The Socio-economic History and Material Culture of the Roman and Byzantine Near East: Essays in Honor of S. Thomas Parker 9781463237738

Citation preview

The Socio-economic History and Material Culture of the Roman and Byzantine Near East

Gorgias Studies in Classical and Late Antiquity

22

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

The Socio-economic History and Material Culture of the Roman and Byzantine Near East

Essays in Honor of S. Thomas Parker

Edited by

Walter D. Ward

gp 2017

Gorgias Press LLC, 954 River Road, Piscataway, NJ, 08854, USA www.gorgiaspress.com Copyright © 2017 by Gorgias Press LLC

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

1

2017

ISBN 978-1-4632-0701-4

A Cataloging-in-Publication record is available from the Library of Congress. Printed in the United States of America

TABLE OF CONTENTS Table of Contents ..................................................................................... v Introduction ............................................................................................ vii WALTER D. WARD S. Thomas Parker: Education, Archaeological Field Experience and Publications............................................................................ xxi Abbreviations .......................................................................................... xli Archaeological Survey Neolithic Period (ca. 9,700–4,900 BCE) Presence in the Southern Transjordan Plateau, the Southern Ghors, Wadis Fidan and Faynan, and the East Side of Wadi ʿArabah ............ 3 BURTON MACDONALD A Revaluation of Iron Age Fortified Sites on the Karak Plateau ... 35 STEPHANIE H. BROWN The roads leading to Zoara in the Roman and Byzantine Periods ............................................................................................. 63 CHAIM BEN-DAVID Material and Written Culture An Assessment and Re-examination of the American Expedition in Petra Excavation in the Residential Area (Area I), 1974–1977: The Early House and Related Ceramic Assemblages ................................................................... 91 TALI ERICKSON-GINI & CHRISTOPHER A. TUTTLE Exploring the Egyptian “Frog” Lamps Unearthed at the Red Sea Port of Roman Aila (Aqaba) and the Roman Fortress of Legio IV Martia at el-Lejjūn, Jordan .....................................151 ERIC C. LAPP From Opaque to Transparent: Colorless Glass and the Byzantine Revolution in Interior Lighting ..............................165 JANET DUNCAN JONES v

vi

ESSAYS IN HONOR OF S. THOMAS PARKER

A Nabataean Inscription from Bir Madkhur With an Excursus on Basileophoric Names and the Nabataean Dynastic Cult ................................................................................................185 DAVID F. GRAF & ANDREW SMITH II Military Following the Roman Paymaster .......................................................225 JOHN W. BETLYON The Modular Planning of Roman Fortifications in the Near East: Principles and Process ......................................................237 JOHN PETER OLESON The Military Presence in Petra during the Roman-Byzantine Period ............................................................................................273 ZBIGNIEW T. FIEMA Economy The Economy of Caesarea Palaestinae in Late Antiquity: Structure and Scale ......................................................................291 KENNETH G. HOLUM “So dearly do we pay for our luxury and our women”: Women and the Margins of the Roman World .....................................327 ELIZABETH ANN POLLARD Human Population Increase and its Effects on the Arid Landscape of Southern Jordan: An Archaeobotanical Study ..............................................................................................349 JENNIFER RAMSAY Index .......................................................................................................385

INTRODUCTION WALTER D. WARD

I received my first Dr. S. Thomas Parker “I know what’s best for you” speech sometime during my first time on an archaeological dig in the summer of 1998 in Aqaba, Jordan. It was during this season, the third of five expeditions that Tom (as his friends, colleagues, and students call him) directed in Aqaba, that I decided to pursue a career in the history and archaeology of the Near East. Tom’s speech outlined the path to become a tenure-track professor in history – starting languages as soon as possible, working with him on an M.A. thesis, attending a highly ranked Ph.D. program, writing a “marketable” dissertation, and then beginning a job hunt. He urged me to pursue history rather than archaeology for a simple reason – there were more jobs available to ancient historians than to classical archaeologists. In this one speech (and several more over my career at NC State), he was able to outline my entire career. For his encouragement, advice, and assistance in pursuing my dream career, I will forever be grateful. This volume – with contributions written by the aforementioned overlapping categories of friends, colleagues, and students – is our way of saying thanks and attempting to repay Tom for everything he has done for us. Dr. S. Thomas Parker began his studies at Trinity University in San Antonio Texas, where he obtained a B.A. in 1972. While a student there, he read the now famous article by Glenn Bowersock, “A Report on Arabia Provincia,” which was published in 1971 in the Journal of Roman Studies. I remember as a student when Tom recounted seeing for the first time Bowersock’s republication of an old aerial photograph of the Lejjūn fortress. Looking at the picture today, one can clearly see the potential for archaeology at the site. The legionary fortress looks beautifully ruined, with still standing, meter-high walls and the clear outlines of the curtain wall, gates, vii

viii

ESSAYS IN HONOR OF S. THOMAS PARKER

internal buildings, and projecting towers. Bowersock hoped that his article would spur renewed interest in the region, which it surely did. Tom often noted in later years that Bowersock himself would become a big supporter of his work. In 1971, Tom also obtained his first archaeological experience at Tell el-Hesi in Israel. Over the the next several years, he worked at digs at Ashkelon, Idalion in Cyprus, Tell Hesban, Umm el-Jimal, and Petra in Jordan. While at Idalion in 1974, his team had to be evacuated to a British naval base and then onto an American navy ship after Turkey invaded the island. In 1972, he began a Ph.D. at the University of California at Los Angeles. As with all students, he took courses in all fields of ancient history. His first article, still cited today, was published in 1975 in the Journal of Biblical Literature, entitled, “The Decapolis Reviewed,” which effectively demolished the idea that the Decapolis was a political league or confederation. Instead, Tom argued that these cities were united only by their Hellenized culture, and as such stood out amongst the various gentes of the region. The term Decapolis was convenient shorthand for these cities. The subject of this article about the Near East forshadowed his later career, however, his article published in the next year, “The Objectives and Strategy of Cimon’s Expedition to Cyprus” in the American Journal of Philology, demonstrates his successful foray into fields other than the Roman Near East. In 1975 Tom began working on his dissertation and his first field project, the survey of the Limes Arabicus, which would ultimately be published as Romans and Saracens: A History of the Arabian Frontier (1986). By the mid-1970s, no systematic survey of the frontier of the province of Roman Arabia had ever been attempted, though previous scholars had visited individual sites. Even though the sites themselves were often clearly visible from the air and ground (like Lejjūn), most of them were undated because only a few had intact dedicatory inscriptions. This meant that some other dating method would be required to reconstruct how the frontier defenses changed over time. Ceramics would prove to be the best route for dating these sites, but until 1973, there were few studies of pottery in the region based on scientific, stratified excavations. In that year, however, James Sauer used coin-controlled contexts from Tell Hesban in Jordan to create a ceramic typology for the region. In 1975, Tom used Sauer’s typology with surface pottery

INTRODUCTION

ix

collected from the legionary fortresses at Lejjūn and Udhruh near Petra. And then in 1976 he worked at Tell Hesban with Sauer, further improving his knowledge of ceramics. By August 1976, Tom felt comfortable enough with the ceramics to begin systematic survey and collection of pottery from the major military sites in the region. Tom Parker’s 1976 survey was designed to survey the entire length of the frontier in Jordan and to examine most sites that were larger than a watchtower. In total, 41 sites were examined, including the two legionary fortresses, 23 castella, a handful of watchtowers, two caravanserai and three non-military sites. By collecting hundreds of sherds from each location, Parker and Sauer (who did in-field readings of sherds) hoped to date each site’s periods of occupation and abandonment. Although surface readings are sometimes notorious for underrepresentation of occupation, subsequent excavation of several of these sites demonstrated the basic validity of dating sites from surface ceramics. When Romans and Saracens was published in 1986, Tom had by then already conducted excavations at several of these sites himself, including Lejjūn. This first book was a substantial contribution to our knowledge of the development of the Roman frontier in the province of Arabia. This survey work impacted Tom’s interpretation of the frontier and its enemies. In debates among fellow scholars, such as E. B. Banning, David Graf, Benjamin Isaac, and Philip Mayerson, Tom forcefully argued that the frontier military structures were positioned on the edge of agricultural production in order to supervise the migration of the nomads from the nearby desert into the sedentary regions. He argued that when there was insufficient military security on the frontier then nomadic populations represented “low-intensity” security problems, such as engaging in raiding and cattle rustling. He argued against a largely cooperative model between nomadic and sedentary populations along the frontier (though not discounting the possibility of cooperation) and against the idea that the primary mission of the Roman military was one of occupation, there to control a restive sedentary population. Instead, he argued that the deployment of the bulk of Roman military along the desert fringe implied that the perceived security threat was external, against the nomads, and not internal. In 1980 Tom launched his first major excavation project, the Limes Arabicus Project, which focused on the Roman frontier east

x

ESSAYS IN HONOR OF S. THOMAS PARKER

of the Dead Sea. The main goal was to understand a major military buildup here around the turn of the fourth century and the apparent abandonment of most of the fortifications about two centuries later. This included large-scale excavation of the legionary fortress at Lejjūn and several smaller but contemporary nearby military sites (most importantly Daʿjaniya) from 1980 to 1989. Within the Lejjun fortress itself, Parker’s team excavated the principia or headquarters building, several barracks, a bath, a church, horreum, portions of the fortifications, and three structures in the extramural vicus. The excavations confirmed the results of the earlier surface survey, i.e., that the fortress was founded under the Tetrarchy ca. 300, likely for legio IV Martia. The fort was damaged in the 363 earthquake and later rebuilt to accommodate a reduced garrison before being abandoned in the mid-sixth century. A more intensive regional survey was also conducted of both the frontier zone and well as portions of the adjacent desert to learn about the nomads. The survey recorded over 500 sites dating from the Paleolithic to the Late Islamic period. Over thirty publications about this project were published, including two major reports, the first, The Roman Frontier in Central Jordan: Interim Report on the Limes Arabicus Project, 1980–1985, appeared in 1987, and the second in 2006, The Roman Frontier in Central Jordan: Final Report on the Limes Arabicus Project, 1980–1989. It is important to note that the material objects including ceramics and glass, as well as organic remains, such as faunal and botanical evidence, were extensively published. Philip Freeman concluded his review of the later volume with the following statement: “[t]he first to go and explore the frontier, he [Tom Parker] is the first to produce a comprehensive synthesis of it. And it must be read” (Freeman 2007, 591). While directing the project at Lejjūn, Tom found time to work with Bert DeVries at Umm el-Jimal between 1977 and 1984, where Parker helped excavate a castellum and served as pottery specialist. In 1979–1980, he was a post-doctoral fellow at the prestigious Dumbarton Oaks Center for Byzantine Studies in Washington D.C. Dumbarton Oaks would later publish his final report of the Limes Arabicus project. While at Dumbarton Oaks he also taught at The American University before accepting a professorship at North Carolina State in 1980, where he has taught since. After completing excavation at Lejjūn, Tom only took four years off from archaeological work. In 1993, he began surveying

INTRODUCTION

xi

the modern city of Aqaba, Jordan for remains of the Nabataean/Roman city of Aila. He was encouraged by Dr. Donald Whitcomb who had previously excavated the Islamic misr of Ayla from 1986–1993. Although there was a legionary base at Aila (which was never discovered), this project (the Roman Aqaba Project) was actually intended to produce new evidence relevant to the ongoing debate about the nature of the imperial Roman economy. By reconstructing the economy of an urban site known to be one of the international ports of the Roman Empire, Tom wanted to address questions about trade between the Mediterranean, the Arabian peninsula, and the Indian Ocean. He also wished to engage in larger debates between so-called “modernists” and “primitivists,” the latter for example presented by Finley in The Ancient Economy. Excavation at Aqaba revealed an extensive stretch of the Byzantine era city wall, two cemeteries, and several Nabataean, Roman, Byzantine, and early Islamic domestic complexes. The most important and perhaps controversial discovery was that of a putative Christian church constructed about the turn of the fourth century and destroyed in the 363 earthquake, which is yet to be fully published. Concurrent with excavation at Aila, the Roman Aqaba Project included a regional survey of Aila’s hinterland on the Jordanian side of Wadi Arabah. The survey was published in 2014 as The Roman Aqaba Project Final Report. Volume I: The Regional Environment and the Regional Survey. Other volumes will focus on the excavations and material culture. This research project has also published about forty articles, most authored or co-authored by Parker himself. Fieldwork at ancient Aila ended in 2002. For the next decade, Parker worked on an intensive study of the ceramics from the site, with a special emphasis on imported amphorae and fine wares, and edited the final publication volumes. He also began a work of synthesis, Rome in the Middle East: Imperial Security Policy from Pompey to Heraclius, 64 B.C.–A.D. 630, which I believe will become the standard scholarly study on the subject. This book will be a capstone of Tom’s previous work, which includes over sixty articles or book chapters on the Roman frontier in the Near East. After a ten year hiatus from fieldwork, Tom returned to the field in 2012. After examining a legionary base and a Red Sea port, he has turned his attention to Petra, where he gained some of his earliest field experience some 27 years before. Though many archaeologists have worked at the spectacular site of Petra, they have

xii

ESSAYS IN HONOR OF S. THOMAS PARKER

focused almost exclusively on its extraordinary monumental structures. In contrast, Tom and his co-director, Megan Perry, have turned their attention to Petra’s non-elite population by excavating simple shaft tombs and several domestic complexes on the North Ridge. The recent discovery of two exquisite marble statues of Aphrodite and other evidence recovered in 2016 suggests that part of the site may have been an upper class dwelling, perhaps a villa urbana. Throughout his career, Tom has made the mentorship of students a priority. I am just one of several students – undergraduate and masters – who have gone on to successful careers in academia. According to his c.v., Tom has chaired the committees of 19 students who have completed their M.A. degrees at NC State. In addition to these students are the hundreds who have worked on his archaeological projects and who he has supported. Many of these students, like me, were given extensive training in analyzing the material culture, especially ceramics, which archaeology discovers on a daily basis. In addition, Tom has been very active in promoting archaeology and conservation in Jordan. He has been a Trustee of the American Center of Oriental Research in Amman, Jordan since 1987, and now serves as Second Vice-President of the Board and Chair of ACOR’s Fellowship Committee. He helped to establish endowments for ACOR to fund student fellowships, to memorialize departed friends and colleagues, such Jennifer Groot and James Sauer. Finally, I would be remiss in not mentioning Tom’s long marriage to Mary and his daughter Grace, both of whom sometimes accompanied him to Aqaba (Grace was only a toddler then).

ARTICLES

The essays collected in this volume have been written by former students and colleagues whom Tom has influenced. These contributors include a wide spectrum of influence; one of Tom’s first undergraduate students, Elizabeth Pollard, is included, as well as one of his most recent M.A. advisees, Stephanie Brown. The essays in this volume are divided into four topics, all of which have been important research areas for Tom’s work: survey, material and written culture, the economy, and the Roman military in the Near East.

INTRODUCTION

ARCHAEOLOGICAL SURVEY

xiii

As Tom Parker’s first original research project concerned a survey of the military frontier of Roman Jordan, it makes sense to begin the volume with three papers which focus on survey work. Burton MacDonald’s contribution, “Neolithic Period (ca. 9,700–4,900 BCE) Presence in the Southern Transjordan Plateau, the Southern Ghors, Wadis Fidan and Faynan, and the East Side of Wadi Arabah,” begins the collected essays both in terms of sequence but also in chronological order. MacDonald writes after conducting five archaeological surveys in the region starting in 1979. His paper argues that there was a large Neolithic presence in the survey region, which lies south of the Wadi al-Hasa and north of Ras anNaqab. He finds a large concentration of sites from the Pre-Pottery Neolithic B period in all areas, except those surveyed in the TafilaBusayra area. Furthermore, he argues that by the Pre-Pottery Neolithic copper ore from Wadi Faynan was being traded to the Karak plateau for use as cosmetics. Stephanie Brown, “A Revaluation of Iron Age Fortified Sites on the Karak Plateau,” returns to the work done during the survey around the fortress at Lejjūn with a concentration on ten Iron Age sites. Most of these sites have been resurveyed in the period since the Lejjūn work, and so Brown is dealing with a wider corpus of materials than previously available. Brown argues that these sites, which in the past were associated with a unified policy of frontier defense organized by the Assyrian government, were in fact constructed at different times for different reasons. Brown argues that five of the Iron Age II sites consisted of 30 m x 30 m fortlets which were constructed by family groups for defense against a nomadic presence mentioned in historical records for Moab of this time period, and only a few of the sites may have been official settlements. The last of the survey articles is Chaim Ben-David’s “The Roads Leading to Zoara in the Roman and Byzantine Periods.” Ben-David’s explorations around Zoara (modern Ghor es-Safi) in Jordan at the tip of the Dead Sea discovered a previously unknown well-built Roman road leading to the site from the north. In this paper, Ben-David describes the three routes now known to approach Zoara from the north – the Kathrabba Ascent, the Zoar Ascent and the Kuniyeh Ascent. These roads, along with the Peutinger Map, demonstrate that Zoara was an important transporta-

xiv

ESSAYS IN HONOR OF S. THOMAS PARKER

tion hub, linking the Negev, the Karak Plateau, the Wadi Arabah, and surrounding regions together.

MATERIAL AND WRITTEN CULTURE

One of Parker’s lasting impacts is the number of students that he has trained in ceramics over his years at NC State and in the field. This may end up being one of his most important contributions to archaeology in the region, as this generation now has a number of qualified ceramic specialists, many of whom have worked at several different sites throughout southern Jordan. In addition to ceramics, Tom’s interim and final reports have always included extensive chapters dedicated to material finds, such as glass, bone, and objects (like lamps). In addition, Tom was responsible for publishing inscriptions from Roman milestones and military installations. The four papers in this section reflect the importance of these pieces of evidence in Tom’s work. The most controversial paper to appear in this volume is the paper, “An Assessment and Re-examination of the American Expedition in Petra Excavation in the Residential Area (Area I), 1974– 1977: The Early House and Related Ceramic Assemblages” by Tali Erickson-Gini and Christopher Tuttle. In 2010, ChrisTuttle began a re-evaluation of the excavation archive of the Temple of the Winged Lions which was excavated from 1973–2005, but never fully published due to the death of the primary excavators, Kenneth Russell (in 2002) and Philip Hammond (in 2008). The material published in this paper concerns Area I, the less-famous residential area near the later discovered temple. This area was of prime importance, because the material unearthed (but not published) from Area I profoundly influenced the dating sequence of the Temple of the Winged Lions (Area II). In this paper, Erickson-Gini and Tuttle argue for a revision of the dating of the Early House in Area I and the corresponding ceramic assemblages previously dated before the Roman annexation of Arabia in 106 CE to the late 2nd and early 3rd c. CE when the structure was abandoned. They argue that other areas of the excavations, such as the Painters’ Workshop, corroborates these revised dates. The impact of these revisions, if accepted by scholars, would mean that structures throughout Petra and Provincia Arabia will also have to be re-dated. This would indicate that building, renovation, and settlement density was much higher in the post-annexation period than previously believed, put-

INTRODUCTION

xv

ting Petra and the region on similar chronological and economic terms as other major cities in the Near East, such as Jerash, Apamaea, and Palymra, which experienced building booms in the second century CE. Continuing the theme of material objects is Eric Lapp’s paper, “Exploring the Egyptian ‘Frog’ Lamps Unearthed at the Red Sea Port of Roman Aila (Aqaba) and the Roman Fortress of Legio IV Martia at el-Lejjūn, Jordan. The “Frog” lamps, named because of the use of the frog as decoration, are rarely found outside of Egypt, but three were discovered at Roman Aila and one from Lejjūn. Lapp describes how the symbol of the frog was associated with rebirth and regeneration in Dynastic Egypt, and he argues that Christians adopted this symbolism. This is proven by frog lamps found with Christian inscriptions, but none of the ones discovered at Aila or Lejjūn contain these Christian embellishments. A petrographic analysis demonstrated that they were imported from Egypt. Janet Jones’s paper, “From Opaque to Transparent: Colorless Glass and the Byzantine Revolution in Interior Lighting,” begins with an introduction to the production and cultural importance of glass for the non-specialist. She focuses on five moments in the history of glass production: natural glass, the discovery of how to manufacture glass, the development of vessels made of transparent glass, glassblowing, and finally, lighting the interior of structures with clear glass. In the final section, she describes the use of glass in Byzantine era churches and then turns to a discussion of glass objects at Roman Aila. She notes that there was a large increase in the finds of interior glass from the third century CE in domestic contexts there, suggesting that the population was becoming more influenced by Greco-Roman cultural ideas. She notes that a large quantity of glass was recovered from the putative church and compares the glass corpus there with glass from churches of the sixth century at nearby sites like Jerash. The final article in this section, “A Nabataean Inscription from Bir Madkhur With an Excursus on Basileophoric Names and the Nabataean Dynastic Cult,” by David Graf and Andrew Smith II publishes for the first time a Nabataean inscription discovered in 2014 at the site of Bir Madkhur, located on the route from Petra to the Negev. The inscription is quite short, consisting of just the end of a name and part of a patronym, dating likely to the third century CE. The patronym is restored to read ‛BD-‛BDT, making it a Naba-

xvi

ESSAYS IN HONOR OF S. THOMAS PARKER

taean basileophoric name. The paper continues with a discussion of the use of basileophoric names in the Nabataean realm and the Nabataean ruler cult. Graf and Smith argue, against some previous interpretations, that there is ample evidence of the incorporation of a Hellenistic-style ruler cult in the Nabataean realm.

