 11:22

Aksum: Water and Urbanization in Northern Ethiopia193

may have been influenced by a progressive desiccation of the plain to the south of the town. 17 Water scarcity affects some 1.6 billion people worldwide and, by 2025, twothirds of the world’s population are expected to be living under water-stressed conditions; see Gilbert (2012) and United Nations’ press release on March 22, 2012: www.fao.org/news. 18 The May Hegˇgˇa watercourse has different names: May Melahso for the upstream section, May Hegˇgˇa for the central part, and May Mat·are for the downstream section. 19 Crummey (2000: 11) argues that the “class relations which characterised the Solomonic state and its predecessors on the Ethiopian highland plateau accompanied the making of those states and were embedded in their very fabric”. 20 For example, Michels (2005: 158) derives a stunning figure of just over 39,000 people living in the greater Aksum area at the peak of urban expansion. This estimate is based upon settlement models and should be taken with great caution, but at least it reflects the extent of residence sites and structures recorded at Aksum.

references Abegaz, B. (2005). “Persistent stasis in a tributary mode of production: the peasant economy of Ethiopia”, Journal of Agrarian Change, 5(3), pp. 299–333. Alemayehu, T., A. Leis, A. Eisenhauer, and M. Dietzel (2011). “Multi-proxy approach (2H/H, 18O/16O, 13C/12C and 87Sr/86Sr) for the evaluation of carbonaterich groundwater in basalt dominated aquifer of the Axum area, northern Ethiopia”, Chemie der Erde, 71, pp. 177–87. Assefa, G. and G. Russo (1997). “General geology of the Aksum area”, in K. A. Bard (ed.), The Environmental History and Human Ecology of Northern Ethiopia in the Late Holocene, pp. 11–18, Naples: Istituto Universitario Orientale. Bard, K. A., M. Coltorti, M. C. DiBlasi, F. Dramis, and R. Fattovich (2000). “The environmental history of Tigray (northern Ethiopia) in the middle and late Holocene: a preliminary outline”, African Archaeological Review, 17(2), pp. 65–86. Barradas, M. (1634 [1996]). Tractatus tres historico-geographici (1634) (ed. R. Pankhurst), Wiesbaden: Harrassowitz. Beckingham, C. F. and G. W. B. Huntingford (eds) (1961). The Prester John and the Indies: a true relation of the lands of Prester John being the narrative of the Portuguese embassy to Ethiopia in 1520 written by Father Francisco Alvarez, London: The Hakluyt Society. Bent, J. T. (1893). The Sacred City of the Ethiopians, London: Longmans. Boardman, S. (1999). “The agricultural foundations of the Aksumite Empire, Ethiopia: an interim report”, in M. van der Veen (ed.), The Exploitation of Plant Resources in Ancient Africa, pp. 137–47, New York: Plenum Publishing Corporation. Bruce, J. (1790). Travels to Discover the Source of the Nile in the Years 1768, 1769, 1770, 1771, 1772 and 1773, Vol. 3, Edinburgh: Robinson.