MILITARY

From Parker’s original survey of the frontier in Jordan to his forthcoming book on the history the Roman frontier in the Near East, the Roman military has been one of the most important features of his research. The three papers in this section honor Tom’s interest in this subject. “Following the Roman Paymaster” by John Betlyon investigates how paying the military in precious metal coins impacted the local economy. Sixty percent of the coins along the military frontier originated at the imperial mints in Alexandria and Antioch. He argues that the majority of the coins found at Aila were “Early Byzantine,” which is striking compared to Berenike on the Egyptian Red Sea coast, where the majority of coins are earlier. Betlyon focuses on four types of coins – “SOLI INVICTO,” “GLORIA EXERCITVS,” “PCONST,” and “FEL TEMP REPARATIO” – as exemplars of military pay. Half of the coins found in the EzZantur villa in Petra were of the “FEL TEMP REPARATIO” type and must have originated with the Roman military presence nearby, likely from Udhruh. John Oleson’s paper, “The Modular Planning of Roman Fortifications in the Near East: Principles and Process,” argues that military sites in the Near East were planned and constructed based on ratios of the Roman foot of 0.296 m (pes monetalis, plural pedes monetales, abbreviated as pm). Oleson begins the paper by discussing his own research at the Trajanic auxiliary fort at Humayma. The rectangular plan of the fort measured 500 x 700 pm and the nearby reservoir measured 100 x 50 x 10 pm. Buildings within the fort also seem to conform to a modular plan based on measurements of pm. In this paper, Oleson tests his hypothesis about modular building with dozens of Roman-constructed sites in the Near East and found that the proportion of 1:1 is most common while that of 3:2 is second most common. Less common proportions were 4:3, 5:4, 9:8, and 16:11. He concludes that authors Polybius and Pseudo-

INTRODUCTION

xvii

Hyginus are generally reliable about Roman camp design when compared to actual, built structures. The last paper in this section, Zbigniew Fiema’s “The Military Presence in Petra during the Roman-Byzantine Period,” seeks to understand the Roman garrison of Petra. Fiema emphasizes the role of the military in providing security and enforcing peace within populations centers, a contrast to Tom’s emphasis on the military as guardians of the frontier and rural zones. Fiema begins by comparing Petra to Bostra, where the III Cyrenaica legion was based. He notes that the competition between these two cities for imperial honorifics was intense, but that Petra’s status remained substantial higher, owing to its use of the honorific metropolis. He then explores the relatively poor information about military units in Petra in the early Roman period, concluding that the only evidence of units stationed in Petra before the fourth century were detachments from the III Cyrenaica. He concludes the paper by looking at unrest – owing to Christian-pagan conflict and the changing status of Petra, first as locals began to take control of security, and then as Petra became a place for the banishment of ecclesiastics and common criminals.

ECONOMY

As discussed above, after excavating at Lejjūn, Parker turned his attention to the Red Sea port at Aila (modern Aqaba, Jordan), where he was particularly interested in understanding the volume and nature of trade flowing through the port. I have already discussed that Parker opposes “primitivist” views of the economy in favor of more expansive trade throughout the empire. Kenneth Holum’s piece, “The Economy of Caesarea Palaestinae in Late Antiquity: Structure and Scale,” presents a different view of the economy by focusing on one of the largest and most important cities in the Palaestina region. Holum argues that the economy of Caesarea can be better explained using a Finley model of the economy than a more trade-intensive model. He argues that the elites of the city were much more interested in wealth in order to increase their own prestige and to enhance their status rather than in creating additional wealth, as in more modern economies. In order to make this argument, Holum discusses land tenure and attitudes towards land owning by the elites in sixth century Caesarea. He argues that the elites poured the resources of the surrounding territory into sus-

xviii

ESSAYS IN HONOR OF S. THOMAS PARKER

taining their urban spaces, demonstrating the feasibility of the Finley model of the ancient city as a consumer, rather than producer. Holum finishes by noting that not all ancient cities were similar and that other cities, such as Aila and Palmyra, may not conform to the Finley model. Elizabeth Ann Pollard’s paper, “’So dearly do we pay for our luxury and our women’: Women and the Margins of the Roman World,” continues with the theme of the economy in the Near East by examining the attitudes of Roman authors concerning trade with the east and products crossing these frontier as associated with uncontrollable women (be they wealthy, witchy, or whore-like) and womanly vice. Pollard’s article begins with the passage of laws in the late Roman Republic against women’s displays of wealth. Next she examines images of the Amazons who were thought to dwell along the Eurasian trade routes, witches such as Meroe and Medea who became associated with eastern trade, and the description of Rome as the “Whore of Babylon,” wanton with precious metals and spices. She argues that these images contributed to the feminizing of long-distance trade and its products. Women’s societal dislocation, their innate nature, and their physical constitution, as articulated in Roman texts, rendered these women and eastern trade as a constant source of potential pollution – and a threat to fluid boundaries that needed protection and control. The final paper in the volume is Jennifer Ramsay’s “Human Population Increase and its Effects on the Arid Landscape of Southern Jordan: An Archaeobotanical Study.” In this paper, Ramsay uses botanical evidence discovered at several sites in southern Jordan – Petra, Aila, Humayma Bir Madhkur, and ‘Ayn Gharandal – to argue that there is ample evidence of local agriculture in the form of cereal grains, chaff and associated weed species and agricultural structures such as field walls, terraces, water channels and check dams. This suggests that the region would have been more agriculturally productive than it appears to be today, and that increasing populations in the early centuries CE were able to produce some of their necessities, especially in the form of wheat and barley.

ACKNOWLEDGMENTS

This volume would not exist with the important mentorship that I received from Tom Parker, and I know that everyone who contrib-

INTRODUCTION

xix

uted to this volume has a lot of gratitude for everything that he has done for them and his inspiration over the years. Each essay includes a special note of thanks to Tom. Dr. Benjamin Dolinka, one of Tom’s former M.A. students, approached me in January 2013 with the idea for this volume. Though Ben was unable to participate in its completion, his ideas and inspiration have been a guiding force in the development of this work. Ben was very helpful in soliciting papers and developing editorial guidelines. Ben also collected the archival photos of Tom. I thank him profusely for coming up with the idea of producing an edited volume in order to honor Tom. I would also like to thank one of my former M.A. students, Timothy Granger, for helping to edit the bibliographies. Also thanks to the dozens of friends and colleagues who gave their advice and support during the development of this project. Over the years, my research has benefited greatly from my experiences at the American Center of Oriental Research in Amman. My short time there in 2014 was instrumental in rounding up the final set of papers. Thanks to all there, especially Barbara Porter and Glenn Corbett. Sarah Harpending at ACOR was kind to find and scan pictures of Tom Parker from the archive. Thanks also to the people at Gorgias Press who were willing to take on this edited volume at its most primordial stage of development. I greatly appreciate the support I have received from my institution, the University of Alabama at Birmingham, especially the Department of History, currently chaired by Colin Davis. Finally, I give the deepest thanks to my family, especially to my wife, Melissa, who has heard me say that I needed to do “just one more thing” before sending this volume to the press more than I can count. I completed the editing of these papers thanks to a generous semester sabbatical. And for my kids – Agatha, Nico, and Ansel – I cannot imagine life without you guys.

A NOTE ABOUT PLACE NAMES

Transliterating the names of places in the Middle East into English is notoriously complicated. It is even tougher when some of the scholars, as in this volume, primarily work in a Hebrew speaking country, while others work in an Arabic speaking one. My general policy has been to allow both Hebrew and Arabic transliterations (so one might see Wadi Arava and Wadi Arabah in the articles in

xx

ESSAYS IN HONOR OF S. THOMAS PARKER

this volume) when the names are relatively similar. If they were not similar, I have adopted the name which I judged to be most common. I have also changed ancient versions of names into their more common counterparts (if they exist – for example, Antioch) when possible in the articles (but not in quotations). This means that some place names will be spelled differently in articles, quotations, and bibliographies, but this seemed the best option, even if it might cause some confusion. Final note: As this volume was going to the printer, Kenneth Holum passed away. He will be missed by many students, colleagues, and friends. His mentorship and scholarship will not be forgotten.

S. THOMAS PARKER: EDUCATION, ARCHAEOLOGICAL FIELD EXPERIENCE AND PUBLICATIONS EDUCATION 1979 1976 1974 1972

Ph.D., University of California, Los Angeles. Dissertation: The Historical Development of the Limes Arabicus (Ann Arbor, MI: University Microfilms, 1979). C.Phil., University of California, Los Angeles M.A., University of California, Los Angeles B.A., Trinity University, San Antonio

ARCHAEOLOGICAL FIELD EXPERIENCE 2016 2014 2012 2002 2000 1998 1997 1996 1994

Co-director and Pottery Specialist, Petra North Ridge Project, Jordan. Co-director and Pottery Specialist, Petra North Ridge Project, Jordan. Co-director and Pottery Specialist, Petra North Ridge Project, Jordan. Director, Stratigrapher, and Pottery Specialist, Roman Aqaba Project, Jordan. Director, Stratigrapher, and Pottery Specialist, Roman Aqaba Project, Jordan. Director, Stratigrapher, and Pottery Specialist, Roman Aqaba Project, Jordan. Director, Stratigrapher, and Pottery Specialist, Roman Aqaba Project, Jordan. Director, Stratigrapher, and Pottery Specialist, Roman Aqaba Project, Jordan. Director, Stratigrapher, and Pottery Specialist, Roman Aqaba Project, Jordan. xxi

xxii 1993 1989 1987 1985 1984 1982 1982 1982 1981 1980 1979 1977 1976 1976 1975 1975 1974 1973 1973 1971

ESSAYS IN HONOR OF S. THOMAS PARKER Director of preliminary reconnaissance, Roman Aqaba Project, Jordan. Director, Stratigrapher, and Pottery Specialist, Limes Arabicus Project, Jordan. Director, Stratigrapher, and Pottery Specialist, Limes Arabicus Project, Jordan. Director, Stratigrapher, and Pottery Specialist, Limes Arabicus Project, Jordan. Stratigrapher and Pottery Specialist, Umm el-Jimal Project, Jordan. Reconnaissance of Roman frontier sites in Turkey and Syria. Director, Stratigrapher, and Pottery Specialist, Limes Arabicus Project, Jordan. Consultant to Department of Antiquities- excavation of Roman tomb in Amman, Jordan. Stratigrapher and Pottery Specialist, Umm el-Jimal Project, Jordan. Director, Stratigrapher, and Pottery Specialist, Limes Arabicus Project, Jordan. Director of preliminary reconnaissance, Limes Arabicus Project, Jordan. Stratigrapher, Area Supervisor, and Pottery Specialist, Umm el-Jimal Project, Jordan. Director, Survey of the Limes Arabicus, Jordan. Area Supervisor at Tell Hesban, Jordan. Area Supervisor at Petra, Jordan. Square Supervisor at Tell el-Hesi, Israel. Square Supervisor at Idalion, Cyprus. Square Supervisor at Tell el-Hesi, Israel. Square Supervisor at Ashkelon, Israel. Student volunteer at Tell el-Hesi, Israel.

PUBLICATIONS Books (ed.) The Roman Aqaba Project Final Report. Volume I: The Regional Environment and the Regional Survey. ASOR Archaeological Reports 19. (Boston: American Schools of Oriental Research, 2014). Co-editor and author of the following chapters:

S. THOMAS PARKER

xxiii

“Introduction”, pp. 1–31. “The Pottery from the Survey”, pp. 311–348. “The Hinterland of Roman Aila” (with A. M. Smith), pp. 357– 377. (ed.) History of the Ancient Mediterranean World (Dubuque, IA: Kendall Hunt, 2009). 2nd ed. 2011. Revised 2nd ed. 2013. (ed.) The Roman Frontier in Central Jordan: Final Report on the Limes Arabicus Project, 1980–1989. 2 vols. (Washington: Dumbarton Oaks, 2006). Editor-in-chief and author or co-author of the following chapters: “Introduction”, pp. 5–22. “The Regional Survey” (with V. Clark and F. Koucky), pp. 25–107. “The Legionary Fortress of el-Lejjūn”, pp. 111–122. “The Principia of el-Lejjūn (Area A)” (with A. Lain), pp. 123– 159. “The Barracks at el-Lejjūn (Areas K, L, R and B.6)” (with J. C. Groot and J. E. Jones), pp. 161–185. “The East Vicus Building (Area P)” (with P. Crawford), pp. 247–257. “The Pottery”, pp. 329–371. “History of the Roman Frontier East of the Dead Sea”, pp. 517–574. Regional editor for the Levant in Richard J. A. Talbert, ed., Barrington Atlas of the Greek and Roman World. (Princeton: Princeton University, 2000). (ed.) The Roman Frontier in Central Jordan: Interim Report on the Limes Arabicus Project, 1980–1985. 2 vols. BAR International Series 340 (Oxford: British Archaeological Reports, 1987), x + 861 pp., 150 figs., 109 plates. Editor-in-chief and author or coauthor of the following chapters: “Introduction: Project Goals, Strategy, Personnel, Funding”, pp. 1–10. “The Late Roman Observation and Signaling System”, pp. 165–181. With V. Clark.

xxiv

ESSAYS IN HONOR OF S. THOMAS PARKER

“An Introduction to the Legionary Fortress of el-Lejjun”, pp. 183–197. “The Roman Castellum of Khirbet el-Fityan”, pp. 429–446. With S. Richard. “The Fortlet of Rujm Beni Yasser”, pp. 447–456. With J. Bloom. “The Pottery”, pp. 525–619. “History of the Roman Frontier East of the Dead Sea”, pp. 793–823. Romans and Saracens: A History of the Arabian Frontier. ASOR Dissertation Series No. 6 (Winona Lake, IN: American Schools of Oriental Research, 1986). xiii + 238 pp., 73 figs., XII plates. Articles 2016 “Petra North Ridge Project”, in G. J. Corbett et al., eds., “Archaeology in Jordan”, American Journal of Archaeology 120 (2016) 664–665 (co-author). “A Diachronic Look at the Agricultural Economy at the Red Sea Port of Aila: An Archaeobotanical Case for Hinterland Production in Arid Environments”, Bulletin of the American Schools of Oriental Research 376 (2016) 101–120 (co-author). “The Petra North Ridge Project: Domestic Structures and the City Wall”, pp. 587–595 in Studies in the History and Archaeology of Jordan XII. Amman: Department of Antiquities. 2015 “The Roman Army at Petra” Pp. 313–318 in Roman Frontier Studies 2012: Limes XXII: Proceedings of the 22nd International Congress of Roman Frontier Studies, Ruse, Bulgaria, September 2012. L. Vagalinski and N. Sharankov (eds.) Sofia: National Archaeological Institute, 2015. “Feeding the Late Roman Army on the Southern Arabian Frontier”, pp. 212–221 in Understanding Roman Frontiers: A Celebration for Professor Bill Hanson. David J. Breeze, Rebecca H. Jones, and Ioana A. Oltean (eds.). Edinburgh: John Donald, 2015. 2014 “The Petra North Ridge Project”, in G. J. Corbett et al., eds., “Ar-

S. THOMAS PARKER

xxv

chaeology in Jordan”, American Journal of Archaeology 118 (2014) 666–667. “Coarse Ware Pottery of the First through Third Centuries at Roman Aila (Aqaba, Jordan)”, pp. 205–215 in Roman Pottery in the Near East: Local Production and Regional Trade. Proceedings of the Round Table held in Berlin, 19–20 February, 2010. B. FischerGenz, Y. Gerber, and H. Hamel (eds.). Roman and Late Antique Mediterranean Pottery 3. Oxford: Archaeopress, 2014. Encyclopedia of the Roman Army (Oxford: Wiley Blackwell), the following articles: “Arabia” (pp. 46–49) “Bosra” (pp. 111–112) “Judaea” (pp. 559–561) “Lejjun” (pp. 583–585) “Masada” (pp. 637–640) “Palaestina/Syria Palaestina” (pp. 717–719) “Qasr Bshir” (pp. 797–799) “Strata Diocletiana” (pp. 929–933) “Syria” (pp. 970–976) 2013 “Petra North Ridge Project: The 2012 Season”, Annual of the Department of Antiquities of Jordan 57 (2013) 399–407 (co-author). “The Roman Port of Aila and its Economic Hinterland”, pp. 735– 743 in Studies in the History and Archaeology of Jordan XI: Changes and Challenges. Amman: Department of Antiquities of Jordan, 2013. 2012 “Response to Bert de Vries, ‘Site Preservation in Jordan: The Case of Umm el-Jimal’”, pp. 187–188 in Archaeology, Bible, Politics, and the Media: Proceedings of the Duke University Conference, April 23–24, 2009, ed. by Eric M. Meyers and Carol Meyers (Winona Lake, IN: Eisenbrauns, 2012). “ACOR’s Named Fellowship Program”, ACOR Newsletter 24.1 (2012) 4–5.

xxvi

ESSAYS IN HONOR OF S. THOMAS PARKER

2010 “Frank Koucky 1927 – January 29, 2010”. American Schools of Oriental Research Newsletter 60.4 (2010) 15 (with L. E. Stager and A. M. Walker). 2009 “The Roman Port of Aila: Economic Connections with the Red Sea Littoral”. Pp. 79–84 in L. Blue, J. Cooper, R. Thomas and J. Wainwright, eds. Connected Hinterlands: Proceedings of the Red Sea Project IV. Held at the University of Southampton September 2008. Society for Arabian Studies Monographs No. 8. British Archaeological Reports International Series 2052. Oxford: British Archaeological Reports, 2009. “The Roman Frontier in Southern Arabia: A Synthesis of Recent Research”. Pp. 142–152 in W. S. Hanson, ed., The Army and Frontiers of Rome: Papers offered to David Breeze on the occasion of his sixth-fifth birthday and his retirement from Historic Scotland. JRA Supplementary Series 74. Portsmouth, RI: Journal of Roman Archaeology, 2009. “Arabia Adquisita: The Roman Annexation of Arabia Reconsidered”. Pp. 1585–1592 in A. Morillo, N. Hanel, and E. Martín, eds., in Limes XX: Roman Frontier Studies. XXth International Congress of Roman Frontier Studies. 3 vols. Madrid: Consejo Superior de Investigaciones Científicas, 2009. “The Foundation of Aila: A Nabataean Port on the Red Sea”, Pp. 685–690 in Studies in the History and Archaeology of Jordan X: Crossing Jordan. Amman: Department of Antiquities of Jordan, 2009. 2008 “The First Archaeological Evidence for Haimation, the Invisible Garum”, Journal of Archaeological Science. 35 (2008) 1821–1827. With Wim Van Neer. 2007 “Harold Forshey Retires from ACOR Board”. ACOR Newsletter 19.2 (2007) 9. “Structural Damage from Earthquakes in the Second-Ninth Century at the Archaeological Site of Aila in Aqaba, Jordan”, Bulletin

S. THOMAS PARKER

xxvii

of the American Schools of Oriental Research 346 (2007) 59–77. With R. Thomas and T. M. Niemi. “Projecting Power on the Periphery: Rome’s Arabian Frontier East of the Dead Sea”. Pp. 349–357 in T. E. Levy, P.M. M. Daviau, and M. Shaer, eds. Crossing Jordan: North American Contributions to the Archaeology of Jordan. London and Oakville, CT: Equinox, 2007. “Beyond Frankincense and Myrrh: Reconstructing the Economy of Roman Aqaba”. Pp. 359–366 in T. E. Levy, P.M. M. Daviau, and M. Shaer, eds. Crossing Jordan: North American Contributions to the Archaeology of Jordan. London and Oakville, CT: Equinox, 2007. 2006 “Nancy Lapp Retires as Chair of the ACOR Fellowship Committee”, ACOR Newsletter 18.1 (2006) 8–9. “Roman Aila and Wadi Arabah: An Economic Relationship.” Pp. 227–234 in Crossing the Rift: Resources, Routes, Settlement Patterns, and Interaction in the Wadi Arabah. P. Bienkowski and K. Galor, eds. British Academy Monographs in Archaeology. Oxford: Oxford University Press, 2006. 2005 “Supplying the Roman Army on the Arabian Frontier”, pp. 415– 425 in Z. Visy, ed. Limes XIX: Proceedings of the XIXth International Congress of Roman Frontier Studies, Pécs. Hungary, September 2003. Pécs: University of Pécs, 2005. “Mary Louise Mussell”, American Schools of Oriental Research Newsletter 55.1 (2005) 6, 9. “Mary-Louise Mussell, 1959–2005”, ACOR Newsletter 16.2 (2005) 8. 2004 “The 363 A.D. earthquake and other ground-rupturing earthquakes along the Dead Sea Transform in Aqaba, Jordan.” Pp. 603– 605 in the 5th International Symposium on Eastern Mediterranean Geology, Thessaloniki, Greece, April 14–19, 2004. Proceedings v. 2. Thessaloniki: Aristotle University (co-author with T. M. Niemi and R. Thomas).

xxviii

ESSAYS IN HONOR OF S. THOMAS PARKER

2003 “The Roman ʿAqaba Project: The 2002 Campaign.” Annual of the Department of Antiquities of Jordan 47 (2003) 321–333. “The Roman Aqaba Project”, in S. H. Savage et al., eds., “Archaeology in Jordan”, American Journal of Archaeology 107 (2003) 473–475. 2002 “The Roman Frontier in Jordan: An Overview”, Pp. 77–84 in Limes XVIII. Proceedings of the XVIIIth International Congress of International Roman Frontiers Studies held in Amman, Jordan. (September 2000). P. Freeman, J. Bennett, Z.T. Fiema & B. Hoffman, eds. British Archaeological Reports 1084 (I) (Oxford: Archaeopress, 2002). “The Roman Aqaba Project”, ACOR Newsletter 14.1 (2002) 11–12. “The Roman ʿAqaba Project: The 2000 Campaign”, Annual of the Department of Antiquities of Jordan 46 (2002) 409–428. 2001 “Roman Aqaba Project”, in S. H. Savage et al., eds., “Archaeology in Jordan”, American Journal of Archaeology 105 (2001) 457–458. 2000 “The Roman ʿAqaba Project: The 1997 and 1998 Campaigns”, Annual of the Department of Antiquities of Jordan 44 (2000) 373–394. “Roman Legionary Fortresses in the East”, in Richard J. Brewer, ed. Roman Fortresses and their Legions: Papers in Honour of George C. Boon. Occasional Papers of the Society of Antiquaries of London 20 (London: Society of Antiquaries of London and National Museums and Galleries of Wales, 2000) 121–138. “The Defense of Palestine and Transjordan from Diocletian to Heraclius”. in L. E. Stager, J. A. Greene, and Michael D. Coogan, eds. The Archaeology of Jordan and Beyond: Essays in Honor of James A. Sauer. Studies in the Archaeology and History of the Levant 1 (Winona Lake, IN: Eisenbrauns, 2000) 367–388. “Roman Aqaba Project”, ACOR Newsletter 12.1 (2000) 4–5. 1999 “An Empire’s New Holy Land: The Byzantine Period”, Near Eastern Archaeology 62:2 (1999) 134–180.

S. THOMAS PARKER

xxix

“The Roman Frontier in Arabia in Light of New Research”, in N. Gudea, ed. Proceedings of the XVII International Congress of Roman Frontier Studies (Zalau: County Council of Salaj and County Museum of History and Art Zalau, 1999) 231–240. “Aila/Ayla”, in G.W. Bowersock, P. Brown, and O. Grabar, eds., Late Antiquity: A Guide to the Postclassical World (Cambridge: Harvard University Press, 1999) p. 283. With D. Whitcomb. “Brief Notice on a Possible Fourth Century Church at ʿAqaba, Jordan”, Journal of Roman Archaeology 12 (1999) 372–376. “Roman Aqaba Project”, in P. Bikai and V. Egan, eds., “Archaeology in Jordan”, American Journal of Archaeology 103 (1999) 511– 513. 1998 “The Roman Aqaba Project: The 1996 Campaign”, Annual of the Department of Antiquities of Jordan 42 (1998) 375–394. In Bert de Vries, ed., Umm el-Jimal: A Frontier Town and its Landscape in Northern Jordan. Volume I: Field Work, 1972–1981. Journal of Roman Archaeology Supplement 26 (Portsmouth, RI: Journal of Roman Archaeology, 1998), the following articles: “The Later Castellum (“Barracks”)”, pp. 131–142. “The Defenses of the Roman and Byzantine Town”, pp. 143– 148. “The Nabataean Temple”, pp. 149–152. “The Nabataean Temple: A Re-examination”, pp. 153–160, With L. de Veaux. “The Pottery (1977)”, pp. 205–218. “An Early Church, Perhaps the Oldest in the World, Found at Aqaba”, Near Eastern Archaeology 61.4 (1998) 254. 1997 From The Oxford Encyclopedia of Archaeology in the Near East (Oxford: Oxford University Press, 1997), the following articles: “Aqaba”, v. 1, pp. 153–156. With D. S. Whitcomb. “The Decapolis”, v. 2, pp. 127–130. “Fortifications: Hellenistic and Roman Periods”, v. 2, pp. 329–334. “Limes Arabicus”, v. 3: pp. 358–361.

xxx

ESSAYS IN HONOR OF S. THOMAS PARKER

“Transjordan in the Persian through Roman Periods”, v. 5, pp. 235–238. “The Roman Aqaba Project”, in P. Bikai and V. Egan, eds., “Archaeology in Jordan”, American Journal of Archaeology 101 (1997) 525–526. “Geography and Strategy on the Southeastern Frontier in the Late Roman Period”, Pp. 115–122 in Roman Frontier Studies 1995: Proceedings of the XVIth International Congress of Roman Frontier Studies. Oxbow Monographs 91 (Oxford: Oxbow, 1997). “Human Settlement at the Northern Head of the Gulf of Aqaba: Evidence of Site Migration”, Pp. 189–193 in G. Bisheh, M. Zaghloul, and I. Kehrberg, eds., Studies in the History and Archaeology of Jordan VI (Amman: Department of Antiquities of Jordan, 1997). “Preliminary Report on the 1994 Season of the Roman Aqaba Project”, Bulletin of the American Schools of Oriental Research 305 (1997) 19–44. 1996 “Roman Aqaba”, ACOR Newsletter 8.1 (1996) 7–8. “The Roman Aqaba Project: The Economy of Aila on the Red Sea”, Biblical Archaeologist 59.3 (1996) 182. “The Roman Aqaba Project: The 1994 Campaign”, Annual of the Department of Antiquities of Jordan 40 (1996) 231–257. “The Roman Aqaba Project”, Old World Archaeology Newsletter 19.2 (1996) 9–11. 1995 “The Typology of Roman and Byzantine Forts and Fortresses in Jordan”, in Studies in the History and Archaeology of Jordan V (Amman: Department of Antiquities, 1995) 251–260. “The Roman Aqaba Project”, in P. Bikai and D. Kooring, eds., “Archaeology in Jordan”, American Journal of Archaeology 99 (1995) 522–523, fig. 15. 1994 “The Roman Aqaba Project: Aila Rediscovered”, Biblical Archaeologist 57.3 (1994) 172. “The Roman Aqaba Project”, ACOR Newsletter 6.2 (1994) 8–9.

S. THOMAS PARKER

xxxi

“A Late Roman Soldier's Grave by the Dead Sea”, Annual of the Department of Antiquities of Jordan 38 (1994) 385–394. “El-Haditha”, in B. DeVries and P. Bikai, ed., “Archaeology in Jordan”, American Journal of Archaeology 98 (1994) 549, fig. 22. 1993 “El-Lejjun”, The New Encyclopedia of Archaeological Excavations in the Holy Land v. 3: pp. 913–915 (Jerusalem: Israel Exploration Society, 1993). 1992 “The Limes and Settlement Patterns in Central Jordan during the Roman and Byzantine Periods”, in Studies in the History and Archaeology of Jordan IV (Amman: Department of Antiquities of Jordan, 1992) 321–325. “Lejjun”, Anchor Bible Dictionary. v. IV: 276–277 (New York: Doubleday, 1992). 1991 “The Nature of Rome’s Arabian Frontier”, in V. A. Maxfield and M. J. Dobson, eds., Roman Frontier Studies 1989: Proceedings of the XVth International Congress of Roman Frontier Studies (Exeter: University of Exeter, 1991) 498–504. “Preliminary Report on the 1989 Season of the Limes Arabicus Project”, Bulletin of the American Schools of Oriental Research Supplement 27 (1991) 117–154. “Limes Arabicus”, in B. de Vries, ed., “Archaeology in Jordan”, American Journal of Archaeology 95 (1991) 274–275, figs. 21–22. 1990 “The Limes Arabicus Project: The 1989 Campaign”, Annual of the Department of Antiquities of Jordan 34 (1990) 357–376. “The Limes Arabicus Project: The 1989 Campaign”, Syria 67 (1990) 476–479. “Preliminary Report on the 1987 Season of the Limes Arabicus Project”, Bulletin of the American Schools of Oriental Research Supplement 26 (1990) 89–136. “New Light on the Roman Frontier in Arabia”, in H. Vetters and M. Kandler, eds., Akten des 14. Internationalen Limeskongresses

xxxii

ESSAYS IN HONOR OF S. THOMAS PARKER

1986 im Carnuntum (Vienna: Österreichische Akademie der Wissenschaften, 1990) 215–230. “The 1989 Season of the Limes Arabicus Project”, ACOR Newsletter 2 (1990) 8. 1989 “The Fourth Century Garrison of Arabia: Strategic Implications for the Southeastern Frontier”, in D. H. French and C. S. Lightfoot, eds., The Eastern Frontier of the Roman Empire: Proceedings of colloquium held at Ankara in September, 1988. BAR International Series 553 (Oxford: British Archaeological Reports, 1989) 355–372. “Digging Diocletian’s Frontier in Jordan”, Minerva: The International Review of Art and Archaeology 1 (1989) 21–24. “The Roman Frontier in the Middle East: Limes Arabicus Project”, American Schools of Oriental Research Newsletter, Fall (1989) 6–7. “The Limes Arabicus Project – The 1989 Campaign”, Liber Annuus 39 (1989) 261–263. “Lejjun”, in D. Homes-Fredericq and J. B. Hennessy, eds. Archaeology of Jordan. Akkadica Supplementum VIII1 (Leuven: Peeters, 1989) 359–367. 1988 “The Limes Arabicus Project: The 1987 Campaign”, Annual of the Department of Antiquities of Jordan 32 (1988) 171–187. “Le Limes Arabicus: La forteresse de Lejjûn et les forts de Khirbet el-Fityân et de Qasr Beshir, 1985 et 1987”, Revue Biblique 95 (1988) 251–267 (in French). “Digging Diocletian’s Frontier: The 1987 Season of the Limes Arabicus Project”, American Schools of Oriental Research Newsletter 39.2 (1988) 11–12. “The Limes Arabicus Project: The 1987 Campaign”, Syria 65 (1988) 419–423. “Preliminary Report on the 1985 Season of the Limes Arabicus Project”, Bulletin of the American Schools of Oriental Research Supplement 25 (1988) 131–174.

S. THOMAS PARKER

xxxiii

1987 “Peasants, Pastoralists, and Pax Romana: A Different View”, Bulletin of the American Schools of Oriental Research 265 (1987) 35–51. “The Roman Limes in Jordan”, in Adnan Hadidi, ed. Studies in the History and Archaeology of Jordan III (London: Routledge and Kegan Paul, 1987) 151–164. “Do You Want to Be a Dig Director?”, American Schools of Oriental Research Newsletter 39.1 (1987) 6–7. “The Limes Arabicus Project: The 1987 Campaign”, Liber Annuus 37 (1987) 393–395, pl. 60, figs. 1–2. 1986 “Retrospective on the Arabian Frontier after a Decade of Research”, in P. Freeman and D. Kennedy, eds. The Defence of the Roman and Byzantine East. British Institute of Archaeology at Ankara Monograph No. 8 and BAR International Series 297 (Oxford: British Archaeological Reports, 1986) 633–660. “Research on the Central Limes Arabicus, 1980–1982”, in Studien zu den Militärgrenzen Roms II. Forschungen und Berichten zur Vor- und Frühgeschichte in Baden Württemberg 20 (1986) 641–648. “A Tetrarchic Milestone from Roman Arabia”, Zeitschrift für Papyrologie und Epigraphik 62 (1986) 246–248. “New Light on Rome’s Arabian Frontier: The 1985 Season of the Limes Arabicus Project”, American Schools of Oriental Research Newsletter 37.2 (1986) 13–14. “The Limes Arabicus Project: the 1985 Campaign”, Syria 63 (1986) 401–405. “The Limes Arabicus Project: The 1985 Campaign”, Archiv für Orientforschung 33 (1986) 269–273. “The Limes Arabicus Project: The 1985 Campaign”, Annual of the Department of Antiquities of Jordan 30 (1986) 233–252. “El-Lejjûn et le Limes Arabicus”, Revue Biblique 93 (1986) 256–261 (in French). 1985 “The Limes Arabicus Project: The 1985 Campaign”, Liber Annuus 35 (1985) 420–424.

xxxiv

ESSAYS IN HONOR OF S. THOMAS PARKER

“Preliminary Report on the 1982 Season of the Central Limes Arabicus Project”, Bulletin of the American Schools of Oriental Research Supplement No. 23 (1985) 1–34, figs. 1–18. 1984 “A Trek along Rome’s Eastern Frontier”, American Schools of Oriental Research Newsletter 2 (1984) 8–12. “Exploring the Roman Frontier in Jordan”, Archaeology 37.5 (1984) 33–39. “Lejjun of the Limes Arabicus” Jordan 9.1 (1984) 20–25. 1983 “An Early Roman Tomb Excavated near ACOR in Amman”, American Schools of Oriental Research Newsletter 7 (1983) 14–15. “The Central Limes Arabicus Project: The 1982 Campaign”, Annual of the Department of Antiquities of Jordan 27 (1983) 213–230. 1982 “Legio IV Martia and the Legionary Camp at el-Lejjun”, Byzantinische Forschungen 8 (1982) 185–210, pls. XV–XXIV. “Preliminary Report on the 1980 Season of the Central Limes Arabicus Project”, Bulletin of the American Schools of Oriental Research 247 (1982) 1–26, figs. 1–18. “Exploring a Roman Frontier: The 1982 Season of the Central Limes Arabicus Project”, American Schools of Oriental Research Newsletter 5 (1982) 5–14. 1981 “The Central Limes Arabicus Project: The 1980 Campaign”, Annual of the Department of Antiquities of Jordan 25 (1981) 171–178. “The Central Limes Arabicus Project: The 1980 Campaign”, American Schools of Oriental Research Newsletter 8 (1981) 8–20. 1980 “Towards a History of the Limes Arabicus”, ed. by W. S. Hansen and L. J. F. Keppie, Roman Frontier Studies 1979, British Archaeological Reports International Series no. 71 (Oxford: British Archaeological Reports, 1980) 865–878. 1978 “A Military Building Inscription from Roman Arabia”, Zeitschrift für Papyrologie und Epigraphik 28 (1978) 61–66, Tafeln VIII–X.

S. THOMAS PARKER

xxxv

“Area C.4, 6, 8, 9, 10”, from Heshbon Excavations 1976, ed. by R. Boraas and L. Geraty, Andrews University Monographs (Berrien Springs, MI: Andrews University, 1978) 71–107. “The 1976 Limes Arabicus Survey”, American Schools of Oriental Research Newsletter 5 (1978) 6–10. 1976 “The Objectives and Strategy of Cimon’s Expedition to Cyprus”, American Journal of Philology 97 (1976) 30–38. “Archaeological Survey of the Limes Arabicus: A Preliminary Report”, Annual of the Department of Antiquities of Jordan 21 (1976) 19–31. 1975 “The Decapolis Reviewed”, Journal of Biblical Literature 94 (1975) 437–441. Review Articles “The Material Culture and Mission of the Late Roman Army on the Southeastern Imperial Frontier”, review of Gwyn Davies and Jodi Magness, The 2003–2007excavations in the Late Roman Fort at Yotvata (Eisenbrauns: Winona Lake, IN 2015), in Journal of Roman Archaeology 29 (2016) 919–923. “Water and More at Humayma”, review of John P. Oleson, Humayma Excavation Project, 1: Resources, History and the Water-Supply System. ASOR Archaeological Reports 15. Boston: American Schools of Oriental Research, 2010), in Journal of Roman Archaeology 25 (2012) 924–926. “New Work on the Euphrates Frontier in Late Antiquity”, review of Michaela Konrad, Resafa V. Der spätrömische Limes in Syrien. Archäologische Untersuchungen and den Grenzkastellen von Sura Tetrapyrgium, Cholle und in Resafa. (Mainz: von Zabern, 2001) in Journal of Roman Archaeology 16 (2003) 722–724. “Two Late Roman Fortlets in Southern Palestine”, reviews of Mordechai Gichon, En Boqeq. Ausgrabungen in einer Oase am Toten Meer. Band I: Geographie und Geschichte der oase das spätrömische-byzantinische Kastell. (Mainz: von Zabern, 1993); and Richard P. Harper, Upper Zohar. An Early Byzantine Fort in Palaestina Tertia: Final Report of Excavations in 1985–1986. British

xxxvi

ESSAYS IN HONOR OF S. THOMAS PARKER

Academy Monographs in Archaeology 9. (Oxford: British School of Archaeology in Jerusalem, 1995), in Journal of Roman Archaeology 10 (1997) 580–586. “Two Books on the Eastern Roman Frontier: Nomads and Other Security Threats”: reviews of Benjamin Isaac, The Limits of Empire: The Roman Army in the East (Oxford: Oxford University, 1990) and David Kennedy and Derrick Riley, Rome's Desert Frontier from the Air (London: Batsford, 1990), in Journal of Roman Archaeology 5 (1992) 467–472. Reviews Kennedy, David L., Settlement and Soldiers in the Roman Near East. Variorium Collected Studies Series CS 1032 (Farnum, Surrey, Burlington and Vermont: Ashgate, 2013), in Levant 46.2 (2014) 315–316. Sauer, James A., and Herr, Larry G. Ceramic Finds: Typological and Technological Studies of the Pottery Remains from Tell Hesban and Vicinity. Hesban 11 (Berrien Springs, MI: Andrews University Press and Institute of Archaeology, Andrews University, 2012), in Bulletin of the American Schools of Oriental Research 371 (2014) 243–246. Kennedy, David, The Roman Army in Jordan (London: Council for British Research in the Levant, 2004) in American Journal of Archaeology 110.3 (2006) 525–526. Villeneuve, E. and Watson, P. M., eds., La céramique byzantine et proto-islamique en Syrie-Jordanie (IVe-VIIIe siècles apr. J.-C.). Actes du colloque tenu à Amman les 3, 4 et 5 décembre 1994. Bibliothèque archèologique et historique 159. Beirut: Institut français d’archéologie de Proche-Orient, 2001, in Bulletin of the American Schools of Oriental Research 341 (2006) 78–80. Shahîd, Irfan, Byzantium and the Arabs in the Sixth Century, Volume 2, Part 1: Toponomy, Monuments, Historical Geography, and Frontier Studies (Washington, D.C.: Dumbarton Oaks, 2002) in Bulletin of the American Schools of Oriental Research 332 (2003) 107–108. David Cherry, Frontier and Society in Roman North Africa (Oxford: Clarendon, 1998) in The American Historical Review 106.3 (2001) 1032–1033.

S. THOMAS PARKER

xxxvii

David Adan Bayewitz, Common Pottery in Roman Galilee: A Study of Local Trade. Bar-Ilan Studies in Near Eastern Language and Culture. (Ramat-Gan: Bar-Ilan Press, 1993) in Bulletin of the American Schools of Oriental Research 299/300 (1995) 127–130. Richard Stoneman, Palmyra and Its Empire: Zenobia’s Revolt against Rome. (Ann Arbor: University of Michigan Press, 1992) in Echos du monde classique/Classical Views 39 (1995) 310–313. Rome's Alpine Frontier. Proceedings of the Conference held at the Center for Old World Archaeology and Art, Brown University, September 27th 1986. (Providence: Brown University, 1990), in Archaeological News 18 (1993) 19–20. Irfan Shahîd, Byzantium and the Arabs in the Fifth Century (Washington: Dumbarton Oaks, 1989), in Speculum 67 (1992) 482–484. M. Chahin, The Kingdom of Armenia (London: Croom Helm, 1987), in The Historian 51 (1989) 483–484. Manfred Lindner, ed., Petra. Neue Ausgrabungen und Entdeckungen (Munich: Delp, 1986) and J. Starcky and M. Gawlikowski, Palmyre (Paris: Librairie d’Amerique et d'Orient, 1985), in American Journal of Archaeology 93 (1989) 158–160. G. W. Bowersock, Roman Arabia (Cambridge: Harvard University, 1983) in Bulletin of the American Schools of Oriental Research 258 (1985) 75–78. Sartre, Maurice, Trois études sur l’Arabie romaine et byzantine, Collection Latomus v. 178 (Brussels: Latomus, 1982) in Gnomon 49 (1983) 770–772. Publications Forthcoming “Nabataean and Late Roman Domestic Complexes at Petra”, Studies in the History and Archaeology of Jordan XIII. Amman: Department of Antiquities. “Military Diet versus Civilian Diet on the Arabian Frontier”, Acta of the XXIII Congress of Roman Frontier Studies. “Petra North Ridge Project: The 2014 Season”, Annual of the Department of Antiquities of Jordan 59 (2015) (co-author). “New Research on the Roman Frontier in Arabia” in Roman Frontier Studies 2009: Acta of the XXI Congress of Roman Frontier Studies.

xxxviii

ESSAYS IN HONOR OF S. THOMAS PARKER

Publications In Preparation Rome in the Middle East: Imperial Security Policy from Pompey to Heraclius, 64 B.C.-A.D. 630. Final Report on the Roman Aqaba Project 1994–2002. Vols. 2–3.

S. THOMAS PARKER

xxxix

S. Thomas Parker in Petra in 1975 (Courtesy of S. Thomas Parker)

Tom Parker in front of his Area A staff at the entrance to the Umm al-Jimal Barracks in 1977 (Photo by Bert de Vries)

xl

ESSAYS IN HONOR OF S. THOMAS PARKER

Tom Parker speaking at the celebrations for ACOR 25 year anniversary at the current Tla’Ali facility. Circa July 1993. To his right are Artemis Joukowsky, HRH Hassan bin Tallal, and HE Yanal Hikmet, Minister of Tourism. (Photo from the ACOR archives, courtesy of Sarah Harpending and Barbara Porter)

Tom Parker supervising excavation at the Petra North Ridge project in 2012 (Courtesy of S. Thomas Parker)

ABBREVIATIONS ADAJ AJA AJPH Ä&L ACOR ANRW ASOR BA BAR BAR-IS BASOR BZ CIIP 2

CIK CIS CSEL CCSL CSCO CRAI IEJ JARCE JECS JESHO JNES

Annual of the Department of Antiquities of Jordan American Journal of Archaeology Americal Journal of Philology Äegypten und Levante / Egypt and the Levant American Center of Oriental Research Aufstieg und Niedergang der römischen Welt American Schools of Oriental Research The Biblical Archaeologist Biblical Archaeology Review British Archaeological Reports, International Series Bulletin of the American Schools of Oriental Research Byzantische Zeitschrift W. Ameling et al., eds, (2011). Corpus Inscriptionum Iudaea/Palaestinae, 1: Caesarea and the Middle Coast 1121–2160. Berlin and Boston: Walter de Gruyter. W. Caskel. (1966). Ğamharat an-nasab; das genealogische Werk des Hišam Ibn Muḥammad al-Kalbī. Leiden: Brill. Corpus Inscriptionum Semiticarum. Corpus Scriptorum Ecclesiasticorum Latinorum Corpus Christianorum, series Latina. Corpus Scriptorum Christianorum Orientalium Comptes rendus des séances de l’année de l’Académie des inscriptions et belles Israel Exploration Journal Journal of the Archaeological Research Center in Egypt Journal of Early Christian Studies Journal of the Economic and Social History of the Orient Journal of Near Eastern Studies xli

xlii JRA JThS LA NEAEHL OC OCP PEF PEQ PG PL RB ROC SHAJ T&MByz ZDPV ZPE

ESSAYS IN HONOR OF S. THOMAS PARKER Journal of Roman Archaeology Journal of Theological Studies Liber Annus New Encyclopedia of Archaeological Excavations in the Holy Land Oriens Christianus Orientalia Christiana Periodica Palestine Exploration Fund Palestine Exploration Quarterly Patrologiae Curus, series Graeca Patrologiae Curus, series Latina Revue Biblique Revue de l’Orient Chrétien Studies in the History and Archaeology of Jordan Travaux et mémoires. Centre de recherche d’histoire et civilisation de Byzance Zeitschrift des Deutschen Palästina-Vereins Zeitschrift für Papyrologie und Epigraphik

NEOLITHIC PERIOD (CA. 9,700–4,900 BCE) PRESENCE IN THE SOUTHERN TRANSJORDAN PLATEAU, THE SOUTHERN GHORS, WADIS FIDAN AND FAYNAN, AND THE EAST SIDE OF WADI ʿARABAH BURTON MACDONALD 1 INTRODUCTION

This paper is a summary of Neolithic presence in the Southern Transjordan Plateau, the Southern Ghors, Wadis Fidan and Fayan, and the east side of Wadi ʿArabah (Figure 1). However, the emphasis will be on the areas I surveyed as director of five archaeological projects in the region (see below). One area that is excluded from this summary is the territory between the southern extremity of “The Tafila-Busayra Archaeological Survey” and the northern limits of “The Shammakh to Ayl Archaeological Survey” (see below), that is, for the most part the Shawbak region. Fujii et al. (2013) are Thomas S. Parker has been engaged in archaeological research in Jordan for over 40 years. Some of his research involved both surveys and excavations in the southern part of Jordan, specifically on the southern Transjordan Plateau and the southern Wadi `Arabah, as part of his interest in the Roman limes (Parker 1976; 1986). Although the time periods and subject of this paper are other than Tom’s specialization, he assisted me on various occasions during pottery “reading” from my five archaeological-survey projects and saved me from many errors. 1

3

4

BURTON MACDONALD

presently carrying out “The Shawbak North Archaeological Project” in this area. Before presenting the archaeological evidence for Neolithicperiod presence in the area of interest I will make some general comments about the time-frame of the Neolithic period and its climate. I will conclude this paper with the Neolithic period and the earliest use of copper ore from the Wadi Faynan area.