173-197 WaterUrban Pt2 Ch08 Sulas.indd 193

12/05/2014 11:22


A History of Water

——— (1813). Travels to Discover the Source of the Nile in the Years 1768, 1769, 1770, 1771, 1772 and 1773, Vol. 5, Edinburgh: Ramsay and Company. Brunner, U. (2005). “Water management and settlement in ancient Eritrea”, in W. Raunig and S. Wenig (eds), Afrikas Horn: Akten der Ersten Internationalen Littmann-Konferenz 2. Bis 5. Mai 2002 in München, pp. 30–43, Wiesbaden: Harrassowitz. Butzer, K. W. (1981). “Rise and fall of Axum, Ethiopia: a geo-archaeological interpretation”, American Antiquity, 46(3), pp. 471–95. ——— (2012). “Collapse, environment, and society”, Proceedings of the National Academy of Science USA, 109(10), pp. 3632–9. Coltorti, M., F. Dramis, and C. D. Ollier (2007). “Planation surfaces in northern Ethiopia”, Geomorphology, 89(3–4), pp. 287–96. Consociazione Turistica Italiana (1938). Guida dell’Africa Orientale Italiana, Milan: Consociazione Turistica Italiana. Conti Rossini, C. (1904). Vitae sanctorum antiquiorum (Gadla Yarpd), seu Acta Sancti Yarpd (C.S.C.O., Scriptores Aethiopici, Series Altera, t. XVII), Rome: K. de Luigi. ——— (1909). Documenta ad illustrandam historiam. I. Liber Axumae (C.S.C.O., Scriptores Aethiopici, Series Altera, t. VIII), Paris: E. Typographeo Reipublicae. ——— (1928). Storia d’Etiopia, 2 vols, Bergamo: Istituto Italiano d’Arti Grafiche. ——— (1942). Proverbi tradizioni e canzoni tigrine, Verbania: A. Airoldi. Crummey, D. (1980). “Abyssinian feudalism”, Past & Present, 89, pp. 115–38. ——— (2000). Land and Society in the Christian Kingdom of Ethiopia: From the Thirteenth to the Twentieth Century, Oxford: James Currey. ——— (2007). “Land tenure”, in S. Uhlig (ed.), Encyclopaedia Aethiopica He-N, pp. 496–9, Wiesbaden: Harrassowitz. D’Andrea, A. C. (2008). “T’ef (Eragrostis tef) in ancient agricultural systems of highland Ethiopia”, Economic Botany, 62(4), pp. 547–66. D’Andrea, A. C., A. Manzo, M. J. Harrower, and A. L. Hawkins (2008). “The Pre-Aksumite and Aksumite of NE Tigray, Ethiopia”, Journal of Field Archaeology, 33(2), pp. 151–76. D’Andrea, A. C., M. P. Richards, L. A. Pavlish, S. Wood, A. Manzo, and H. S. Wolde-Kiros (2011). “Stable isotopic analysis of human and animal diets from two pre-Aksumite/Proto-Aksumite archaeological sites in northern Ethiopia”, Journal of Archaeological Science, 38, pp. 367–74. Dramis, F., M. Umer, G. Calderoni, and M. Haile (2003). “Holocene climate phases from buried soils in Tigray (northern Ethiopia): comparison with lake level fluctuations in the Main Ethiopian Rift”, Quaternary Research, 60, pp. 274–83. Fattovich, R. (2008). Kings and Farmers. The Urban Development of Ancient Aksum (Tigray, Ethiopia), ca. 700 bc–ad 1500 (Working Papers in African Studies 4), Boston: African Studies Center, Boston University. ——— (2010). “The development of ancient states in the northern Horn of Africa, c.3000 bc–ad 1000: an archaeological outline”, Journal of World Prehistory, 23, pp. 145–75. Fattovich, R., K. A. Bard, L. Petrassi, and V. Pisano (2000). The Aksum Archaeological Area: A Preliminary Assessment, Naples: Istituto Universitario Orientale. Finneran, N. (2007a). The Archaeology of Ethiopia, London: Routledge.