Figure 1. Geographical area of interest (prepared by Hilary Hatcher)

NEOLITHIC PERIOD PRESENCE

THE TIME-FRAME OF THE NEOLITHIC PERIOD

5

The beginning and end of the Neolithic are not clearly defined. The former is associated with the beginning of agriculture which was a gradual process that began in the previous Epipaleolithic period. Along with the onset of farming were changes in material culture such as the first instance of true arrowheads in the form of alKhiam and similar projectile points (Rollefson 2008:71). As for the end of the Neolithic, the transition to the Chalcolithic remains obscure. This is due to the fact that the material correlates for the change are a matter of dispute (Rollefson 2008:71, 98). Within the Neolithic, the two main chronological divisions are the Pre-Pottery and the Pottery Neolithic. Relative to the former, it is divided into the Pre-Pottery Neolithic A, B, and C (often designated as PPNA, PPNB, and PPNC). Furthermore, the Pre-Pottery Neolithic B is divided into Early, Middle, and Late (Rollefson 2008:71, Table 1). The distinctions among the various divisions of the Neolithic are based on such factors as: lithic technology and typology; subsistence economy; architecture; and ritual (Rollefson 2008). An awareness of these distinctions and divisions will be helpful in the discussion that follows. Dates for the various periods of the Neolithic are: Pre-Pottery Neolithic A (9,700–8,600 BCE); Pre-Pottery Neolithic B (8,600– 6,900 BCE); Pre-Pottery Neolithic C (6,900–6,500 BCE); and Pottery Neolithic (6,500–4,900 BCE). Of course, there is not complete agreement among scholars, as is evident in the various works cited in this paper, relative to this chronology. For the Levant in general, the period of the Pottery Neolithic is a time when there is the re-establishment of population that had suffered a crisis at the end of the seventh and the beginning of the sixth millennium BCE (example of ʿAyn Ghazal; see, for example, Kohler-Rollefson 1988; Rollefson and Kohler-Rollefson 1989; 1992). There appears to have been a slow recovery towards stability throughout the sixth and fifth millennia BCE. The reason for this change does not appear to be climate alone since there is no evidence for a significant change in climate in the second half of the sixth and in the fifth millennia BCE. Gopher (1995:221) sees the change due a restructuring of society in combination with external influences and local environmental conditions. The most widely attested Pottery Neolithic tradition in Jordan is that which is called the Yarmoukian. Rollefson sees this tradition

6

BURTON MACDONALD

as a direct outgrowth of the Pre-Pottery Neolithic C (2008:95). The tradition is distinguished by its distinctive ceramics, for example, banded herringbone impression (Kafafi 1993). There is no absolute indication as to when the Yarmoukian tradition came to an end. The Yarmoukian tradition and the pottery recovered from the Pottery Neolithic A levels of Jericho are clearly distinguished. The easiest features to recognize are the use of red slips and a red pigment applied in geometric motifs. Although the Jericho style is characteristic of the area west of the Jordan River, it appears to have constituted the major ceramic tradition in southern Jordan (Rollefson 2008:97).

THE CLIMATE

The beginning of the Holocene period, ca. 12 thousand years ago, is marked by the expansion of deciduous oak and the masking of herbaceous vegetation in the pollen record. This reforestation occurred along with an increase in humidity, following the meltwater discharge into the global ocean (Bard et al. 1987). This also coincided with the beginning of a period of high freshwater discharge into the eastern Mediterranean from the Nile and other rivers (Fontugne et al. 1994). Moreover, el-Moslimany (1994) found that during the almost three-thousand-year period between 9,500 and 6,750 bp, there was a high percentage of arboreal pollen at Ali Kosh and Tepe Sabz – both located in southern Iran – implying increased rainfall. Kafafi (1996), on the basis of paleobotanical and geomorphic evidence from Pre-Pottery Neolithic B levels at Beidha, posits that the eastern desert of Jordan was very moist from 8,800 to 8,250 bp. Weninger posits an abrupt major (50 m) rise in the Dead Sea levels at the onset of Middle Pre-Pottery Neolithic B (ca. 10 thousand years cal. B.P. [8,000 BCE]) (2009: 9, Fig. 4) and an abrupt dip in Dead Sea levels at the end of the Late Pre-Pottery Neolithic B/C (ca. 8.6 thousand years cal B.P. [ca. 6,600 BCE]). The latter is during a global cold period, during which extreme drought is registered in the southern Levant (Weninger 2009: 9, Fig. 4). Frumkin et al. (1991), through a study of salt caves at Mount Sedom along the southwestern shore of the Dead Sea, determined that 7,000 B.P. was a wet period. They support this conclusion on the basis of their humidity indicators: abundance of wood (more trees means more water); cave width (wider indicates more humidity, since there is more water to erode the caves); and height of out-

NEOLITHIC PERIOD PRESENCE

7

let (higher outlets mean the water was high enough to erode through at that point which is indicative of more humidity). All the above evidence is in line with wetter conditions in the Jordan Valley (Schuldenrein and Goldberg 1981; Goldberg and Bar-Yosef 1982) and the southern and eastern parts of the area that were probably influenced by rains derived from the south, probably the African monsoon (Goldberg 1986: 241–242).

THE ARCHAEOLOGICAL EVIDENCE

Neolithic-period presence is well represented on the southern Transjordan Plateau which extends for a distance of ca. 115 km north-south, from Wadi al-Hasa in the north to the edge of the escarpment at Ras an-Naqab in the south. In addition, I will consider Neolithic presence in areas to the west in the Southern Ghors, Wadis Fidan and Faynan, and the eastern side of Wadi ʿArabah (Figure 1). I begin with the evidence for the Neolithic period from within the territories of the survey projects that I directed on the Southern Transjordan Plateau, namely, “The Wadi al-Hasa Archaeological Survey” (WHS) (MacDonald et al. 1988), “The Tafila-Busayra Archaeological Survey” (TBAS) (MacDonald et al. 2004), “The Ayl to Ras an-Naqab Archaeological Survey” (ARNAS) (MacDonald et al. 2012), and “The Shammakh to Ayl Archaeological Survey” (SAAS) (MacDonald et al. 2010; 2011; 2016) between 1979 and 2012. I will then turn my attention to evidence from “The Southern Ghors and Northeast ʿAraba Archaeological Survey” (SGNAS) (MacDonald et al. 1992) along with Neolithic presence in the Wadis Fidan and Faynan region, and the eastern side of Wadi ʿArabah. In conjunction with the results from the archaeological surveys that I conducted, I will include Neolithic-period sites documented, and some excavated, by other researchers, including Parker and Smith II (2014). The Wadi al-Hasa Archaeological Survey (WHS) Territory Pre-Pottery Neolithic The Pre-Pottery Neolithic is well documented at a number of sites in the WHS territory. Two of these sites, namely, Khirbat. alHammam/Abu Ghrab and Abu Hudhud, are located in the central

8

BURTON MACDONALD

segment of Wadi al-Hasa. In addition, there are nine sites from the period located in the eastern segment of the Wadi al-Hasa. They are: a mixed-Epipalaeolithic/Pre-Pottery Neolithic site; three single-component Pre-Pottery Neolithic ones; and five Pre-Pottery Neolithic B burin sites (MacDonald et al. 1988).

Figure 2. Map of sites referred to in the chapter

Khirbat al-Hammam/Abu Ghrab (WHS Site 149) is located in agricultural fields, between Wadis Laʿban and ʿAfra, leading down to Wadi al-Hasa. Farmers working in the area informed WHS team

NEOLITHIC PERIOD PRESENCE

9

members that the area was once the location of a large ancient ruin. On visits to the site WHS survey-team members collected lithics from the Early Pre-Pottery Neolithic period along with Pottery Neolithic sherds (MacDonald 1988: 128). Rollefson and Kafafi (1985) examined the site in 1984 and judged it to be an agricultural village dating to phases of the Late Pre-Pottery Neolithic B period. The chipped stone tools they collected at the site included: sickle blade; transverse burin; damaged, leaf-shaped spear point; multiple burins on notches; a wedge; scrapers; and a polished adze/celt (Rollefson and Kafafi 1985: 66, Fig. 3; 67, Fig. 4). Peterson carried out excavations at the site in 1999 and 2006 and accessed it to be 6–7 hectares in size and comprised of substantial domestic structures. Her collections at the site included stone and bone tools, a small number of ornamental objects, and grinding stones (Peterson 2004) and the fragments of a skull of a three-four year old individual (Peterson 2004: 5). Peterson posits that “collections of lithic and organic specimens allow a multidimensional reconstruction of Neolithic lifeways” (2004: 3). She concludes that the site has a complex occupational history with PrePottery Neolithic A, Middle Pre-Pottery Neolithic B, and Late PrePottery Neolithic B components (dating to at least 8,300 radiocarbon years bp) (Peterson 2003; 2004; 2007; see also Peterson and Neeley 2007; Peterson et al. 2010: 408). She estimates that there are stratified deposits in excess to 2 m deep exposed in a road cut running through the site. The site has a great deal of archaeological potential for future excavation. Abu Hudhud (WHS Site 1008) is located in the central segment of Wadi al-Hasa, immediately to the east of its confluence with Wadi al-ʿAli. Rollefson visited the site in 1982 and 1990 and collected ten Helwan points, cressed blades cores, picks, scrapers, lustered blades, transverse burins, borers/drills, and a broken basalt pestle. On the basis of the Helwan points, collected at each visit, Rollefson suggests that the surface scatter belongs to the Early PrePottery Neolithic B period (1996: 159). The stone alignments on the surface of the site lead him to believe that the site was at least a semi-permanent settlement consisting of oval or subrectangular

10

BURTON MACDONALD

structures (ca. 3.00 x 2.50 m) (Rollefson 2008:74). However, there is the possibility that these stone alignments could date to another or other periods (Rollefson 1996: 160). 2 At the eastern end of Wadi al-Hasa, Clark et al. (1988) studied four WHS collections from non-Natufian, Epipalaeolithic and PrePottery Neolithic sites for comparative purposes. Three were considered single-component Pre-Pottery Neolithic ones. All are relatively pristine Pre-Pottery Neolithic. The fourth is a mixed Epipalaeolithic/Pre-Pottery Neolithic collection. The results of Clark et al.’s analyses indicate that lithic artifacts assigned to the Pre-Pottery Neolithic from these sites are representative of at least the gamut of morphological types present in these time frames. In their opinion, they display, moreover, a considerable uniformity in raw material. However, it is not known whether this preference for a certain kind of raw material has any general currency for other contemporaneous WHS collections (Clark et al. 1988: 107, Table 13, and 108). According to Clark et al. (1988), as far as the debitage fraction is concerned, what distinguishes these four collections from those of the Upper Paleolithic is a much higher incidence of bladelets. Although the cores do not allow for much discrimination, the mixed core category is dominated by nuclei from which small blades and bladelets were detached, rather than the flakes and blades of the Upper Paleolithic. Moreover, from their study of the collections, they determined that there is no longer any trace of Levallois technique (Clark et al. 1988: 104–05). As indicated above, WHS team members also reported five Pre-Pottery Neolithic B burin sites in the eastern segment of Wadi al-Hasa. According to Clark et al., “these sites in general seem to represent ephemeral campsites produced by repeated, short-term, sporadic visits to the ridge crest” (1988: 108). The sites’ location provides a sweeping vista of the surrounding terrain and the deeply incised bed of Wadi al-Hasa, a natural route for seasonally migratoWHS Site 1007, from which Rollefson collected both PPN and Epipalaeolithic material, is located immediately to the west of WHS 1008. It is presently separated from it by an erosional gulley. In antiquity, it could have been connected (Rollefson 1996:159). 2

NEOLITHIC PERIOD PRESENCE

11

ry animals. The burin-dominated assemblages might be the remnants of limited activity stations which function, at least in part, to monitor the movement of game (Clark et al. 1988: 108). Pottery Neolithic WHS team members collected Pottery Neolithic sherds from five sites within the WHS territory. These sites are: Khirbat. Hammam/Abu Ghrab (WHS 149); Ras al-Siq (WHS 307); Khirbat edhDharih (WHS 524); and WHS Sites 857 and 870 (MacDonald et al. 1988: 128–131; 129, Table 21; 130, Fig. 38). The first of these sites has been treated above as a Pre-Pottery Neolithic one. Ras al-Siq is located, like Khirbat Hammam/Abu Ghrab, on slopes to the east of Wadi Laʿban, leading down to Wadi al-Hasa. It consists of an upper and lower segment, both of which are cut by a small wadi. Both segments showed signs of walls or fences. However, it is not certain that these features are contemporaneous with the Neolithic ceramics collected. Khirbat edh-Dharih is also located on a slope. However, in its case, the slope leads down to Wadi Laʿban. WHS team members discovered the site due to the fact that remnants of it were visible in a road cut on the west side of the main road between Wadi alHasa and at-Tafilah (MacDonald et al. 1988: 129–130). The site is now completely destroyed. WHS Sites 857 and 870 are located at higher elevations than the three, previously-described ones (MacDonald et al. 1988: 131). They are located on a ridge and a plateau and to the east and south of Wadi el ʿAli and Wadi al-Hasa respectively. While Khirbat Hammam/Abu Ghrab, Ras al-Siq, and Khirbat edh-Dharih are close to present sources of water, WHS Sites 857 and 870 are presently removed from an easy access to water. Of course, as has been noted previously, the wadis in question would have been down-cut considerably over the centuries. Site 857 consists of a number of structures which could be towers and/or tombs (MacDonald et al. 1988: 131). It is impossible to determine, without excavation, whether or not all or some of the structures are contemporaneous and whether or not they are to be dated to the Pottery Neolithic period. Site 870 consists of two structures, one of which is circular and measures ca. 7 m in diameter, and a ceramic and lithic scatter overlooking a large bend in Wadi al-Hasa (MacDonald et al. 1988:

12

BURTON MACDONALD

131). Thus, like the Pre-Pottery burin sites described above, the site would have had the advantage of monitoring the movements of game in the wadi. 3 The Tafila-Busayra Archaeological Survey (TBAS) Territory Pre-Pottery Neolithic None of the sites in the TBAS region indicated the presence of an early or Pre-Pottery Neolithic component. This is in sharp contrast to Epipalaeolithic presence in the Jurf ad-Darawish region, located in the southeastern segment of the territory. Relative to this, Neeley states, “this gap in the settlement record of the TBAS might indicate an abandonment of the region or the reorganization of a dispersed late Paleolithic population into larger, more centrally located settlements outside the area surveyed” (2004: 45). He gives the example of the Pre-Pottery Neolithic site of Khirbat AlHammam/Abu Ghrab in the WHS territory to the north. In any case, the late Neolithic/Chalcolithic remains suggest a change in the pattern of settlement within the TBAS region. Unlike the previous Upper Palaeolithic and Epipalaeolithic periods, the Late Neolithic/Chalcolithic settlement indicates an increased use of the areas distant from the Pleistocene lakes. In Neeley’s opinion, this reversal There are indeed other Neolithic sites in the immediate vicinity of the ones described above. However, they are all located to the north of Wadi al-Hasa. One of these, Al-Hamma, is located on a terrace in the Wadi al-Hasa near the Tannur Dam Project, several kilometers to the northeast of WHS 149, Khirbat Hammam/Abu Ghrab. Preliminary analyses suggest that occupation at the site dates to LPPNB (Rollefson 1999; Makarewicz and Goodale 2004; Makarewicz and Austin 2006). In addition, since 2007, a team from the University of the Aegean, Rhodes, Greece, participating in a survey in the as-Safi area, discovered two PPNB sites, Wadi Hamarash 1 (Sampson 2010a; 2011; 2013) and Wadi Hamarash 4 (Sampson 2010b; 2011), located at the western end of Wadi alHasa. The former is a PPNB while the latter is a PPNA site. However, they too are located north of wadi and will not be considered in what follows. 3

NEOLITHIC PERIOD PRESENCE

13

from an earlier trend could be due to a climate change or a shift to agro-pastoral subsistence activities (2004: 45). Pottery Neolithic TBAS team members collected Neolithic-Chalcolithic sherds at three sites. Moreover, they collected Neolithic/Chalcolithic sherds at two sites. All of these sites are located in the southern segment of the Pleistocene lakes to the south of Jurf ad-Darawish. In contrast to the WHS territory, there were no sites in the TBAS one that yielded clearly-defined ceramics from the Neolithic period (Neeley 2004:54). The Shammakh to Ayl Archaeological Survey (SAAS) Territory Within the SAAS territory proper, the Neolithic period is represented by two excavated sites, Basta (SAAS Site 23; MacDonald et al. 2016: 144–145) and al-Basit (Wadi Musa 8; ʿAmr and al-Momani 2001). SAAS team members found no identifiable Neolithic pottery in the other 364 sites documented. Moreover, the period is represented by what may be sherds from the period at only Random Square 048, located in the eastern segment of Zone 3. Basta is an excavated Late Pre-Pottery Neolithic B (LPPNB) village (Nissen and Gebel 1987; Nissen et al. 2004; Gebel et al. 2006). It is located ca. 12 km to the southeast of Wadi Musa, 19 km southeast of Beidha (see below), and within the modern village by the same name. The LPPNB component of the site, which lasted from ca. 7,500–6,900 BCE, left behind, on the slopes, major architectural features. Post-LPPNB Neolithic squatters modified the ruins (6,900–6,500 BCE), and Pottery Neolithic people inserted flimsy structures and casual deposits (6,500–6,200 BCE) at it (Gebel et al. 2006: 205). Due to the fact that the majority of the LPPNB village is under the modern village, only one percent of it could be excavated. SAAS team members recorded the site so that our site inventory of the territory would be as complete as possible. The LPPNB village of Basta covers an area of about 10–12 hectares. Its inhabitants practiced four subsistence strategies: herding; farming; gathering; and hunting. They supplemented lower levels of meat consumption of domestic caprines, especially goat and sheep, by hunting parties which concentrated on the local game which was available in the immediate neighbourhood all year

14

BURTON MACDONALD

round (45 percent of the animal bones in Basta belonged to wild species). In addition, they probably trapped herds whose known migration routes passed in the vicinity of their settlement. On the floral side, Basta has provided the ordinary selection of carbonized field crops, for example, two-row barley and emmer wheat, and garden fruits, for example, pistachio, nuts, almonds, and wild figs (Nissen et al. 2004: 29). The most striking result of the Basta project is its LPPNB architecture with its multi-room domestic units built on terraces. Relative to this, the site was the first one to show the extraordinary preservation of walls which were constructed from dressed ashlars laid out in neat courses with small wedge stones driven into the joints (Nissen et al. 2004: 27). In the opinion of the excavators, Basta may have been part of an expansion of settlers around 7,600/7,500 BCE that started in northern and central Jordan and expanded towards the south. This would have resulted in the occupation of semi-arid territories next to rich springs at the border of the highlands to the east of the Dead Sea Rift Valley (Nissen et al. 2004: 23; see Bender 1974:9). A number of factors are set forth for the collapse of Basta and other contemporaneous sites from southern Jordan. These include: their location in environmentally sensitive fringes; unfavourable climate fluctuation ca. 6,900 BCE; local, man-made environmental disasters, for example, wood harvesting, monocultures, overgrazing; and a social collapse that could not find the social answers expeditiously enough to secure the dynamics and consequences of growth (Nissen et al. 2004: 25–26). It could have been a number of these factors rather than just one that led to the village’s demise. In 1996, during the first season of ʿAmr et al.’s survey of the “Wadi Musa Water Supply and Wastewater Project” Al-Basit (Wadi Musa 8) was discovered. The surveyors dated the site to “Neolithic (predominant), Iron II, Nabataean, Byzantine” (ʿAmr et al. 1998: 519). The Department of Antiquities of Jordan carried out a small scale excavation of the site in 1997 (Fino 1998). The excavator describes the site as “a medium-sized village, c. 5–10 ha. in size, located in the town of Wadi Musa” (Fino 1998: 103). He dates it to the LPPNB period and states that “the 1997 excavation season at al-Basit demonstrates that the architecture and material culture of the site is comparable with other LPPNB sites such as ʿAyn Jammam, Basta and Baʿja” (1998: 108). He notes that the excavated

NEOLITHIC PERIOD PRESENCE

15

area was limited and that the site was in need of further exposure before the whole area is turned over to modern construction activities (1998: 108). In 2011, members of the SAAS project, using the coordinates published by both ʿAmr et al. (1998) and Fino (1998), attempted to locate the site with the aid of hand-held GPS units. The only materials they could find in the area where the site was supposed to have been were a number of new houses and a few undetermined sherds. We did not find any lithics in the immediate area. It seems that recent development has removed/destroyed the site. The Petra Region There are several important Neolithic sites in the Petra region to the west of the SAAS territory. They are Beidha, Baʿja II, Baʿja V, Shaqarat Mazyad, and Sabra 1. D. Kirkbride-Helbaek excavated the site of Beidha, which is located to the north of Petra, over eight seasons (seven between 1958 and 1967 and a final one in 1983) (Kirkbride 1966). Her fieldwork revealed that the site is a multicomponent one: 1) an early Natufian encampment (primarily during the eleventh millennium BCE); 2) a PPNB village (primarily during the seventh millennium BCE); and 3) terraced Nabataean agricultural fields (during the first millennium BCE) (Byrd 2005: 6). It is the second component of the site which is of interest here. Neolithic Beidha was a small agricultural village, between 1,500 and 3,500 square meters in extent. During its excavation, Kirkbride-Helbaek uncovered more than 65 buildings within a long sequence of the Pre-Pottery Neolithic B habitation period. During this time, an architectural progression took place from clusters of oval post-houses, through individual oval and subrectangular buildings, and ultimately to full rectangular buildings with two stories. The layout and distribution of the Neolithic settlement were restricted by several constraints, including a sandstone outcropping on the north and the edge of the alluvial terrace on the south. A seasonal watercourse may have constituted the settlement’s western boundary. The Beidha settlement grew and spread by the founding of new structures and the decay and desertion of old ones. Throughout its history, its organization appears to have been relatively unplanned. No formal or planned community emerged.

16

BURTON MACDONALD

Byrd (2005) argues that small, individual households (best considered nuclear families) constituted the primary residential and economic unit throughout the occupation of Beidha. Production activities and storage were increasingly carried out within buildings. This was especially so with the establishment of two-storey corridor buildings, resulting in a three-fold increase in the interior areas of domestic units. The upper stories of these buildings were undoubtedly the venue for eating, sleeping, and entertaining (Byrd 2005: 132). Over time, according to Byrd, the distinction between private and public space at Beidha increased. Larger, nondomestic buildings coexisted with the domestic dwelling throughout most of the settlement’s history. With time, they increased in size and structural elaboration. Supra-household decision making and related ceremonial activities were probably carried out within these corporate structures (Byrd 2005: 132). Byrd summarizes his work on the Neolithic site of Beidha by stating: “the Neolithic settlement of Beidha is significant for its developmental sequence of architectural style and construction technique … and for understanding community organization and layout. These developments provide considerable insights into how one small community embraced the social and economic changes taking place with the establishment and initial spread of sedentary food-producing communities throughout Southwest Asia” (2005: 133). Baʿja II is located in the rugged sandstone mountains ca. 10 km north of Petra/Wadi Musa and ca. 5 km north of Beidha. Due to its location, it can be reached, through a narrow gorge, only with difficulty. This latter feature raises the possibility of water storage through the use of 1–2 dams in this otherwise arid area (Gebel and Bienert 1997: 14; Gebel and Hermansen 1999: 18). The site is a small, ca. 1.2–1.5 hectares, LPPNB village. It dates to the second half of the eighth millennium BCE (Gebel and Hermansen 2004: 15; Gebel 2014: 235). It is characterized by pueblo-type dense domestic areas. There is good wall preservation at the site with some still standing at least 2.2 m high (Gebel and Bienert 1997: 15). Access to most rooms was probably by ladders through openings in the ceilings and roofs, for only rarely were rooms connected by doors and windows. Outer walls were shared. Interiorly,

NEOLITHIC PERIOD PRESENCE

17

some rooms were connected by shallow doorways and other kinds of openings. The excavators assert that one room at the site had a special function. Its floors and walls were red-stained (re-plastered in places up to three times). In addition, specialists exposed a wall painting at the site (Gebel and Hermansen 2004: 17–18). Gebel and Hermansen (1999: 18) think that Baʿja II’s location, in a hard-to-reach area, is due to competition for restricted habitat. This would have taken place in a period of dense population as early as the seventh millennium BCE. The excavators posit that extended families lived in the houses at Baʿja II. They estimate that 50–60 families lived at the site. Depending on child mortality, the proportion of family members herding outside, and so forth, this would result in a settlement of 400–500 people (Gebel and Hermansen 1999: 19). Sandstone rings of bracelet-size were made at the site for export. They are represented in the archaeological record by broken finished products as well as by their production waste and as stored pre-forms. The inhabitants of Baʿja II hunted for fur. This is in evidence in the form of leopard, foxes, and hyrax remains. The excavators uncovered intra-mural family burials at the site. These burials were placed in two tiny (each < 1 square meter) neighbouring rooms that had no access by floor-related passages (Gebel and Hermansen 2004: 17). During the excavation of Baʿja I, a Nabataean and Late Islamic site, a Neolithic occupation was discovered in the same area. It was given the name Baʿja V (Bienert et al. 2000: 139). The collected projectile points from the site offer a preliminary dating: LPPNB to Early Pottery Neolithic, that is, through the Yarmoukian period. The earlier phase of occupation at Baʿja V corresponds to the LPPNB site of Baʿja II, which is nearby. Further work is required to determine the duration of the site’s occupation as well as its character (Bienert et al. 2000: 140). Shaqarat Mazyad, a MPPNB site, is located ca. 13 km north of Petra. It is described as “the village on the edge” because of its location at the southernmost part of a plateau that overlooks several wadis (Kaliszan et al. 2002). It is comparable in size to Beidha. It has circular semi-subterranean houses up to 7 m in diameter. As is the case with Beidha, its houses occur in clusters, each cluster with

18

BURTON MACDONALD

shared walls. The floors of some of the houses are plastered (Kaliszan et al. 2002: 17). The use of vertical wooden posts in the walls, which was characteristic of Beidha, is also a common feature at Shaqarat Mazyad (Rollefson 2008:77). The walls themselves are ca. 0.65 m thick (Kaliszan et al. 2002: 17). Among the tool group, arrowheads are dominant, knives are almost as common, and sickle blades, although present, are not very common. Among the faunal material analyzed, goat/sheep are by far the most dominant group, with a slight tendency for more goat (Kaliszan et al. 2002: 19). The site of Sabra, a primarily major Nabataean settlement, is located ca. 8 km from the centre of Petra. There are also several Palaeolithic sites in the area. Gebel and Starck (1988) identified another site as Sabra 1. They date it to the early Pre-Pottery Neolithic (1988: 78) while Rollefson designates it as Pre-Pottery Neolithic A (2008: 71). According to Gebel and Starck, the site consists of “superimposed and overlapping alignments of burnt sandstone and limestone fragments with charcoal indicate a succession of fireplaces” (1988: 78). They uncovered at the site “burnt and unburnt animal and human bones, land snails … flint artefacts, and a few ornaments. No grinding slabs, handstones, etc.” (1988: 78). Sickle blades are very rare. The Ayl to Ras an-Naqab Archaeological Survey (ARNAS) Territory ARNAS team members collected neither Neolithic lithics nor sherds within the survey territory. Nevertheless, there is one Neolithic site documented in the ARNAS inventory. ʿAyn Jammām South (ARNAS Site 184; MacDonald et al. 2012: 183) is a Neolithic farming village located on a steep slope at the edge of the escarpment near Ras an-Naqab. Personnel from the Department of Antiquities of Jordan excavated the site in 1992 and 1996. On the basis of two C14 dates, Bisheh et al. date the site to “8030 + 120 b.p. and 8520 + 190 b.p., which would place the visible structures between the end of the PPNB and the recently defined PPNC period” (1993: 122). However, Waheeb states, “the preliminary assessment of the architecture and material discovered

NEOLITHIC PERIOD PRESENCE

19

in the field suggests a continuation from the Late PPNB up to Late PNA” (1996: 342) or from ca. 8,500–8000 bp (for the LPPNB) to ca. 7,500–5,700 bp (for the PN) (Waheeb and Fino 1997: 215). 4 The site’s architecture was well preserved, including some walls that stood 3 m high. According to the excavators, the Late Pre-Pottery Neolithic buildings show many similarities with Basta and Baʿja. Both their floors and walls were plastered. With the appearance of ceramics at the site, some architectural development distinguished the newer structures (Waheeb and Fino 1997: 214). The Southern Ghors and Northeast ʿArabah Archaeological Survey (SGNAS) Territory It must be pointed out that major Neolithic sites are located along the east side of the Dead Sea to the north of the SGNAS territory and in the vicinity of the Bab adh-Dhraʿ (Schaub and Rast 1989; Rast and Schaub 2003; Ortner and Frohlich 2008). These sites are Zahrat adh-Dhraʿ 2 – PPNA (Edwards et al. 2002; 2004; Edwards and House 2007) and Dhraʿ – PPN and PN (Finlayson et al. 2003) period ones. It is most likely that there would have been contacts in some form between the occupants of these sites and the people who occupied the ones I will consider below. Eight SGNAS sites contain Neolithic components. However, six of these sites also contain later components of either Chalcolithic or Early Bronze Age. Four of these sites are located in the Southern Ghors and four are in Wadi Fidan. Three of the latter four appear to represent substantial occupations (MacDonald et al. 1992). From SGNAS survey evidence, it appears that areas in the Southern Ghors are occupied for the first time during the Neolithic period. The reason for no earlier occupation is the presence of Lake Lisan. Neolithic components are found in Wadi Fifa and just There is a mistake in the final report on the ARNAS project. In the publication, the “Description” for Site 183, ʿAyn Jammam North, ought to be the “Description” for Site 184, ʿAyn Jammam South, that is, the Neolithic village, and vice versa (MacDonald et al. 2012: 183). In other words, the “Description” for the sites has been mistakenly interchanged. 4

20

BURTON MACDONALD

to its north in Wadi Madsus al-Shamali. However, the one in the latter wadi is dated from a single point fragment. Moreover, it is the only one of the eight sites that is not situated at or near the mouth of a wadi. In summary, Neolithic sites in Wadis Fifa and Fidan are clustered together. These two wadis presently have springs and could have been more attractive for settlement than other wadis of the area (Neeley 1992: 37). In dividing these Neolithic sites into pre-pottery or pottery components, SAAS Sites 21, 76, 92, and 95 contain pottery without lithic diagnostics and are considered to be later developments. As indicated above, three of these four sites are located in the region of Wadi Fifa. Of these three, Sites 92 and 95 are lithic and sherd scatters while Site 76, the eastern segment of ancient Fifa, is a multi-period one and the location of a major Early Bronze Age I cemetery (MacDonald et al. 1992: 257), a portion of which was excavated in 1989–1990 (Chesson and Schaub 2007:255; and see below). However, it could also have been the location of Neolithic and/or Chalcolithic settlement(s) since SGNAS team members collected, among several archaeological periods represented at the site, Neolithic and Chalcolithic sherds (MacDonald et al. 1992: 257). The Wadis Fidan and Faynan Area Jabal Hamrat Fidan 001 (SGNAS Site 12) is located on a small “island” at the mouth of Wadi Fidan. It has been excavated but not published. It is generally ascribed to the Pre-Pottery Neolithic B period (Meadows 2001:162). The Neolithic sites of Wadi Faynan 16, Ghuwayr I and Tall Wadi Faynan are located just to the south of the SGNAS territory. They are within the area of interest here. With the development of moister climate at the beginning of the Holocene, ca. 10,000 BCE, people started to camp near springs at the foot of the Transjordan Plateau for several months of the year (Barker and Mattingly 2007: 427). This conclusion comes from the work of “The Dana-Faynan-Ghuwayr Early Prehistoric Project” which surveyed and evaluated Wadi Faynan 16 (WF16), located on a pronounced knoll at the confluence of Wadis Ghuwayr, Fayan and Shayqar, and by means of a small-scale excavation. Mithen and Finlayson, project directors, determined that WF16 is a PPNA settlement with good preservation and a rich cultural inventory (2007: 470). They concluded that PPNA activity was feasibly

NEOLITHIC PERIOD PRESENCE

21

present at the site by 11,600 BP, that is, ca. 9,500 BCE, and lasted until 10,200 BP, coinciding with the dramatic global warning at the start of the Holocene (Mithen and Finlayson 2007: 472). Plant taxa recovered from the excavated areas indicate a riverine environment. In addition, the animal and bird fauna as well as the plant remains from the trenches indicate that a diverse range of resources was available within the region of the site. Moreover, the site’s position on a knoll would have provided a good vantage point from which its inhabitants could observe the movements of game and the seasonal migrations of birds using the wadis as their access to the southern Transjordan Plateau (Mithen and Finlayson 2007: 476). Evidence indicates that at least four different architectural styles were used at WF16: boulder walls; mud walls; irregular stone walls; and stone walls with orthostats. The structures were predominantly circular but varied considerably in size, some ca. 3.5 m in diameter, and internal features. The knoll was covered with structures, probably numbering 20–30. These structures, however, were not necessarily all in use contemporaneously. Mithen and Finlayson, nevertheless, think that all the structures date to the PPNA period or earlier (2007: 476). An especially interesting structure at the site appears to have been built by digging a pit and then lining its walls with a mud mixture. A mud floor was then constructed and surrounded by two tiers of benches, about 1 m deep and 0.50 m high, recalling an amphitheater. Postholes indicate that a roof covered a section of the structure. It could have been used for the collective processing of plants and/or it offered a place for communal gatherings. The occupants of the WF16 village could have been either hunter-gatherers or early farmers. On the other hand, the site could represent intermittent occupation by small groups, perhaps no more than hunting and gathering parties foraging in Wadi Faynan (Mithen and Finlayson 2007: 479). Ghuwayr I, which Simmons and Najjar (2006; 2007) excavated, is a small, ca. 1.25 hectares, MPPNB village in Wadi Faynan. A series of radiocarbon determinations indicates an occupation of ca. 500 years (ca. 10,033/9,542 to 9,510/9,285 ca. B.P./ca. 7500 cal. BCE [Barker and Mattingly 2007: 427]) To date, only a small portion, ca. 900 square meters, of the site has been excavated. This has

22

BURTON MACDONALD

shown, however, that depths of cultural deposits exceed 5 m in places. The site has spectacularly preserved architecture. Relative to this, Simmons and Najjar (2006; 2007) have exposed a distinct village layout with room blocks consisting of several large (ca. over 4 square meters) rooms flanked by smaller (ca. 1 square meter) units. Some rooms also suggest ritual functions. Moreover, there is evidence for two-storey dwellings. Non-domestic architecture indicates communal activities, for example, an outdoor stairway/“public area,” terracing, and entryway, beyond the household. The artifacts at the site are abundant and varied. The large chipped-stone assemblage is characteristically PPNB. In addition, ground stone is varied and numerous. Other materials include shell and a limited use of locally-available malachite. The excavators uncovered the remains of barley, emmer wheat, einkorn wheat, and pea. Seeds and fruits from shrubs and trees including figs and pistachio and wild herbs are common. The faunal assemblage is large with ca. 76 percent belonging to caprines. Gazelle, cattle and bird bones are also part of the assemblage. From the above, it may be concluded that the inhabitants of Ghuwayr I’s subsistence base was mixed, containing both domestic and wild resources (Simmons and Najjar 2006; 2007). Palaeoenvironmental data from Ghwayr I suggest a richer and less degraded environment than today’s. Conditions during the life of the site may have been more humid and exhibited less extremes than currently. The site was abandoned after the Neolithic. Simmons and Najjar wonder if its demise was due to ecological degradation brought on by erosion, overgrazing, and over-farming (2007: 239). Najjar et al. excavated Tall Wadi Faynan, located on the south bank of Wadi Faynan ca. 2 km west of Khirbat Faynan. They labeled it a Pottery Neolithic one (Najjar et al. 1990). However, Bourke refers to it as a Neolithic/Chalcolithic site (2008: 112). He sees it as being likely the main focus of Neolithic/Chalcolithic occupation in the region and radiometric data suggest an extensive Chalcolithic period occupation at the site. According to Bourke, “published Neolithic ceramic material …. have reasonably close form parallels in the early assemblage at Tulaylat Ghassul …, and the lithic assemblage from the site suggests a mixed farming regime, probably based on floodwater irrigation” (2008: 117). How-

NEOLITHIC PERIOD PRESENCE

23

ever, Barker and Mattingly (2007:427) date the site to ca. 5,500 BCE, thus, within the Neolithic time-frame. Hauptmann summarizes the dating for the site: “… Below it starts a phase dating to the Chalcolithic period (stratum 2). The third phase (stratum 3) dates to the Yarmukian and is thus contemporary with the Pottery Neolithic A. Three charcoal samples resulted in 14C-dates from 5520–4910 BCE … and therefore demonstrate a settlement in the second half of the 6th millennium BCE, for which for a long time a decrease in population had been assumed – at least for the larger settlements in Palestine…. Two other 14C samples (4675–4165 BCE) date to the Chalcolithic period (Joffe and Dessel 1995)” (2007: 109). The Southeast ʿArabah Area Evidence for Neolithic sites in the southern Wadi ʿArabah, which Parker and Smith II surveyed as part of “The Roman Aqaba Project” (Parker and Smith II 2014), is lacking. This is in contrast to Chalcolithic or Chalcolithic/Early Bronze Age presence in the territory of “The Southern ʿAraba Archaeological Survey.” The absence, however, could be due to a recognition problem and/or that sites from the period are buried under later sediments (Parker and Smith II 2014: 357–77). The above is in keeping with the survey work of Henry et al. (2001) in the area. They documented no Neolithic sites for Wadi ʿArabah to the south of Wadi Faynan. Nevertheless, Henry et al. (2001: 16), as a result of their work in the southern segment of Wadi ʿArabah, conclude that there was prehistoric occupation of the eastern flank of the wadi over much of the Late Pleistocene and Holocene, spanning Middle Palaeolithic to Chalcolithic periods. However, they also acknowledge that the site density is low and, for the most part, results from ephemeral encampments (Henry et al. 2001).

THE NEOLITHIC PERIOD AND THE EARLIEST USE OF COPPER ORE FROM THE WADI FAYNAN AREA

Excavations and test trenches at sites in both Wadi Faynan and at Ghuwayr I have brought to light large amounts of pure, intensively green-colored copper ores, which had been collected at the outcrops of the mineralizations of the Dolomite Limestone Shale

24

BURTON MACDONALD

(DLS), partly also of the Massive Brown Sandstones (MBS). Greenstone was found at all these sites and were an important source of green mineral pigments next to a number of other “greenstone” from other sources in the region (Hauptmann 2007: 257). There is thus evidence at the sites for the earliest use of copper ore from the Wadi Faynan area. The ores must have been collected in order to exchange them as “exotic goods.” These goods undoubtedly played a significant role in the exchange or trade systems involving different raw materials. However, at this time, the greenstones were used only for cosmetic purposes, in the form of powder or as raw material for beads (Hauptmann 2007: 305). In addition to the above, ore from the Faynan region has been found in settlements of the Southern Levant. For example, trade connections between Faynan and the Jordan Valley are proven by ‘greenstone’ artifacts from PPNA levels at Jericho, where ca. 100 beads and pendants were found. At ʿAyn Ghazal, in the area of Amman, numerous disc- and butterfly-shaped beads have been found. Among the beads from the site are unprocessed pieces of ore consisting of fiber-like Cu silicates and malachite, originating from Faynan. Perhaps such pieces served as the basis for the production of green powder used for cosmetic purposes. The eye shadow and eyeliner around the eyes of some of the ʿAyn Ghazal figurines (clay statues) have a foundation of green material: the X-ray diffracometry proved it to be powdered Cu silicate (Tubb 1985; Hauptmann 2007: 258). Finally, at Beidha (see above), on the plateau, a deliberately shaped roll of powdered Cu silicates, Cu chlorides and malachite a few centimeters long has been uncovered. Hauptmann assumes that it is actual cosmetic powder of the same kind as that which was used at ʿAyn Ghazal. In all probability, the material originates from the DLS ores in Faynan (Hauptmann 2007: 259–260). In summary, it can be said that during the PPNB mostly ores from the DLS were used. But they form only a relatively small amount of the ‘greenstones’ discovered in all the PPNB settlements (Hauptmann 2007: 260–261). From the above, it is evident that greenstone was collected in the Faynan region. From there it was brought/“traded” to the northwest, north and east. Thus, in the PPN period, the copper ore resource of the Dead Sea Rift Valley was valued on the plateau as well as elsewhere.

NEOLITHIC PERIOD PRESENCE

CONCLUSIONS

25

All periods of the Pre-Pottery Neolithic are represent in most segments of the study territory. This is especially true for the PrePottery Neolithic B period which is represented on the Transjordan Plateau in the territories of the WHS, SAAS, and ARNAS projects as well as in the Petra region. In addition, PPNB is present in both Wadis Fidan and Faynan. Although evidence for the Pottery Neolithic is present on the Transjordan Plateau as well as in the Southern Ghors and Wadi Faynan, it is not as common as the PrePottery Neolithic period. But the evidence recovered to date cannot be said to be extremely impressive as far as population numbers are concerned. It is important to note that PPN and PN sites are not as yet documented for the TBAS territory. This appears to indicate an abandonment of the Jurf ad-Darawish region following the Epipalaeolithic Period. Such is certainly not the case for the eastern end of Wadi al-Hasa. It must be noted, however, relative to the PN in the TBAS region, that survey-team member documented a number of Neolithic-Chalcolithic sites. The people who lived on the southern Transjordan Plateau as well as in the area of the Dead Sea Rift Valley during the Neolithic Period would have engaged in farming while at the same time continuing previous-period practices of hunting and gathering. They would have raised and/or hunted goats and sheep as well as migrating animals. In addition, they planted and harvested two-row barley and emmer wheat and garden fruit such as pistachio, nuts, almonds, and wild fig. They would also have continued to gather the edible native plants of the region. PPN and PN sites are documented for the plateau and Wadi ʿArabah as far south as the Faynan region. However, to date, neither PPN nor PN sites have been reported in Wadi ʿArabah south of the Faynan area. Hauptmann states that ore from the Faynan area was used since the PPN period (ninth/eighth millennium BCE). However, at this time, it was used only for cosmetic purposes, in the form of powder or as raw material for beads. From the Faynan region, it was brought/“traded” to the northwest, north and east. Thus, in the PPN period, the copper ore resource of the Dead Sea Rift Valley was valued on the plateau.