173-197 WaterUrban Pt2 Ch08 Sulas.indd 194

12/05/2014 11:22

Aksum: Water and Urbanization in Northern Ethiopia195

——— (2007b). “May Kerwah”, in S. Uhlig (ed.), Encyclopaedia Aethiopica He-N, p. 386, Wiesbaden: Harrassowitz. Fletcher, R. (2011). “Low-density, agrarian-based urbanism. Scale, power, and ecology”, in M. E. Smith (ed.), The Comparative Archaeology of Complex Societies, pp. 285–320, Cambridge: Cambridge University Press. French, C., F. Sulas, and M. Madella (2009). “New geoarchaeological investigations of the valley systems in the Aksum area of northern Ethiopia”, Catena, 78(3), pp. 218–33. Gebru, T., Z. Eshetu, Y. Huang, T. Woldemariam, N. Strong, M. Umer, M. C. DiBlasi, and V. Terwilliger (2009). “Holocene palaeovegetation of the Tigray Plateau in northern Ethiopia from charcoal and stable organic isotopic analyses of gully sediments”, Palaeogeography, Palaeoclimatology, Palaeoecology, 282, pp. 67–80. Gilbert, N. (2012). “Water under pressure”, Nature, 256(483), pp. 256–7. Harlan, J. R. (1971). “Agricultural origins: centers and non-centers”, Science, 174(4008), pp. 468–74. Hassan, F. (2007). “Extreme Nile floods and famines in medieval Egypt (ad 930–1500) and their climatic implications”, Quaternary International, 173–4, pp. 101–12. Kiros, F. G. (1993). The Subsistence Crisis in Africa: The Case of Ethiopia, Nairobi: Organisation for social science research in Eastern Africa. Kloos, H. (1991). “Peasant irrigation development and food production in Ethiopia”, The Geographical Journal, 157, pp. 295–306. Littmann, E., D. Krencker, and Th. Von Lüpke (1913). Deutsche Aksum-Expedition, 4 vols, Berlin: Reimer. Lucero, L. J., J. D. de Gunn, and V. L. Scarborough (2011). “Climate change and classic Maya water management”, Water, 3(2), pp. 479–94. Ludolf, H. (1684). A New History of Ethiopia (2nd edn), London: Samuel Smith. Lusini, G. (2003). “Aksum: Mäshafä Aksum”, in S. Uhlig (ed.), Encyclopaedia Aethiopica A-C, pp. 185–6, Wiesbaden: Harrassowitz. MacDonald, A. M., H. C. Bonsor, B. É. Ó. Dochartaigh, and R. G. Taylor (2012). “Quantitative maps of groundwater resources in Africa”, Environmental Research Letters, 7(2), doi:10.1088/1748-9326/7/2/024009. Machado, M. J., A. Pérez-González, and G. Benito (1998). “Palaeoenvironmental changes during the last 4000 yr in the Tigray, northern Ethiopia”, Quaternary Research, 49, pp. 312–21. Marshall, M. H., H. F. Lamb, S. J. Davies, M. J. Leng, Z. Kubsa, M. Umer, and C. Bryant (2009). “Climatic change in northern Ethiopia during the past 17,000 years: a diatom and stable isotope record from Lake Ashenge”, Palaeogeography, Palaeoclimatology, Palaeoecology, 279, pp. 114–27. McCann, J. (1995). People of the Plow: An Agricultural History of Ethiopia, 1800– 1990, Madison, WI: The University of Wisconsin Press. ——— (2001). “Variabilità climatica e classi sociali nell’altopiano dell’Etiopia”, Storia Urbana, 95, pp. 71–92. Michels, J. W. (2005). Changing Settlement Patterns in the Aksum-Yeha region of Ethiopia: 700 bc–ad 850 (BAR International Series 1446), Oxford: Archaeopress. Monneret de Villard, U. (1938). Aksum: Ricerche di topografia generale, Rome: Pontificium Institutum Biblicum.