26

WORKS CITED

BURTON MACDONALD

ʿAmr, K., Al-Moumani, A., Farajat, S., and H. Falahat. (1998). Archaeological Survey of the Wadi Musa Water Supply and Wastewater Project Area. ADAJ 42, 503–48. Bard, E., Arnold, M., Maurice, P., Duprat, J, Moyes, J., and J.C. Duplessy. (1987). Retreat velocity of the North Atlantic polar front during the last deglaciation determined by 14C accelerator mass spectrometry. Nature 328, pp. 791–94. Barker, G., and Mattingly, D. (2007). Archaeology and desertification: English summary. In: G. Barker, D. Gilbertson, and D. Mattingly, eds., Archaeology and Desertification: the Wadi Faynan Landscape Survey, Southern Jordan,. Wadi Faynan Series Volume 2. Levant Supplementary Series Volume 6. Oxford: Oxbow Books, pp. 425–428. Bender, F. (1974). The Geology of Jordan. Trans. M. K. Khdeir. Berlin: Gebrüder Borntraeger. Bienert, H.-D., Lamprichs, R., and Vieweger, D. (2000). Baʿja – The Archaeology of a Landscape 9000 Years of Human Occupations: A Preliminary Report on the 1999 Field Season. ADAJ 44, 119–48. Bisheh, G., Farajat, S., Palumbo, G., and M. Waheeb. (1993). The Cultural Resources Management Project in Jordan: Archaeological Survey of the Ras an-Naqab – Aqaba Highway Alignment, 1992. ADAJ 37, 119–33. Bourke, S. J. (2008). The Chalcolithic Period.” In: R. Adams, ed., Jordan: An Archaeological Reader. London: Equinox, pp. 109–60. Byrd, B. F. (2005). Early Village Life at Beidha, Jordan: Neolithic Spatial Organization and Vernacular Architecture. The Excavations of Mrs. Diana Kirdbride-Helbaek. British Academy Monographs in Archaeology. New York: Oxford University. Chesson, M. S., and R.T. Schaub. (2007). Fifa, Khirbat al-Khanazir, and Bab adh-Dhraʿ Cemeteries. In: T. E. Levy, P. M. M. Daviau, R. W. Younker, and M. Shaer, eds., Crossing Jordan: North American Contributions to the Archaeology of Jordan. London: Equinox, pp. 253–60. Clark, G.A., Majchrowizz, D., and N. R. Coinman. (1988). A Typological and Technological Study of Upper Paleolithic Collec-

NEOLITHIC PERIOD PRESENCE

27

tions from the WHS with Observations on Adjacent TimeStratigraphic Units. In: B. MacDonald, ed., The Wadi el Hasa Archaeological Survey 1979–1983, West-Central Jordan. Waterloo, ON: Wilfrid Laurier University, pp. 87–127. Edwards, P.C., Meadows, J., Metzger, M., and G. Sayej. (2002). Zahrat adh-Dhraʿ 2: A Pre-Pottery Neolithic A site on the Dead Sea Plain in Jordan. BASOR 327, 1–15 Edwards, P. C., Meadows, J., Sayej, G., and M. Westway. (2004). From the PPNA to the PPNB: New Views from the Southern Levant after Excavations at Zahrat adh-Dhraʿ 2 in Jordan. Paléorient 30 (2), 21–60. Edwards, P. C., and E. House. (2007). The Third Season of Investigations at the Pre-Pottery Neolithic A Site of Zahrat adhDhraʿ 2 on the Dead Sea Plain, Jordan. BASOR 347, pp. 1– 19. El-Moslimany (1994). Evidence of Early Holocene Summer Precipitation in the Contintental Middle East. In: O. Bar-Yosef, and R. S. Kra, eds., Late Quaternary Chronology and Paleoclimtes of the Eastern Mediterranean. RADIOCARBON, 34. Tucson: The University of Arizona, pp. 121 – 130. Finalyson, B., Kuijt, I., Arpin, T., Chesson, M., Dennis, S., Goodale, N., Kadowaki, S., Maher, L., Smith, S., Schurr, M., and J. McKay. (2003). Dhraʿ, Excavation Project, 2002 Interim Report. Levant 35, 1–38. Fino, N. (1998). Al-Basit Neolithic Site in Southern Jordan. ADAJ 42, pp. 103–11. Fontugne, M., Arnold, M., Labeyrie, L., Paterne, M., Calvert, S.E., and J-C. Duplessey. (1994). Paleoenvironment, sapropel chronology and Nile River discharge during the last 20,000 years as indicated by deep-sea sediment records in the eastern Mediterranean. In: O. Bar-Yosef, and R. S. Kra, eds., Late Quaternary Chronology and Paleoclimtes of the Eastern Mediterranean. RADIOCARBON 34. Tucson: The University of Arizona, pp. 75–88. Frumkin, A., Magaritz, M., Carmi, I., and I. Zak. (1991). The Holocene Climatic Record of the Salt Caves of Mount Sedom, Israel. Holocene, 1(3), 191–200.

28

BURTON MACDONALD

Fujii, S., Yamafuji, M., Odaka, T., and Y. Arimatsu. (2013). Preliminary Report of the Shawbak North Archaeological Project (SNAP): 2012–2013. ADAJ 57, 359–71. Gebel, H. G. K. (2014). Non-Formal Sling Balls? Evidence of Geofact Commodification at LPPNB Baʿja, South Jordan. In: B. Finlayson and C. Makarewicz, eds., Settlement, Survey, and Stone: Essays on Near Eastern Prehistory in Honour of Gary Rollefson. Berlin: ex oriente, pp. 235–41. Gebel, H. G., and J. M. Starck. (1988). Late Epipalaeolithic – Aceramic Neolithic Sites in the Petra Area. In: A. N. Garrard, and H. G. Gebel,eds., The Prehistory of Jordan: The State of Research in 1986. British Archaeological Reports International Series 397 (i). Oxford: British Archaeological Reports, pp. 67– 100. Gebel, H. G., and H.-D. Bienert (1997). The 1997 Season at Baja, Southern Jordan. Neo-Lithic, 3/97, 14–18. Gebel, H. G. K., and B. D. Hermansen. (1999). Baʿja Neolithic Project 1999: Short Report on Architectural Findings. NeoLithics, 3/99, 18–21. Gebel, H. G. K., and B. D. Hermansen. (2004). Baʿja 2003: Summary on the 5th Season of Excavation. Neo-Lithics, 2/04, 15– 18. Gebel, H. G., Nissen, H. J., and Z. Zaid. (2006). Basta II: The Architecture and Stratigraphy. Bibliotheca neolithica Asiae meridionalis et occidentalis & Yarmouk University, Monograph of the Faculty of Archaeology and Anthropology, Volume 5. Berlin: ex oriente and Yarmouk University. Goldberg, F. (1986). Late Quaternary Environmental History of the Southern Levant. Geoarchaeology: An International Journal, 1(3), 225–44. Goldberg, F., and Bar-Yosef, O. (1982). Environmental and Archaeological Evidence for Climatic Change in the Southern Levant and Adjacent Areas. In: J. L. Bintcliff and W. van Zeist, eds., Palaeoclimates, Palaeoenvironments and Human Conditions in the Eastern Mediterranean in Later Prehistory. BAR IS-133. Oxford: British Archaeological Reports, pp. 399–414. Gopher, A. (1995). Early Pottery-Bearing Groups in Israel – the Pottery Neolithic Period. In: T. E. Levy, ed., The Archaeology of

NEOLITHIC PERIOD PRESENCE

29

Society in the Holy Land. London and Washington: Leicester University, pp. 205 – 221. Hauptmann, A. (2007). The Archaeometallurgy of Copper: Evidence from Faynan, Jordan. Natural Science in Archaeology. New York and Berlin: Springer. Henry, D. O., Bauer, H. A., Kerry, K. W., Beaver, J. E., and J.J White. (2001). Survey of Prehistoric Sites, Wadi Araba, Southern Jordan. BASOR 323, 3–19. Joffe, A., and J.-P. Dessel. (1995). Redefining Chronology and Terminology for the Chalcolithic of the Southern Levant. Current Anthropology, 39 pp. 297–322. Kafafi, Z. (1993). The Yarmoukian in Jordan. Paléorient 19, 101–14. Kafafi, Z. (1996). The Environment Impact on the Neolithic Settlement Patterns in Jordan. In: The Syrian Peninsula: Cultural Heritage and Contacts International Seminiar. Deir ez-Zor, Syria. Kaliszan, L. R., Hermansen, B. D., Jensen, C. H., Skuldbol, T. B. B., Bille, M., Bangsgaard, P., Ihr, A., Sorensen, M. L., and B. Markussen. (2002). Shaqarat Mazyad – The Village on the Edge. Neo-Lithics 1/02, 16–19. Kirkbride, D. (1966). Five seasons at the Pre-Pottery Neolithic village of Beidha in Jordan. Palestine Exploration Quarterly 98, 8– 72. Kohler-Rollefson, I. (1988). The Aftermath of the Levantine Neolithic Revolution in the Light of Ecological and Ethnographic Evidence. Paléorient 14(1), 87–93. MacDonald, B., Banning, E. B., Clark, G. A., Majchrowicz, D. and Coinman, N. R., Coinman, N., Clark, G. A. and Lindly, J., Donahue, J. and Beynon, D. E., Harlan, J. R., Betlyon, J., King, G., and M. Piccirillo, (1988). The Wadi el Hasa Archaeological Survey 1979–1983, West-Central Jordan. Waterloo, ON: Wilfrid Laurier University. MacDonald, B., ʿAmr, K., Broeder, N. H., Skinner, H. C. W., Meyer, C., Neeley, M. P., Reese, D. S., and D.S. Whitcomb. (1992). The Southern Ghors and Northeast ʿAraba Archaeological Survey 1985–1986, Southern Jordan. Sheffield Archaeological Monographs 5. Sheffield: Collis.

30

BURTON MACDONALD

MacDonald, B., Herr, L. G., Neeley, M. P., Gagos, T., Moumani, K., and M. Rockman. (2004). The Tafila-Busayra Archaeological Survey 1999–2001, West-Central Jordan. Archaeological Reports 9. Boston: American Schools of Oriental Research. MacDonald, B., Herr, L. G., Quaintance, D. S., and H. M. Lock. (2010). The Shammakh to Ayl Archaeological Survey, Southern Jordan: Preliminary Report (First Season 2010). ADAJ 54, 329–44. MacDonald, B., Herr, L. G., Quaintance, D. S., al-Hajj, W., and A. Jouvenel. (2011). The Shammakh to Ayl Archaeological Survey, Southern Jordan: Second Season 2011. ADAJ 55, pp. 363–76. MacDonald, B., Herr, L. G., Quaintance, D. S., Clark, G. A., and M.A. Macdonald. (2012). The Ayl to Ras an-Naqab Archaeological Survey, Southern Jordan 2005–2007. Archaeological Reports 16. Boston: American Schools of Oriental Research. MacDonald, B., Herr, L. G., Quaintance, D. S., Clark, G. A., Hajan, H., and J. Eggler. (2016). The Shammakh to Ayl Archaeological Survey, Southern Jordan (2010–2012). Archaeological Reports 24. Boston: American Schools of Oriental Research. Makarewicz, C. A., and N.B. Goodale. (2004). Results from the First Excavation Season at el-Hemmeh: A Pre-Pottery Neolithic Site in the Wadi el-Hasa, Jordan. Neo-Lithics, 2/04, 5–10. Makarewicz, C., and A. E. Austin. (2006). Late PPNB Occupation of el-Hemmeh: Results of the Third Excavation Season 2006. Neo-Lithics, 2/06, pp. 19–23. Meadows, J. (2001). Arid-Zone Farming in the Fourth Millennium BC: The Plant Remains from Wadi Fidan 4. In: A. Walmsley, ed., Australians Uncovering Ancient Jordan: Fifty Years of Middle Eastern ArchaeologySydney, Australia: The Research Institute for Humanities and Social Sciences, The University of Sydney and The Department of Antiquities of Jordan, pp. 153–64. Mithen, S., and B. Finalyson. (2007). WF16 and the Pre-Pottery Neolithic A of the southern Levant. In: B. Finalyson and S. Mithen, eds., The Early Prehistory of Wadi Faynan, Southern Jordan: Archaeological Survey of Wadis Faynan, Ghwayr and al-Bustan and evaluation of the Pre-Pottery Neolithic A site of WF16. Wadi Faynan Series 1. Levant Supplementary Series 4. Oxford: Council of

NEOLITHIC PERIOD PRESENCE

31

British Research in the Levant and Oxbow Books, pp. 470– 86. Najjar, M., Abu-Dayya, A., Suleiman, E., Weisgerber, G., and A. Hauptmann. (1990). Tell Wadi Faynan: the first Pottery Neolithic Tell in the South of Jordan. ADAJ 34, 27–56. Najjar, M., Hauptmann, A., Weisgerber, G., and H.G. Bachmann. (1995). The Early Bronze Age at Faynan/Wadi ʿAraba: A Period of New Technology in Copper Production. In: K. ʿAmr, F. Zayadine, and M. Zaghloul,eds., Studies in the History and Archaeology of Jordan V: Art and Technology Throughout the Ages. Amman: Department of Antiquities, pp. 519–21. Neeley, M. P. (1992). Lithic Period Sites. In: B. MacDonald, ed., The Southern Ghors and Northeast ʿArabah Archaeological Survey. Sheffield Archaeological Monographs 5. Sheffield: Collis, Department of Archaeology and Prehistory, University of Sheffield, pp. 23–51. Neeley, M. P. (2004). Exploring Prehistoric Land-Use Patterns in the Tafila-Busayra Survey Area. In: B. MacDonald, ed., The Tafila-Busayra Archaeological Survey 1999–2001, West-Central Jordan. Archaeological Reports 8. Boston: American Schools of Oriental Research, pp. 35–46. Nissen, H. J. and H.G. Gebel. (1987). Report on the First Two Seasons of Excavations at Basta (1986–1987). ADAJ 31, 79– 119. Nissen, H. J., Muheisen, M., and H. G. K. Gebel. eds. (2004). Basta I: The Human Ecology. Bibliotheca neolithica Asiae meridionalis et occidentalis & Yarmouk University, Monograph of the Faculty of Archaeology and Anthropology, Volume 4. Berlin: Ex oriente and Yarmouk University. Parker, S. T. (1976). Archaeological Survey of the Limes Arabicus: A Preliminary Report. ADAJ 21, 19–31. Parker, S. T. (1986). Romans and Saracens: A History of the Arabian Frontier. American Schools of Oriental Research Dissertation Series 6. Philadelphia: American Schools of Oriental Research. Parker, S. T., and Smith II, A. M. (2014). The Hinterland of Roman Aila. In: S. T. Parker, and A. M. Smith II, eds., The Roman Aqaba Project Final Report. Volume I: The Regional Environment and

32

BURTON MACDONALD

the Regional Survey. Archaeological Reports Series 19. Boston: American Schools of Oriental Research, pp. 357–77. Peterson, J. (2003). The 1999 Test Excavations at Khirbat alHammam (WHS 149) Wadi al-Hasa, Jordan. ADAJ 47, 117– 28. Peterson, J. (2004). Khirbat Hammam (WHS 149): A Late PrePottery Neolithic B Settlement in the Wadi el-Hasa, Jordan. BASOR 334, 1–17. Peterson, J. and M.P. Neeley. (2007). Crossing the Boundary to the Domestication Economies: A Case Study from West-central Jordan. In: T. E. Levy, P. M. M. Daviau, R. W. Younker, and M. Shaer, eds., Crossing Jordan: North American Contributions to the Archaeology of Jordan. London: Equinox, pp. 203 – 10. Peterson, J., Neeley, M., Hill, B., Jones, J., Crawford, P., Kurzawsia, A., Sullivan, N., Wasse, A., and C. White. (2010). The Origins and Development of Agriculture in the Wadi al-Hasa Region: 2006 Test Excavations at Khirbat al-Hammam (WHS 149), TBAS 102, and TBAS 212. ADAJ 54, pp. 387–412. Ortner, D. and B. Frohlich, eds. (2008). The Early Bronze Age Tombs and Burials of Bab edh-Dhraʿ, Jordan. Reports of the Expedition to the Dead Sea Plain, Jordan Volume 3. Lanham, MD: Altamira. Rast, W. E. and R.T Schaub. (2003). Bab edh-Dhraʿ: Excavations at the Town Site (1975–1981). Part 1: Text. Winona Lake: Eisenbrauns. Rollefson, G. O. (1996). An EPPNB Settlement in the Wadi elHasa, Central Jordan. In: S. K. Kozlowski and H. G. K. Gebel, eds., Neolithic Chipped Stone Industries of the Fertile Crescent, and Their Contemporaries in Adjacent Regions (Proceedings of the Second Workshop on PPN Chipped Lithic Industries. Berlin: ex oriente, pp. 159–60 Rollefson, G. O. (1999). El-Hemmeh: A Late PPNB – PPNC Village in the Wadi el-Hasa, Southern Jordan. Neo-Lithics, 2/99, 6–8. Rollefson, G. O. (2008). The Neolithic Period. In: R. B. Adams,ed., Jordan: An Archaeological Reader. London: Equinox, 71–108

NEOLITHIC PERIOD PRESENCE

33

Rollefson, G. O., and Kafafi, Z. (1985). Khirbat Hammam: A PPNB Village in the Wadi el-Hasa, Southern Jordan. BASOR 258, 63–69. Rollefson, G. O., and I. Kohler-Rollefson. (1989). The Collapse of Early Neolithic Settlements in the Southern Levant. In: I. Hershkovitz, ed., People and Culture in Change: Proceedings of the Second Symposium on Upper Paleolithic, Mesolithic, and Neolithic Populations of Europe and the Eastern Mediterranean Basin. BAR-IS 508. Oxford: British Archaeological Reports, pp. 59–72. Rollefson, G. O., and I. Kohler-Rollefson. (1992). Early Neolithic Exploitation Patterns in the Levant: Cultural Impact on the Environment. Population and Environment, 13, 243–54. Sampson, A. (2010a). Wadi Hamarash: A New MPPNB Site in Jordan. ADAJ 54, 73–84. Sampson, A. (2010b). The Excavation of 2010 at the Wadi Hamarash-Suwayf (Ghawr as-Safi). ADAJ 54, 85–94. Sampson, A. (2011). The 2011 Excavations at Wadi Hamarash 1 and 4, As-Safi, Jordan. ADAJ 55, 73–83. Sampson, A. (2013). Wadi Hamarash 1: An Early PPNB Settlement at Wadi al-Hasa, Jordan. Vol. I. Rhodes: University of the Aegean, Laboratory of Environmental Archaeology. Schaub, R. T., and W.E. Rast. (1989). Bab edh-Dhraʿ: Excavations in the Cemetery Directed by P. W. Lapp (1964 – 67). Winona Lake: Eisenbrauns. Schuldenrein, J. S., and P. Goldberg. (1981). Late Quaternary Paleo-Environments and Prehistoric Site Distribution in the Lower Jordan Valley: A Preliminary Report. Paléorient 7, pp. 57–71. Simmons, A., and M. Najjar. (2006). Ghwair I: A small but complex Neolithic community in southern Jordan. Journal of Field Archaeology 31, 35–39. Simmons, A., and M. Najjar. (2007). Life in the Resort Corridor – Ghwair I, a Small but Elaborate Neolithic Community in Southern Jordan. In: T. E. Levy, P. M. M. Daviau, R. W. Younker,and M. Shaer,eds., Crossing Jordan: North American Contributions to the Archaeology of Jordan. London: Equinox, pp. 231–41

34

BURTON MACDONALD

Tubb, K. (1985). Preliminary Report on the ʿAin Ghazal Statues. Mitteilungen der Deutschen Orient Gesellschaft 117, 117–34. Waheeb, M. (1996). Archaeological Excavations at Ras an-Naqab – ʿAqaba Road Alignment: Preliminary Report (1995). ADAJ 40, 339–48. Waheeb, M., and N. Fino. (1997). ʿAyn el-Jamman: a Neolithic Site Near Ras en-Naqb, Southern Jordan. In: H. G. K. Gebel, Z. Kafafi, and G. O. Rollefson,eds., The Prehistory of Jordan, II. Perspectives from 1997. Studies in Early Near Eastern Production, Subsistence, and Environment 4. Berlin: ex oriente, pp. 215–19. Weninger, B. (2009). Yarmoukian Rubble Slides. Evidence for Early Holocene Rapid Climate Change in Southern Jordan. NeoLithics, 1/09, 5–11.

A REVALUATION OF IRON AGE FORTIFIED SITES ON THE KARAK PLATEAU STEPHANIE H. BROWN INTRODUCTION

Between the years of 1980 and 1989 S. Thomas Parker directed the Limes Arabicus Project (Parker 1987, 2006), which investigated the Roman period occupation of the Karak Plateau in central Jordan. One aspect of Parker’s project was a vehicular survey of the eastern Karak Plateau, which identified sites associated with many different periods of human occupation (Clark 1987). This paper is concerned with the nature of ten Iron Age fortified sites identified by Parker’s team during their survey. Based largely upon Nelson Glueck’s 1930s survey of Eastern Palestine scholars believed for many years that there existed a fortified Moabite frontier, made up of contemporary fortified sites that ran north/south along the eastern edge of the Karak Plateau as part of a larger system of defense against a threat from the eastern desert (Glueck 1939). During the past thirty years some scholars have begun to doubt the validity of this idea. The question remains, however, if the original interpretation is incorrect, what then is the nature and function of these Iron Age fortified sites on the eastern Karak Plateau? The Limes Arabicus Project published evidence from all ten of the fortified sites examined in this paper. Several other scholars have also studied some of these sites, but to date there is not a clear consensus as to their nature and function. The very fact that all the sites are fortified and located atop hills that command excellent views of the surrounding areas, and that several include apparent watchtowers implies that their function is at least partially defense. This analysis, however, will suggest that previous notions of a fortified Moabite frontier that secured the entire Karak Plateau at any 35

36

STEPHANIE H. BROWN

time during the Iron Age is simply incorrect. This paper will examine the ten sites in detail and will offer a more precise date for their construction and occupation. Finally, the paper will attempt to synthesize the material by placing the sites within their respective historical environments and will then put forth several suggestions as to their natures and functions.

RESEARCH DESIGN

In comparison to other regions in Jordan the Karak Plateau has seen little archaeological excavation despite the fact that several surveys have recorded many sites. The ten sites examined in this study were surveyed and published by S. Thomas Parker in his Limes Arabicus Project, however, they were surveyed at a time when there were very few sites that could offer stratified sequences of Iron Age ceramics to aide in the initial dating of the sites. In the subsequent decades several Moabite sites have been or are being excavated, mostly north of the Karak Plateau (e.g., Dhiban, Wadi ath-Thamad, Tell Madaba, etc.) and several regional surveys have reported Iron Age ceramic evidence at various sites in the region, providing more evidence of Moabite ceramic typology. In light of this recent research, this project reexamined most of the ceramic evidence collected by the Limes Arabicus from these sites in an attempt to provide a more precise date for their occupation. There were a few obvious challenges to this research. The first was the present condition of these sites, which are relatively small. Some have suffered dramatically from modern development, in some cases greatly altering the surrounding landscape even in the twenty five years since Parker’s survey. Furthermore, illicit digging and/or stone-robbing has in many cases further damaged the sites’ architectural features. This rapid deterioration was an important factor in the decision to reexamine the sites while it was still possible. A second problem was simply the challenge of interpreting surface ceramic evidence. Nevertheless, reexaminations of the existing surface collections in light of the greatly improved knowledge of Iron Age pottery has suggested the temporal parameters of Iron Age occupation at these sites. These ten fortified sites (Figure 1) were chosen on the basis of their identification by the Limes Arabicus Project as a “fort” and the recovery of a significant number of Iron Age sherds from each site. The ten sites are: Qasr ed-Dabaʿa (Parker’s site 194), Khirbet Tha-

IRON AGE FORTIFIED SITES ON THE KARAK PLATEAU

37

mayil (Parker’s site 456), Qasr Saliya (Parker’s site 2), Qasr Abu elKharaqa (Parker’s site 110), Qasr el-ʿAl (Parker’s site 124), Khirbet ʿArbid (Parker’s site 464), esh-Sharif (Parker’s site 477), Khirbet Medeinet ʿAliya (Parker’s site 149), Parker’s site 7, and Parker’s site 57. Many of these were mentioned in Glueck’s survey as well as Miller’s Archaeological Survey of the Kerak Plateau and a few of them were revisited in the 1990s by Bruce Routledge.

Figure 1. Map of the Iron Age fortified sites surveyed by the Limes Arabicus Project (Used with permission by Dumbarton Oaks Publications)

38

STEPHANIE H. BROWN

Site and LAP Site # (#corresponds with Fig. 1.1) Qasr ed-Daba’a Site # 194 Khirbet Thamayil Site # 456

Iron Age Occupation (Sherd count from LAP) Ir II, 113 sherds Ir II, 107 sherds

Dimensions

Tower Dimensions

Interesting Features

27 x 30 m

9x7m

24.5 x 31.5 m

7.5 x 10 m

Qasr Saliya Site #2

9 Ir, 10 Ir I, 101 Ir II, 178 UD (Ir?) Ir, 31 sherds (8 Ir II sherds from 1976)

19.2 x 25 m

N/A

60 x 68 m

16 x 20 m

Khirbet ‘Arbid Site # 464

Ir II, 66 sherds

47.5 x 50 m

10 x 8 m

Esh-Sharif Site # 477 Khirbet Madayna ‘Aliya Site # 149

Ir, 56 sherds

60 x 60 m

18 x 18 m

Ir I, 116 sherds

275 x 110 m

N/A

Qasr Abu elKharaqa Site # 110

Ir I, 109 sherdss

60 x 60 m

22 x 18 m

LAP Site # 7

Ir II, 18 sherds

31 x 30 m

7x7m

LAP Site # 57

Ir, 61 sherds

31 x 39 m

N/A

Two rooms inside tower Poss. Cisterns inside enclosure Traces of internal rooms and cisterns Traces of interior rooms, cisterns outside and road (?) Rooms inside courtyard, Routledge suggests Islamic Many rooms, cistern outside Many internal buildings (“pillar houses”) and dry moat Rooms against curtain wall and inside tower, cistern outside Inside – rec. building, cistern (?) Elliptical w/traces of interior walls, possible cistern

Qasr el-‘Al Site # 124

Table 1. Ten fortified sites on the eastern Karak Plateau surveyed by the Limes Arabicus Project (LAP)

Several additional challenges accompanied the dating of the ceramic evidence from this site. Though ongoing excavation projects in Moab, such as those taking place at Dhiban, Tall Madaba, the Wadi ath-Thamad, and Mudaybiʾ now permit closer dating of Iron Age pottery (usually called “Moabite”) from this region, no complete typology has been published. Furthermore, the differences between contemporary Moabite ceramics and the better-studied Ammonite ceramics farther north are greater than previously thought. In order

IRON AGE FORTIFIED SITES ON THE KARAK PLATEAU

39

to address these obstacles, the author spent several months studying the ceramic collections from Dhiban, Tell Madaba, and Mudaybiʾ as well as consulting with their respective project directors prior to the examination of the ceramics from the ten sites in this study. Therefore, the dates given for these ten sites reflect the author’s knowledge of this largely unpublished material. Both the Iron I (ca. 1200–1000 BCE) and Iron II (ca. 1000– 539 BCE) periods are represented in the ceramic collection from these sites. The sherd count represented in Table 1 was published in Parker’s final report of the Limes Arabicus Project and includes diagnostic as well as non-diagnostic sherds. The diagnostic sherds were shipped back to the United States while non-diagnostic sherds were discarded after the field reading. Of these ten sites, the author was able to locate and examine the diagnostic sherds for seven. The three sites whose associated ceramic evidence could not be examined are: Parker’s site 57, Qasr el-ʿAl and Qasr Abu el-Karaqa. Because the author was not able to locate the sherds associated with these three sites, no closer dating of the sites can be provided. However, the Limes Arabicus Project offered broad general dates that were consistently accurate. Therefore, it can be assumed that the general dates assigned to the sites can be trusted.

THE KARAK PLATEAU

The Karak Plateau is surrounded to its north, south, and west by sharp, deep cliffs created by the rivers that formed the wadis to the Plateau’s north and south and by the Jordan Valley to its west. These wadis make access to the plateau from these directions extremely difficult. To the north and northeast of the Karak Plateau the Wadi Mujib and its tributaries wrap around the edge of the plateau, creating many smaller wadi systems that form a significant barrier. The southern half of the plateau slopes more gradually eastward from ca. 1200 m (the highest point of the plateau, located in its southwest quadrant) to ca. 650 m (the elevation of the desert plateau), making access to the plateau from the desert relatively simple. However, any traveler attempting to approach the plateau from the southwest would be forced to deal with the difficulties of desert travel, such as lack of vegetation and water, which creates another barrier to the accessibility of the Karak Plateau. The eastern border of the Karak Plateau has two main geographic features: the southern Wadi Mujib/Wadi ed-Dabaʿa drain-

40

STEPHANIE H. BROWN

age system, discussed above, and the Fajj al-ʿUseikir. The Fajj is a large valley, with a wide flat floor that runs northwest from the desert onto the plateau. The Fajj is flanked by high ridges divided by small wadis draining into the Fajj. Several Iron Age sites lie on some of these high points, a few of which will be discussed below.

Figure 2. Map of soil associations on the Karak Plateau (Used with permission by John Wineland)

IRON AGE FORTIFIED SITES ON THE KARAK PLATEAU

Figure 3. Rainfall map of Central Jordan (Image created by Andrew T. Wilson)

41

The Karak Resources Project, under the direction of Gerald Mattingly and John Wineland, conducted an environmental study of the Karak Plateau. One of the most illuminating aspects of this study is an analysis of the agriculturally viable areas of the plateau, which went beyond the typical rainfall map to take into account an analysis of soil quality on the Karak Plateau. The project generated a map (http://www.vkrp.org/studies/environmental/plateau-

42

STEPHANIE H. BROWN

soils/info/soil-associations.asp) (Figure 2) that ranks the local soils with a number 1–12. The number 1 represents the extremely rocky soil found in the wadis while the number 12 is the sandy soil found as one travels east into the desert. The mid-range numbers represent the soil best suited for agriculture. The map shows that all ten sites in this study are located in areas that are associated with soil number 12, indicating that the soil near these sites is extremely sandy and not well suited for agriculture. Jordan’s climate also plays a large role in dictating the best areas for human settlement. Though the Yarmuk, Zerqa, and Jordan Rivers all may be tapped for irrigation, rainfall has the most influence over settlement patterns. Jordan experiences two distinct seasons per year: cold rainy winters and hot dry summers. Rain falls for only a few months during winter. The Karak Plateau averages about 700 mm per annum. However, the amount of rainfall quickly drops off as one moves toward the eastern plateau, where for example el-Lejjūn receives only 200 mm (Koucky 1987: 16). The nomadic and semi-nomadic populations of today and of antiquity have dealt with these rainfall patterns by moving their herds to graze seasonally. Settled agriculturalists, however, were for the most part concentrated in areas of sufficient rainfall Most of the sites examined by this study lie even further east than el-Lejjūn, and receive less than 200 mm of rain each year. A rainfall map of the Karak Plateau (Figure 3) shows that the zone of fortified sites surveyed by the Limes Arabicus Project falls between the 100 and 200 mm rainfall lines. Dry farming usually requires at least 200 mm of rainfall per year. The combination of low rainfall and sandy soil suggests that the land around the fortified sites in this study is quite marginal and could not normally sustain significant agriculture.

THE IRON AGE FORTIFIED SITES OF THE L IMES ARABICUS PROJECT

As mentioned above, in Glueck’s survey of eastern Palestine he reported a line of fortified sites on the eastern border of Moab that he believed were constructed to protect Moab from an eastern threat attested in biblical sources. Three important surveys since Glueck’s have contributed to our knowledge of the ten fortified sites on the eastern Karak Plateau on which this paper focuses: Miller’s Archaeological Survey of the Kerak Plateau, Parker’s Limes Arabi-

IRON AGE FORTIFIED SITES ON THE KARAK PLATEAU

43

cus Project, and Routledge’s 1990s survey that was published in his dissertation, Intermittent Agriculture and the Political Economy of Iron Age Moab. Each survey published evidence on several of these fortified sites. In order to assess the function of the sites, knowledge of the relative dates offered by their associated ceramic evidence is crucial. Because of the limited knowledge in the 1980s of “Moabite” ceramics, the dating of these fortified sites published by Miller and Parker is broadly accurate but somewhat nebulous. For example, most sites are simply labeled as “Iron I” or “Iron II” and some only as “Iron Age.” In the thirty years since the Parker’s and Miller’s surveys, new excavations in central Jordan have contributed to our knowledge of Iron Age ceramics.

IRON AGE SITES Site 57 Parker’s site 57 lies above the east bank of the Wadi Mujib overlooking a possible entrance into the wadi. The site is elliptical, unique among the sites in this study. The dimensions of the basalt structure are ca. 31 x 39 m with an outer wall of two rows of blocks preserved to four courses in places. Inside the west end of the enclosure is a platform (ca. 7 x 4 m, with a base ca. 19 x 24 m) of large blocks. Also, at the northern end of the enclosure are traces of an interior wall. A possible cistern lies just west of the enclosure (Parker 2006: 60). Unfortunately, not much more can be said about this site without further study. The Limes Arabicus Project was unable to date the pottery from this site closer than generically “Iron Age.” Because the associated ceramic evidence has not been located, this site cannot be dated more accurately at present. Iron I Fortified Sites Of the ceramic evidence examined, only Khirbet Medayna ʿAliya (KMA) yielded exclusively Iron I sherds. The other fortified site that the Limes Arabicus Project reported as yielding predominantly Iron I sherds was Qasr Abu el-Kharaqa, whose pottery could not be located, making it impossible to validate that call. However, based on the general accuracy of the Limes Arabicus Project’s pottery calls, it seems reasonable to conclude that Abu al-Kharaqa is also primarily an Iron I site. According to the Limes Arabicus survey

44

STEPHANIE H. BROWN

report, Qasr Saliya also yielded some Iron I pottery; but it may have been represented by non-diagnostic sherds only, because there did not appear to be any Iron I sherds among the ceramic material associated with Qasr Saliya. Khirbet Madayna al-ʿAliya (KMA) KMA is an Iron I fort that sits on a large peninsula protruding into the Wadi Mujib where it joins the Wadi Mukheiris. This site is thought to be a “sister site” to Khirbet al-Mudayna al-Muʾarradja (KMM), whose proximity to KMA (the two sites are located only five kilometers apart), remarkably similar features, and similar names confused early travelers and archaeologists. Following Routledge’s excavations at the site in the mid-1990s, he concluded that KMA was a single period site dating to the second half of the eleventh century BCE (Routledge 2000: 43). The unique location of KMA provides almost 360 degrees of natural defense, as its surrounding slopes are extremely steep. The only vulnerable location is a narrow land bridge that provides access to the site from the plateau to the west. This walled site is of considerable size (ca. 275 x 110 m), and the defensive architecture is unmistakable. The entire site is enclosed; and the portion of the site designed for habitation is surrounded by a casemate wall, over 4 m wide, which was incorporated into the architecture of most of the buildings on the site. At the center of the site exists a large, open court. The inhabitants of the site dug a large dry moat (ca. 35 m long x 19 m wide x 5 m deep) commanded by a massive tower near the land bridge. A 3–4 m wide road preserved for 147.5 m leads to the site along the northern slope of the promontory. At the eastern end of the site is a small gate (2.5 m) that Routledge suggests provided access to three cisterns dug into the steep slopes as well as the wadi bottom below (Routledge 2000: 48–49). The ca. 35–45 domestic structures at KMA are of the so-called “pillared” house variety. The houses were constructed of local, unworked stone, with no use of mudbrick. The construction is simple, without plaster. Many houses take the “four-room” design while others form an “L-shaped” pattern (Routledge 2000: 51). Routledge stresses the self-sufficiency of the site of KMA. Though he readily admits that the striking defensive architecture strongly suggests that defense was a serious concern of the inhabit-

IRON AGE FORTIFIED SITES ON THE KARAK PLATEAU

45

ants, he argues that it was not its only raison d’être. In other words the architecture does not necessarily imply that the primary function of the site and its inhabitants was military. Therefore, one cannot necessarily conclude from the available evidence that there was any sort of regular garrison, clearly defined border, or nonmilitary sites in need of protection associated with this site (Routledge 2000: 56). Routledge also suggests that the site may not be as marginal as it seems. Though the site exists in a very low rainfall zone, there is access to the perennial waters at the bottom of the wadi. The evidence suggests that the inhabitants made use of local resources from the wadi, including freshwater crab, partridge, and timber such as ash and poplar (Routledge 2000: 38). Many of these resources are utilized by local agro-pastoralists living in the area today, who also make use of the rich sediment found in the wadi beds for small-scale agricultural activities (Porter et al. 2014: 134), underscoring the productive capacity of this site despite its environmental marginality. Qasr Abu al-Kharaqa Qasr Abu al-Kharaqa was also interpreted by the Limes Arabicus Project as an Iron I fortified site, though, the author was not able to locate its ceramic evidence. The site is ca. 60 x 60 m with a large tower in its southwest corner that still stands to 21 courses (ca. 8 m). There is a large door in the north wall, and though the LAP survey reported several interior rooms inside the tower, when the author visited the site in 2009 the tower was completely filled with rock tumble. Like KMA, Abu al-Kharaqa has many rooms built against the interior of the curtain wall. Excavation could determine if Abu alKharaqa also had a large central courtyard. Outside of the enclosure to the southeast is a cistern that was still in use in the early 1980s (Parker 2006: 65). The site has a spectacular view to its west and northwest, overlooking the upper drainage of Wadi Mujib. The site is directly across a small wadi and within clear sight of Qasr elʿAl, an Iron II fortified site discussed below. The intervisibility of these two sites may suggest that even though the Limes Arabicus Project’s survey did not report a single Iron II sherd, Qasr Abu alKharaqa may have been reused during the Iron II period (Parker 2006: 66).

46

STEPHANIE H. BROWN

Iron II Fortified Sites Of the ten sites examined in this study five can be definitively labeled as Iron II and two more (Qasr el-ʿAl and Parker’s site 7) can be tentatively called Iron II. The Limes Arabicus Project dated Parker’s site 7 to Iron II and the author is inclined to accept that reading based on the general reliability of the other readings. However, what remains of the diagnostic ceramic evidence is scant (only six sherds and only one rim); and these sherds are quite eroded. These sherds are clearly of Iron Age ware but no closer dating seems possible. The pottery from Qasr el-ʿAl was called simply “Iron” by the Limes Arabicus Project but the report mentions an earlier collection in 1976 that recovered eight Iron II sherds (Parker 2006: 67). Qasr el-ʿAl Qasr el-ʿAl is a large (ca. 60 x 68 m) fortified site dating to the Iron II period. Unfortunately, the ceramic evidence from the site could not be found, precluding close dating within the Iron II period. Qasr el-ʿAl lies east of the Wadi Mujib, overlooking the upper drainage of the same wadi. Its view extends as far as 20 km to the southwest to include Khirbet el-Fityan, a Roman fort built on top of a supposed Iron Age fort (Parker 1987: 429–46). Glueck considered this site part of his “chain of Moabite fortresses” (Glueck 1939: 101–03). Qasr Abu al-Kharaqa is clearly visible from Qasr elʿAl and construction of the two sites is very similar. Qasr el-ʿAl has a large tower (ca. 16 x 20 m) that still stands to 6 m in some places. The wall enclosing Qasr el-ʿAl is ca. 1.5 m thick, and there is evidence of rooms built against the enclosure wall. Some of these rooms are well preserved, and one still retains its lintel in place above the doorway. The site also includes many cisterns and caves. Unfortunately, the site has experienced a great deal of modern disturbance and looting, and even the many modern burials inside the rock tumble of the tower have been disturbed. Esh-Sharif The site of esh-Sharif (Parker’s site 477) corresponds to Miller’s site 238, though Miller’s site 238 includes Parker’s site 476 (a cistern that likely provided water to site 477) and 478 (a robbed structure of uncertain function) in addition to site 477 (the fortified site of esh-Sharif) (Parker 2006: 101). Esh-Sharif is the southernmost

IRON AGE FORTIFIED SITES ON THE KARAK PLATEAU

47

site in this study, and it sits atop a large hill that overlooks the northwest portion of the Fajj el-ʿUseikhir, the deep and wide valley that cuts through the hills and wadis of the southeastern Karak Plateau and provides easy entrance to the plateau from the eastern desert. The site offers extensive views in all directions and includes many ancient sites that line along both sides of the Fajj. Glueck called this site, “Qaseir et-Tamrah.” It is enclosed by a wall ca. 2 m thick and up to 3 m high and encompasses an area ca. 60 x 60 m, and the site includes a central structure, possibly a tower (ca. 18–20 x 18–20 m). The Limes Arabicus Project reports that the central structure was at least two stories high and contained many internal rooms (Parker 2006: 101). When reexamining the site the internal rooms were obvious. However, there was less evidence of a second story, perhaps because the structure suffered further damage in the almost thirty years between the Limes Arabicus Project and the author’s visit in 2009. Outside the central structure is evidence of other buildings and courtyards, including a rectangular two-roomed building northwest of the enclosure in which the rooms are connected by a doorway (Miller 1991: 102). Both the Limes Arabicus Project Survey and Miller’s Archaeological Survey of the Kerak Plateau date the ceramics from esh-Sharif as generically Iron Age. After re-examining the ceramics from eshSharif collected by the Limes Arabicus Project, the author concluded that this site can safely be dated to the Iron II period. A closer date is slightly more problematic because of the lack of comparanda, but perhaps a date during the Iron IIB period (900–700 BCE) would be appropriate. The ceramic forms appear most similar to those at Dhiban, though the ware is distinctly more coarse and filled with basalt inclusions, typical of Iron II pottery on the eastern Karak Plateau. Site 7 The author was also unable to locate the pottery collected from Parker’s site 7; however, the Limes Arabicus Project dated the Iron Age pottery to the Iron II period. The site is a relatively simple rectangular enclosure (ca. 31 x 30 m), and was constructed for defense with two rows of large stones making up its exterior wall. Inside the enclosure there is a square tower (ca. 7 x 7 m) preserved to six courses in the northeast corner, a rectangular building (ca. 5 x 12 m), and a possible rock-cut cistern under the enclosure. This forti-

48

STEPHANIE H. BROWN

fied site overlooks a road crossing through Wadi es-Su’eida to the south and another wadi to the north (Parker 2006: 55). Qasr Saliya The construction of Qasr Saliya can be safely dated to the Iron II period, though it should be noted that the Limes Arabicus Project found a small number of Iron I sherds, implying perhaps that the Iron II occupation was secondary. The ceramics from Qasr Saliya seem to fit neatly into the Iron IIB range, likely the eighth century. Qasr Saliya is a relatively small site (ca. 19.2 x 25 m) located northeast of the Mujib on a large hill that offers outstanding views to the north, east, and south. The site is constructed from large dry-laid stones but lacks a tower. There is a possible entrance ramp on the east side of the site and traces of many rooms inside the enclosure (Parker 2006: 55). Outside the enclosure to the south and west are many cisterns and wall traces, perhaps suggesting a small settlement associated with this site. Because the Limes Arabicus Project treated both the enclosure and extramural settlement as the same site, it is impossible to know from whence the Iron I sherds were collected. It is possible that the settlement outside of the fortification was predominately used during the Iron I period but not reoccupied during the Iron II period. Only further survey or excavation can provide clarification. Qasr ed-Dabaʿa Qasr ed-Dabaʿa lies just south of the modern Karak-Qatrana highway and offers a commanding view of Wadi ed-Dabaʿa where the modern road crosses the wadi. This was a likely crossing point in antiquity as the wadi is very steep and treacherous elsewhere. The pottery assemblage from Qasr ed-Dabaʿa is homogenous in nature and quite similar to the pottery from Khirbet ʿArbid and Khirbet Thamayil. The site itself is ca. 28 x 27.2 m with a central tower (9.2 x 7.2 m). The walls of the tower are ca. 1.5 m thick and the interior space is divided into two rectangular rooms. There seems to be a small opening in the enclosure wall at the eastern corner (Routledge 1996: 120). Also within the enclosure is a cistern filled completely with rock tumble and debris (Parker 2006: 74).