173-197 WaterUrban Pt2 Ch08 Sulas.indd 195

12/05/2014 11:22


A History of Water

Mountnorris, G. A., Earl of (Viscount Valentia) (1809). Voyages and Travels in India, Ceylon, the Red Sea, Abyssinia, and Egypt: in the Years 1802, 1803, 1804, 1805, and 1806, Vol. 3, London: W. Miller. Munro-Hay, S. C. (1991). Aksum: An African Civilization of Late Antiquity, Edinburgh: Edinburgh University Press. ——— (2003). “Aksum Seyon”, in S. Uhlig (ed.), Encyclopaedia Aethiopica A-C, pp. 183–5, Wiesbaden: Harrassowitz. Nyssen, J., H. Vendenreyken, J. Poesen, J. Moeyersons, J. Deckers, M. Haile, C. Salles, and G. Grovers (2005). “Rainfall erosivity and variability in the northern Ethiopian highlands”, Journal of Hydrology, 311, pp. 172–87. Pankhurst, R. (1985). The History of Famine and Epidemics in Ethiopia Prior to the Twentieth Century, Addis Ababa: Relief and Rehabilitation Commission. Parkyns, M. (1853 [2005]). Life in Abyssinia, London: Elibron Classics. Phillipson, D. W. (1997). The Monuments of Aksum, Addis Ababa: Addis Ababa University Press and British Institute in Eastern Africa. ——— (1998). Ancient Ethiopia, London: The British Museum Press. ——— (2000). Archaeology at Aksum, Ethiopia, 1993–7, London: British Institute in Eastern Africa and Society of Antiquaries of London. ——— (2009). “The first millennium bc in the highlands of northern Ethiopia and south-central Eritrea: a reassessment of cultural and political development”, African Archaeology Review, 26, pp. 257–74. ——— (2012). Foundations of an African Civilisation: Aksum and the Northern Horn 1000 bc–ad 1200, Suffolk, UK and Rochester, NY: James Currey. Pikirayi, I. (in press). “Water and social formation in pre-colonial Zambezia: rethinking the development and demise of complex societies in southern Africa”, in V. Scarborough (ed.), Water History and Humanity, Paris: UNESCO. Raadvad, T. (2007). The Archaeological Site and Cultural Landscape of Qohaito, Eritrea: Site Management and Implementation Plan, Lingby, Denmark: Consulting Architects MAA. Ricci, L. (1974). “Scavi archeologici in Etiopia”, Africa (Rome), 29(3), pp. 129–75. ——— (1990). “Appunti archeologici”, Rassegna di Studi Etiopici, 32, pp. 129–65. Salt, H. (1814). A Voyage to Abyssinia and Travels in the Interior of that Country Executed under the Orders of the British Government in the Years 1809 and 1810, London: Rivington. Scarborough, V. L. (2009). “Beyond sustainability: managed wetlands and water harvesting in ancient Mesoamerica”, in C. T. Fisher, J. B. Hill, and G. M. Feinman (eds), The Archaeology of Environmental Change: Socionatural Legacies of Degradation and Resilience, pp. 62–82, Tucson, AZ: University of Arizona Press. Schmidt, P. R., M. C. Curtis, and Z. Teka (eds) (2008). The Archaeology of Ancient Eritrea, Asmara: Red Sea Press. Schneider, R. (1996). “Remarques sur le nom ‘Aksum’”, Rassegna di Studi Etiopici, 38, pp. 183–90. Sernicola, L. (2008). “Il modello d’insediamento dell’altopiano tigrino (Etiopia settentrionale / Eritrea centrale) in epoca Pre-Aksumita e Aksumita (ca. 700 a.C.–800 d.C.). Un contributo da Aksum”, unpublished PhD dissertation, University of Naples “L’Orientale”. Sernicola, L. and F. Sulas (2012). “Continuità e cambiamento nel paesaggio rurale di Aksum: dati archeologici, etnografici e paleoambientali”, in A. Bausi,

173-197 WaterUrban Pt2 Ch08 Sulas.indd 196

12/05/2014 11:22

Aksum: Water and Urbanization in Northern Ethiopia197

A. Brita and A. Manzo (eds), Æthiopica et Orientalia: Studi in onore di Yaqob Beyene, Vol. 1, pp. 549–74, Naples: University of Naples “L’Orientale”. Simoons, F. J. (1960). Northwest Ethiopia: Peoples and Economy, Madison, WI: University of Wisconsin Press. Sulas, F. (2010). “Environmental and cultural interplay in highland Ethiopia: Geoarchaeology at Aksum”, unpublished PhD dissertation, University of Cambridge. Sulas, F., M. Madella, and C. French (2009). “State formation and water-resource management in the Horn of Africa: the Aksumite Kingdom of the northern Ethiopian highlands”, World Archaeology, 41(1), pp. 2–15. Tadesse, N., S. Tadios, and M. Tesfaye (2010). “The Water Balance of May Negus Catchment, Tigray, Northern Ethiopia”, Agricultural Engineering International: CIGR Ejournal, 12(1306), pp. 1–29. Tainter, J. A. (2006). “Archaeology of overshoot and collapse”, Annual Review of Anthropology, 35, pp. 59–74. Terwilliger, V. J., Z. Eshetu, Y. Huang, M. Alexandre, M. Umer, and T. Gebru (2011). “Local variation in climate and land use during the time of the major kingdoms of the Tigray Plateau in Ethiopia and Eritrea”, Catena, 85, pp. 130–43. Tvedt, T. and T. Oestigaard (2010). “A history of the ideas of water: deconstructing nature and constructing society”, in T. Tvedt and T. Oestigaard (eds), A History of Water. Volume 1, The Idea of Water, from Ancient Societies to the Modern World, pp. 1–36, London: I.B.Tauris. Vigoni, P. (1881). Abissinia. Giornale di un viaggio, Milan: Hoepli. Wenig, S. (1997). “Fieldwork in Eritrea”, Nyame Akuma, 48, pp. 20–1. Wenig, S. and M. C. Curtis (2008). “Qohaito: an ancient highland urban center”, in P. R. Schmidt, M. C. Curtis, and Z. Teka (eds), The Archaeology of Ancient Eritrea, pp. 287–300, Asmara: Red Sea Press. Wylde, A. B. (1901). Modern Abyssinia, London: Methuen & Co.

173-197 WaterUrban Pt2 Ch08 Sulas.indd 197

12/05/2014 11:22


Machu Picchu: Water Engineering in the Mountains

Kenneth R. Wright Introduction Machu Picchu is a magnificent royal estate in a spectacular mountaintop location. Water engineering by the Incas was key to making this site viable and special. They developed five different types of settlements during their nearly 100-year empire: (1) the Inca capital of Cusco; (2) administrative capitals located at strategic places across the empire; (3) wayside stations called tambos where travelers could rest; (4) villages that served as agricultural or other resource/labor centers; and (5) royal estates (Salazar, 2004). Machu Picchu falls into the last category, as evidenced by its remoteness, its high-status design and construction, its many temples, lack of economic infrastructure, and limited housing. Machu Picchu supported a resident population of about 300 people, with a maximum of 1,000 when the royal entourage visited (Rowe, 1990). Machu Picchu was established by the Inca ruler Pachacuti to commemorate his military conquest in the middle and lower Urubamba valleys around ad 1450. It was a retreat where Pachacuti and his elite guests “engaged in celebrations, diplomatic feasting, religious ceremonies and rituals, astronomical observations and administrative affairs of the empire” (Salazar, 2004). The grandness of Machu Picchu would have reminded Pachacuti’s guests, and perhaps his subjects, of his power and divinity. Machu Picchu was abandoned in ad 1572, but it likely ceased operation by ad 1540 due to the earlier collapse of the Inca Empire (Maurtua, 1906; Rowe, 1990). The Machu Picchu site is well suited for a royal estate because it is characterized by special examples of features revered by the Incas—water and mountains. Machu Picchu is surrounded on three sides by the steep, roaring Urubamba River, 450 m below, and located between two regional faults that created the two sharp peaks of Huayna Picchu and Machu Picchu Mountains. This made Machu Picchu a prime location for the Incas to worship water and apus (mountain gods). Furthermore, the development of Machu Picchu in such a remote location served to emphasize Pachacuti’s dominion over the land and the people, as he led his workers to physically alter the sacred landscape.

198-220 WaterUrban Pt2 Ch09 Wright.indd 198

12/05/2014 11:22

Machu Picchu: Water Engineering in the Mountains199

The non-profit research foundation Wright Paleohydrological Institute (WPI)1 performed investigations between 1995 and 2004 on Incan water use and handling at Machu Picchu. Although the Incas did not have a written language, the well-preserved remains of Machu Picchu show that they had an advanced understanding of such principles as urban planning, hydrology, hydraulics, drainage, and durable construction methods. However, without a water supply, the site never could have existed. The water engineering performed in this mountainous environment is a testament to the skills of the Incas. A canal system conveyed spring water to Machu Picchu, where it flowed through a series of well-crafted fountains and then into the drains; it infiltrated the soil and eventually ran into the river below. This system, although on one level a convenient water supply for the populace of the site, is also a reminder of the revered water cycle regulated by the Incas. Water falls from the sky, over the mountains and into the forested catchment basin; it is then collected into the canal system from which it nurtured human life, before it flows into the river and finally the sea. Ultimately, Machu Picchu served as a place where the divine emperor was able to commune with the mountain deities and take part in the great cycle of life, land, and water. The planners and civil engineers of the Incan Empire are responsible for the achievement of Machu Picchu. Mountainous environment Machu Picchu is about 1,400 km south of the equator on the eastern slope of the Peruvian Andes. The site lies near the headwaters of the Amazon River, at longitude 72º 32’ and latitude 13º 9’. The mountain of Huayna Picchu (2,720 m) flanks the developed site on the north, while the mountain of Machu Picchu (2,760 m) juts up in the southeast. The dramatic location is a result of tectonic forces and valley downcutting by the Urubamba River. Across the river from Machu Picchu to the northeast is the 2,560 m Putucusi peak that has a rounded profile like a half-orange when viewed from Machu Picchu. Even today, all three peaks are considered to be holy mountains by the local Quechua Indians. The Machu Picchu site enjoys magnificent views of other mountains that were holy to the Incas. To the east is the triangular-shaped mountain called Yanantin (4,486 m) and the glacier-capped flanks of Mount Veronica (5,850 m). Some 20 km south is Mount Salcantay at an elevation of 6,257 m. The importance of Mount Salcantay to the Incas can be inferred from the fact that an arrow stone at Machu Picchu’s holiest location, the Intiwatana summit, points to it. Machu Picchu is geologically located within the Cordillera Oriental (Eastern Cordillera), between the High Plateau and subandine zones of the Peruvian Andes (Marocco, 1977) on a portion of the complex Vilcabamba

198-220 WaterUrban Pt2 Ch09 Wright.indd 199

12/05/2014 11:22

200A History of Water

Batholith 40 km2 in size (Caillaux, n.d.). This 250-million-year-old intrusion is white-to-gray-colored granite, characterized by its abundance of quartz, feldspar, and mica (predominantly biotite). The composition of the mineral rendered the granite of Machu Picchu a sturdy and longlasting construction material. The Incan workmen masterfully utilized the inherent, rectangular joint pattern of the granite building stones to construct Machu Picchu. The competent use of this durable stone guaranteed the retreat’s longevity. Machu Picchu’s most significant geologic features are the numerous faults and abundant rock fractures. The two main faults are named for two prominent local peaks: the Huayna Picchu fault and the Machu Picchu fault. These high-angle reverse faults formed a wedge-shaped structural block that dropped relative to the peaks on either side. This block, or graben, is shown in Figure 9.1, with Machu Picchu situated on the graben. Relationship of water to Incan society Water was important to the Incas. The crowning achievements of the most carefully wrought Inca landscape architecture are glorious fountains that filled the environment with the sight and sound of falling water. Examples exist at Tipón, Ollantaytambo, Tambomachay, Winaywayna, and many other places developed by the Incas. Fountains were used for ritual purification. The importance of water to the Incas is also evident in many of the religious practices that the modern Quechua Indians inherited from the Incas. Springs are worshipped, as are flowing fountains, rain, thunderstorms, and the sea. The Incas worshipped gods whose identities related to water, such as Pariacaca, god of water, Kon, the rain god, Mama Cocha, the sea mother, and Paricia, a god who punished people with floods (Cobo, 1990). It is impossible to separate the idea of water from its role in agriculture. Without adequate water, sufficient crops cannot be grown. Communal ownership of land and surplus food was a fundamental practice of many Andean societies. The Incan state was efficient, storing extra food in warehouses to protect against times of famine. This insured the Incas against the volatility of nature and allowed them to demand that a large portion of the people, who would have otherwise been farmers, work on other tasks like soldiering, the construction of roads, and building sites like Machu Picchu. Site overview Most people, both then and now, approach Machu Picchu through its Main Gate. The Main Gate is a doorway of precise construction located

198-220 WaterUrban Pt2 Ch09 Wright.indd 200

12/05/2014 11:22

Machu Picchu: Water Engineering in the Mountains201

Figure 9.1.  Machu Picchu is laid out along the top of a huge graben ridge. The water supply drainage area is shown beyond, on the north slope of Machu Picchu Mountain.

198-220 WaterUrban Pt2 Ch09 Wright.indd 201

12/05/2014 11:23

202A History of Water

almost dead center on the western side of the oblong development. To the south of the Main Gate are agricultural terraces, and to the north is the developed urban sector that is home to most of Machu Picchu’s architectural and religious monuments. From the Main Gate heading northward, some of the most famous features are the Temple of the Sun, the Stairway of the Fountains, the Royal Residence, the Temple of the Three Windows, the Intiwatana Pyramid, the Artisan’s Wall, the Plaza, and the Sacred Rock. Dividing the agricultural and urban sectors is the Main Drain, which collects and funnels away stormwater. The general layout of Machu Picchu is shown in Figure 9.2. Hydrology On an elevated mountain ridge between two prominent peaks is not an intuitive location for ancient people to have located a reliable and potable source of groundwater. Machu Picchu would not have been built at this location had it not been for the availability of the perennial springs that the Incan engineers discovered and developed on the steep north face of Machu Picchu Mountain. Nearly 2,000 mm of annual rainfall, a modestsized tributary drainage basin, igneous bedrock, and extensive faulting associated with the site collectively supplied Pachacuti and his engineers with a dependable local source of water. The elevation of the steep drainage basin tributary to the Inca springs ranges from 2,458 m to 3,050 m, a vertical rise of 592 m. This tributary drainage basin covers 16.3 ha, including two sub-basins of 5.9 and 10.4 ha. It was possible to demarcate the drainage basins after exhaustive field examinations, site observations, photographic interpretation, and use of topographic mapping. The drainage basin is well covered with tropical forest vegetation, and bisected by the final section of the ancient Inca Trail from Intipunku at the ridge top, to the royal estate of Machu Picchu. An inflow–outflow assessment was performed for the 10.4 ha topographic drainage basin, using the observed annual rainfall of 1,960 mm. The results were estimates of annual evapotranspiration of the forest cover of 1,760 mm and an approximate spring yield of 40,000 m3 per year. The spring was assumed to be 100 percent efficient and a mean daily evapotranspiration of 4.82 mm was used (Wright et al., 1997a). The function of this evaluation was to ascertain whether the Machu Picchu fault was responsible for a hydrogeologic zone of capture greater than that defined by the topographic ground-surface drainage basin. The conclusion was that the yield of the primary spring represents drainage from a hydrogeologic catchment basin as much as double the size of the topographic drainage basin. This tends to correlate with field surveys made on the north slope of the mountain of Machu Picchu, as well as reviews of topographic mapping.

198-220 WaterUrban Pt2 Ch09 Wright.indd 202

12/05/2014 11:23

Machu Picchu: Water Engineering in the Mountains203

Figure 9.2.  Hiram Bingham mapped Machu Picchu in 1913, which WPI refined over the course of several field studies.

198-220 WaterUrban Pt2 Ch09 Wright.indd 203

12/05/2014 11:23

204A History of Water

Since Hiram Bingham first learned of and shared the existence of Machu Picchu with the world outside Peru in 1911, it has been theorized that the royal retreat was abandoned in 1540 in response to a scarcity of water. This conjecture has no merit when the Quelccaya Glacier’s ice core data are analyzed (Wright et al., 1996). The ice core data suggest that the “Little Ice Age” began in about ad 1500 and the precipitation for the final decade of occupancy was greater than that during any of the other decades of the period of habitation (Wright et al., 1997b). Before the emperor chose Machu Picchu for his royal estate, the civil engineers would have had to verify that the water supply was viable. While the Machu Picchu primary spring is a natural phenomenon, its consistent yield is augmented by a novel and well-engineered stone spring collection system that functions to this day. It is on the north slope of Machu Picchu Mountain at an elevation of 2,458 m. This ancient example of groundwater and hydraulic engineering is no simple spring works. Rather, it is a carefully planned and constructed permeable stone wall set into the steep hillside. Based on field measurements, the linear stone wall is approximately 14.6 m long and up to 1.4 m high. At the foot of the wall is a rectangular collection trench with a cross-section approximately 0.8 m wide and about 0.6 m high. The spring works were accessed for operation and repairs via a terrace approximately 1.5 m to 2 m wide, supported on the steep hillside by a stone wall along the entire length of the collection area. A cross-section of the primary spring water collection system is illustrated in Figure 9.3. Water from a secondary spring surfaces 40 m above the ancient Incan water supply canal and 80 m west of the primary spring area. This water

Figure 9.3.  The ancient Incan spring headworks of Machu Picchu has endured for five centuries in the thick forest.

198-220 WaterUrban Pt2 Ch09 Wright.indd 204

12/05/2014 11:23

Machu Picchu: Water Engineering in the Mountains205

flows from the point of emergence to the Inca canal on the ancient canal terrace. Supplementary evidence of the Inca water collection system on the mountainside was searched for, and several field explorations into the dense forest area east of the primary spring were conducted. These explor­ ations revealed no evidence that the canal extended farther east than the primary spring, even though the surface topography shows the existence of a significant gully beyond. As a result, it was concluded that the water collection area works did not extend eastward from the primary spring. The Incan engineers evidently elected to refrain from extending the system to the next defined surface drainage, despite the fact it would have been physically viable to do so. About 10 m east of the primary spring, a wall of rounded cobbles was noted that may have afforded some protection against landslides or may have been a religious offering for the mountainside water source. The quality of water from the Inca spring is remarkably good (see Table 9.1). Based upon field measurements, conductivity typically ranged from 25 to 35 micro-Siemans per centimeter, pH ranged from 6.45 to 7.3, and temperature ranged from 14°C to 16°C, as compared to the annual average air temperature of 15.6ºC. The original Inca spring was used for many years to provide water for the travelers who visited Machu Picchu each year. However, when the spring proved to be inadequate for the rising number of visitors, the government constructed a pipeline to supply water from another source located several kilometers to the south. Measurements of the primary spring yield and water quality were taken at various times, and compared to the average monthly precipitation at Machu Picchu. The winter months of May through August are the dry season, while the summer months of November through March comprise the wet season. The spring yield is variable, ranging from a measured low of 23 liters per minute to 125 liters per minute. There is notable variation in the flow from season to season. The disparity implies that the spring flow is obtained from a relatively local hydrogeologic source that is affected by seasonal variation in precipitation. This correlates with the inflow–outflow evaluation and field surveys. Canal The ancient Machu Picchu water supply canal illustrates the Incan capacity for continuing a suitable grade for great distances, in addition to their faculty to build for the ages—even on steep, unstable slopes. Water was transported 749 m from the spring water source to the city center by means of a small local water canal constructed with cut stones, as shown in Figure 9.4.

198-220 WaterUrban Pt2 Ch09 Wright.indd 205

12/05/2014 11:23

206A History of Water

Table 9.1.  Machu Picchu domestic spring water quality Units

October 1994

February 1995

January 1996

Inorganics Total dissolved solids Total alkalinity Total Kjeldahl nitrogen Ammonia-N Chloride Sulfur

mg/L mg/L mg/L mg/L mg/L mg/L

40.00a 11.10a