IRON AGE FORTIFIED SITES ON THE KARAK PLATEAU

49

Khirbet Thamayil Khirbet Thamayil lies about five kilometers southwest of Qasr edDabaʿa on a small promontory that overlooks a wadi system to its east and southeast. Routledge, Parker, and Miller all surveyed this site, and Routledge did several probes. Thamayil is a fortified tower site (ca. 36.7 x 27 m). The exterior wall and the interior tower (ca. 7.5 x 10 m) are constructed from large uncut boulders of limestone and basalt, laid in two rows about 1m wide (Parker 2006: 99). The tower still stands to ca. 5 m in places and is built against the southwest side of the inner-most of two rectangular enclosure walls (the inner enclosure measures ca. 26 x 16.5 m). This inner wall was discerned by Routledge’s more intensive survey. A subsequent test probe (1 x 12 m) did not offer very much stratigraphic information (Routledge 1996: 99–104). Outside the enclosure is a complex of terraces measuring ca. 80 x 55 m. On the southeastern side of the complex is a secondary terrace (ca. 23 x 18 m). Several depressions in the northwest quadrant of the complex may be filled cisterns, and Miller reported a slanted rampart about 1 m high, built of fieldstones (Miller 1991: 106). Routledge also dug a test probe (1 x 4 m) to examine a walled rectangular space about 14 m southeast of the outermost enclosure. The probe exposed a stratigraphic section 5.6 m in height. Here Routledge found a walled rectangular enclosure (ca. 9 x 10 m) with a double wall on the southwest side. He also noted in the balk two flat-lying limestone slabs, just above the bedrock, superimposed on each other, and the fragments of four basalt artifacts were associated with this area. Routledge suggests that this area may have been used for food processing (Routledge 1996: 104–06). Khirbet ʿArbid Like Khirbet Thamayil, Khirbet ʿArbid was surveyed by Parker, Miller, and Routledge; and Routledge also conducted several probes at the site. It sits atop a low hill overlooking Wadi edDakakin to its south. To the north and south the site is bounded by small wadis with evidence of terracing. The structure itself (47.5 x 50 m) encloses a small tower (10 x 8.5 m at the base) preserved up to six courses (2.5 m in height). The construction of both the tower and the outer enclosure is of huge, roughly cut limestone slabs and smaller uncut stones (mostly basalt) (Parker 2006: 100). The tower

50

STEPHANIE H. BROWN

sits atop a sort of terrace (15 x 15m) preserved to 7 courses on its southern side (Miller 1991: 99). Outside the enclosure described above is an even larger enclosure (45 x 43 x 44.3 x 51m) containing numerous interior walls and rooms. All three surveys found large amounts of Late Islamic, or Ottoman period, pottery, along with Iron IIC ceramics. Most of the Iron Age sherds came from the enclosed area containing the tower (Routledge’s area A). Therefore Routledge opened a test probe in this area aiming to uncover stratified Iron Age remains. However, this and several other test pits led him to conclude that the architectural remains must be associated with the Late Islamic period. He does insist, however, that the large amount of Iron IIC ceramic evidence strongly suggests an occupation of the site in this period, though its nature cannot be determined (Routledge 1996: 106–14).

INTERPRETATION

The Iron Age fortified sites identified by the Limes Arabicus Project both complement and contribute to our knowledge of settlement trends on Karak Plateau. The Iron I sites discussed above, KMA and Qasr Abu al-Kharaqa, as well as other Iron I sites in Moab, such as Khirbet Madayna al-Muʾarradja (KMM), Baluʾ, Khirbet alMudayna ala al-Mujib, and Lahun, are examples of large nucleated sites that seem to typify the nature of settlement on the Karak Plateau during the late Iron I period (Routledge 1996: 74). This period marks the beginnings of Iron Age settlement on the Karak Plateau, and the large sites of the Iron I period, allowed for many people to live together inside the protection of the sites’ fortified walls. Based on the locations of the large fortified Iron Age sites of the Karak Plateau, which sit at the edges of large wadi systems with expansive views, and the quality of their surrounding lands, they seem to have been constructed for their strategic military value rather than agricultural intensification. Though the fortified nature of these sites would have benefited their inhabitants by protecting their homes and property, it would have limited the agro-pastoral activities in which the occupants could have engaged. Following the Iron I period, there appears to be a break in settlement on the Karak Plateau, that is followed by two different settlement phases in during the Iron II period: one in the Iron IIB period, which would correspond historically to the period of Mo-

IRON AGE FORTIFIED SITES ON THE KARAK PLATEAU

51

abite state formation and rule; and a second in the Iron IIC period, which in central Jordan corresponds historically with increased Neo-Assyrian involvement and then abandonment. The Iron II settlement on the Karak Plateau seems to be more dispersed, and three general trends about the settlement can be observed. First, there is more Iron II than Iron I ceramic evidence on the Karak Plateau. Second, most of the Iron II pottery can be dated to the Iron IIC period (700–539 BCE). And third, the Iron II period reflects an increase in what Routledge calls, “single-unit” sites (Routledge 1996: 75–89). Both the Limes Arabicus Project and Miller’s Survey of the Karak Plateau project report a larger number of these isolated, “single-unit” sites during the Iron II period in significant contrast to the nucleated compounds of the Iron I period. Though most Iron II sites on the eastern Karak Plateau are “singleunit” sites, there are a few exceptions: for example, the large fortified complexes of Mhai (Miller’s site 436), Mudaybiʾ, the esh-Sharif ridge (Miller’s site 238), and Qasr el-ʿAl (Parker’s site 124). The resumption of settlement on the Karak Plateau during the Iron IIB period corresponds to the period of statehood initiatives in Moab as reported in the well-known Mesha inscription, which was followed by the period of Assyrian domination of Moab. The tendency away from nucleation during the Iron IIB-C period likely implies that the eastern Karak Plateau was more secure during this period, as one might expect during a period when this area fell under first Moabite and then Assyrian control. During the Iron IIB period the inhabitants of the Karak Plateau could have lived outside the walls of a fortified site, either in tents or in small domestic dwellings, perhaps with their own land. With a few fortified sites such as Qasr Saliya and esh-Sharif, which date to the Iron IIB period, to protect the plateau as well as the agro-pastoralists that inhabited the eastern fringe, the entire area would have been more secure. There is also a significant settlement shift in the Iron IIC, or the seventh century, when several small fortified sites were constructed: Khirbet Thamayil, Khirbet ʿArbid, and Qasr ed-Dabaʿa (all examined by this study), and also many other small single-unit sites. The Iron IIC period on the eastern Karak Plateau corresponds with a period of increased Neo-Assyrian involvement in the Southern Levant, and then a brief period of independence experi-

52

STEPHANIE H. BROWN

enced by the Moabites between Assurbanipal’s withdrawal from Moab and the fall of Nineveh to the Neo-Babylonians.

DISCUSSION

The fortified sites in this study fall within several different settlement phases on the Karak Plateau, and it is therefore likely that they were built to serve different functions. This discussion will focus on the possible functions of the Iron II fortified sites that make up the bulk of this study. These sites have traditionally been called forts, and in many ways this is understandable, considering that they are enclosed by substantial curtain walls constructed of megalithic boulders, some as casemates. All command good views of the surrounding landscape and usually are intervisible with other apparently contemporary sites. Almost all lie at the highest, most defendable location in their region. Clearly these sites were built with a defensive purpose, but the question remains – what were they defending and from whom? A notable commonality is the number of sites that are ca. 30 x 30 m. Five of the seven Iron II sites have approximately these dimensions. An enclosure of 30 x 30 m seems too small to be called a fort, perhaps “fortlet” would be a more appropriate term. Some of the towers were divided into small rooms, and there is often some evidence of internal walls which may suggest more rooms outside the tower. These small fortified sites would have been big enough to provide shelter for a handful of people, but not big enough to provide permanent or semi-permanent residence for a large military garrison. However, several sites appear to have had cisterns; and there is evidence of terracing near many of these sites. Three of these small sites can be safely dated to the Iron IIC period: Khirbet ʿArbid, Khirbet Thamayil, and Qasr ed-Dabaʿa. One indicator is the presence at all three sites of “rectangular-rimmed cooking pots,” (Figure 4) a form that seems unique to Moab and is attested both from tombs at Dhiban and a sixth century context at Baluʾa. All three sites were intensely surveyed by Routledge in the mid-1990s. Khirbet Thamayil and Khirbet ʿArbid were also surveyed by Miller. Khirbet ʿArbid is slightly bigger (47.5 x 50 m), but following his limited excavation at the site, Routledge posits that a Late Islamic occupation of Khirbet ʿArbid has completely obscured the original Iron Age architecture.

IRON AGE FORTIFIED SITES ON THE KARAK PLATEAU

Figure 4. Rectangular rimmed cooking pots (plus one Iron IIC krater) from Khirbet Thamayil, Qasr ed-Dabba’a, and Khirbet ‘Arbid (Image created by Stephanie H. Brown)

53

The relatively small size of these Iron II fortified sites causes one to wonder about their purpose. Glueck suggested that they were a part of a larger system of border forts that guarded the eastern border of Moab (as well as Ammon and Edom) from the nomadic Arabs in the desert, and both Miller and Parker seem to accept this explanation. The sheer expense of building and manning the fortified sites on the eastern Karak Plateau speaks to the likelihood of a serious eastern threat. There is textual evidence of such a nomadic threat to Moab in both the Hebrew Bible as well as in Assyrian sources. Ezekiel 25 says, “Therefore, I will lay open the flank of Moab from the towns on its frontier, the glory of the country, Beth-jeshimoth, Baal-meon, and Kiriathaim. I will give it along with Ammon to the people of the east as a possession” (Ezekiel 25: 9–10). Though several Assyrian inscriptions reference an eastern threat to Moab by the Qedarite Arabs, perhaps the most illuminat-

54

STEPHANIE H. BROWN

ing source is the text of the Rassam Cylinder, in which Assurbanipal records a number of victories against Yuhaithiʾ the Qedarite. He persuaded the inhabitants of Arabia (to join) him and then plundered repeatedly those peoples which Ashur, Ishtar, and the (other) great gods had given to me to be their shepherd and had entrusted into my hands. Upon the oracle-command of Ashur and Ishtar, (I called up) my army and defeated him in bloody battles inflicting countless routs on him (to wit) in the giru of the towns of Azaril (and) Hirata (-) kasaia, in Edom, in the pass of Iabrudu, in Beth-Ammon, in the district of Haurina, in Moab, in Sa’arri, in Harge, in all the inhabitants of Arabia who had revolted with him, but he escaped… (Luckenbill 1927: 314).

There is clear evidence in the southern Levant, from the Negev, for forts built under the administration of the Assyrian Empire; Arad, for example, is tied directly to the royal administration by official letters from the king (Aharoni 1981). Though there is no such documentary evidence to support the possibility that any of the Iron II forts in Moab were tied to the royal administration of the Assyrian Empire, there are perhaps three excavated fortified sites whose size and archaeological remains may suggest significant “state” involvement, whether by the Assyrians or the Moabites themselves. The best example is found at Khirbet Mudaybiʾ, a fortified site which measures 83.5 x 88.75 m, making it larger than any of the Iron II sites from the northern part of the eastern Karak Plateau. Research at the site has revealed a late Iron II fortified settlement with a four-chambered gate and at least five so-called “proto-Aeolic capitals,” (Drinkard 2002) probably from the era of Assyrian domination of Moab (Mattingly and Pace 2007: 153–159). The presence of the proto-Aeolic capitals, powerful emblems that are found throughout the Southern Levant tied to royal architecture, such as palaces and gates, almost insists that the site had some sort of involvement with a central authority. The other two sites that such attachment to a centralized power structure, Aroer and Lahun, lie just north of Wadi Mujib on the Dhiban Plateau. At both of these sites an Iron II fort was built above an older Iron I fortified village, but in both instances, the Iron II occupation was much smaller than the Iron I village and

IRON AGE FORTIFIED SITES ON THE KARAK PLATEAU

55

seems to fall into Routledge’s “single-unit” site category. The size of these sites suggests that they were built by a central authority, likely the Moabite state, and this suggestion is supported by the mention of Aroʾer’s construction in the Mesha inscription, attributed to Mesha himself. Routledge emphasizes an agricultural role in association with some of the smaller Iron II defended sites. He cites a study of watchtowers and fortifications near Amman that suggests their use was primarily for agricultural purposes based on three factors: the physical locations of the sites; associated installations; and similar structures from the recent past (Routledge 2004: 195). He claims that many of the watchtowers near Amman were not in good strategic military locations (i.e., not on hilltops) but rather were on hillsides overlooking arable wadis and did not provide panoramic views of surrounding areas. He also discusses the Greek, Akkadian, and Hebrew words for “tower” (pyrgos, dimtu, migdal) which in all languages include an agricultural building near cultivated land (Routledge 2004: 195–99). The Iron Age fortlets surveyed by the Limes Arabicus Project do not fit easily into the model created for the Amman fortifications, mainly because the former are located in agriculturally marginal areas in strategic positions atop hills that afford commanding views, often including nearby towers, and are rarely associated with any other agricultural installations. However, these fortified sites often provide access to shallow tributary wadis and lie within marginally cultivatable land. Furthermore, the largest clusterings on the eastern Karak Plateau occur on the south side of the Fajj and around Khirbet Thamayil, both areas that provide easy access to wadis. Routledge claims that the clustering of these and other Iron Age sites on the Karak Plateau would be poor military planning by the Moabites if their nature was purely defensive (Routledge 2004: 200). He also claims that because of the limited agricultural potential of the eastern edge of the Karak Plateau many fortified sites likely supported a pastoral lifestyle along with the agriculture practiced by the Moabites. These fortified sites could have secured livestock feed (Routledge 2004: 200–01). When discussing agriculture one must keep in mind the type of agriculture that can be practiced on the eastern Karak Plateau. As explained in the introduction, the line of the ten fortified sites surveyed by the Limes Arabicus Project is located between the 100

56

STEPHANIE H. BROWN

and 200mm rainfall lines; and in order to dry farm one usually needs at least 200 mm of rainfall per year. Furthermore, the Virtual Karak Resources Project’s soil analysis of the Karak Plateau, indicated that all of the sites examined in this share the same type of extremely sandy soil, which suggests that this area could not sustain much agriculture. As a result, one must consider the role that pastoralism could have played in the region. Here, the position of the fortified sites would make sense, as this marginal land was and still is suitable for pastoralists to graze their herds. While the link between agricultural activities and Iron Age fortified sites is an important one, the Iron Age fortified sites surveyed by the Limes Arabicus Project seem to have been primarily defensive in nature, as they appear to guard the eastern edge of the Karak Plateau. It seems fairly obvious that those responsible for the construction of these sites would select defendable locations that were also conducive to cultivation, especially since they were occupying such a marginal area. The open space within the defended enclosures as well as the many cisterns at these sites suggest the possibility that domestic herds might have sheltered there in times of insecurity Additional evidence that may support the important role of pastoralism on the eastern edge of the Karak Plateau is that all recent surveys of the desert fringe have recorded numerous so-called “ring sites,” the associated artifacts of which date from late prehistoric periods through the modern period. Today these “ring sites” are often built to fence in sheep/goats and it seems quite possible that they may have served a similar function in antiquity. The small Iron IIC fortlets could have provided security for local pastoralists. The view offered from the tower would have given a shepherd ample opportunity to get his herd inside the fortlet if there was a threat. Furthermore, the tower could have, as Routledge suggests, been used to store fodder as a sort of silo. Another element relative to the position and function of the fortified sites on the eastern Karak Plateau is ʿAin Lejjūn, a perennial water source that today serves the local population near the desert fringe. Overlooking ʿAin Lejjūn and just northwest of elLejjūn is the Roman fort of Khirbet el-Fityan, excavated by the Limes Arabicus Project. Khirbet el-Fityan lies on the north bank of Wadi el-Lejjūn and has a panoramic view of the eastern desert. The four-sided fort is protected by a 1.8 m thick curtain wall. The inter-

IRON AGE FORTIFIED SITES ON THE KARAK PLATEAU

57

nal structures that were excavated date to the Late Roman period, however, excavation revealed that at least one side of the Roman castellum’s curtain wall (D.4:002) was built atop an Iron Age wall, possibly suggesting an earlier Iron Age fort there guarding the spring (Parker 1987: 429–46). Furthermore, Parker’s “signaling experiment,” which demonstrated the inter-visibility of a group of fourteen Roman forts and watchtower sites, suggests that Khirbet el-Fityan was the key hub among surrounding forts, of which many had associated Iron Age evidence. This seems to underscore the strategic position of Khirbet el-Fityan both as a strategic vantage point and as controlling the area’s main perennial water source, beneficial for agriculturalists and pastoralists alike. Another possible function of these sites relates to trade and the protection of trade routes. It has long been assumed that the via nova Traiana, constructed in CE 111–114, along an earlier Nabataean trade route, had in turn followed an Iron Age trade route, perhaps even the drk hmlk referenced in the Hebrew Bible (Num. 21:22). Although this belief is sufficiently widespread to provide the modern popular name of “The King’s Highway,” there is no direct evidence to suggest that the via nova Traiana was in fact the major north-south trade route through Jordan during the Iron Age. In fact, as anyone who has travelled on the modern “King’s Highway” knows, it is a much longer and more difficult way of northsouth travel through Jordan. This is because the via nova Traiana traverses several deep wadis, especially Wadi Mujib. If one just goes a few kilometers east and travels along the desert fringe (where the fortified sites in this study lie) travel is much easier. This does not deny use of the route later taken by the via nova Traiana in the Iron Age, which runs through the agricultural heartland of the region, connects several major Iron Age sites such as Karak and Dhiban, and is likely the “road through the Arnon” that is referenced in the Mesha Inscription. However, a north-south route a bit further east would avoid the deep Wadi Mujib as did a secondary Roman route and the Late Islamic Hajj road centuries later. A route farther to the east would better explain the location of Wadi ath-Thamad, an Iron IIC industrial textile site which would have conducted a great deal of trading, well east of the “King’s Highway.” There would have been two major disadvantages to the eastern route. First, the route has fewer water sources, which would have made sites with access to water extremely important. And second, it was much

58

STEPHANIE H. BROWN

more exposed to raids from the desert, which may explain in part the chain of later Roman and Late Islamic forts along its line, as well as the existence of the Iron II fortified sites. The last question to be addressed is who built these Iron II fortified sites and against whom were they defending? As mentioned earlier, there are several references to an eastern threat to Moab in both the Hebrew Bible as well as contemporary Assyrian sources including Ezekiel 25:9–10 and perhaps the most illuminating source, the Rassam Cylinder of Assurbanipal. The Rassam Cylinder is enlightening because it not only demonstrates a clear threat to Moab (and several neighboring states) and names the threat as the people of Arabia (or the Qedarite Arabs); but it implies that Assyria, at least under Ashurbanipal (ca. 668–627 BCE), considered itself to have divine responsibility to protect its western client kingdoms. When political conflicts in southern Mesopotamia caused Ashurbanipal to withdraw his troops from the western empire sometime after ca. 640 BCE, the Karak Plateau was left vulnerable. Therefore, the date of some of these Iron II fortified sites might suggest either construction when Moab was under Assyrian tutelage or as a local reaction after the retreat of the Assyrian Empire. The possibility of any sort of central authority, either an independent Moabite state or one responding to imperial pressure to protect the empire’s frontier, was behind the construction of these sites is debatable. The semi-standardization of the architectural plans of the 30 x 30 m sites might support that view, and while the fortlets would have been too small to serve as strategic military bases for a large garrison, they could have served as small outposts housing a small group of soldiers. Regrettably there is no evidence as to the state of Moab after Assyria’s retreat. We do know from the Mesha Inscription and from Assyrian sources that Moab had an army, but the state of this army in the seventh century is unknown. Furthermore, the size of the Iron II fortlets, as well as their proximity to each other makes them ideal for local families or militia to watch over their homes, water sources, herds, and fields against the raid of small bands of Qedarite Arabs, with or without the aid of a central authority. The larger sites (ca. 60 x 60 m; e.g., Qasr Saliya and esh-Sharif), however, especially those that may date to the eighth and/or seventh centuries, could have served as military bases built by a central authority, perhaps even with Assyrian aid.

IRON AGE FORTIFIED SITES ON THE KARAK PLATEAU

59

CONCLUSION

As evident from the information presented in here, the ten fortified sites examined in this study vary greatly in size, architecture, and date of occupation. Furthermore, some of these fortified sites may have been state-sponsored, either by the Moabite state or with assistance from the Assyrian Empire. Finally, the reexamination of the ceramic evidence from these sites clearly indicates that the sites were occupied at different points during both the Iron I and Iron II periods. Therefore, all these sites together never formed a fortified border that protected the northern half of the Karak Plateau at any time during the Iron Age. Though Glueck’s grand view concerning the fortified sites was perhaps over-simplified, his general explanation about their purpose may have been correct. It does seem that groups of Qedarite Arabs posed a threat to the Karak Plateau; however, Glueck’s assumption that all these sites were a part of an single system of defense built by a powerful Moabite kingdom is based on the assumption that all the Iron II sites were contemporary, which we now know to be false. In the seventy years since Glueck published his survey more emphasis has been placed on the actions and agency of the local, indigenous populations of the Southern Levant, which undoubtedly played a larger role in the creation of the Southern Levantine archaeological landscape then attributed to them by traditional historical sources. Therefore, it is necessary to consider local agency in the construction of the Iron Age fortified sites in this study, especially the fortlets, which were small enough to have been constructed without any type of institutionalized labor force. Even without the sanction of a “state” the local seminomadic or sedentary population of the Karak Plateau needed to maintain their security and protect their limited but valuable resources from an eastern threat. Whether intentionally or not, the efforts of these individuals worked to protect the frontier of the Moabite state as well.

BIBLIOGRAPHY

Aharoni, Y. (1981). Arad Inscriptions. Jerusalem: Israel Exploration Society. Clark, V.A. (1987) The Desert Survey. In: S.T. Parker, ed., The Roman Frontier in Central Jordan. Interim Report on the Limes Arabicus

60

STEPHANIE H. BROWN

Project, 1980–1985, BAR-IS 340, Oxford: Archaeopress, pp. 107–63. Drinkard, J. (2002). The Volute Capitals of Mudaybi’, Virtual Karak Resoures Project, www.vkrp.org/studies/historical/capitals/ info/capitals-Mudaybi.asp Glueck, N. ( 1939). Explorations in Eastern Palestine III. New Haven: American Schools of Oriental Research. Koucky, FL. (1987). The Regional Environment. In: S.T. Parker, ed., The Roman Frontier in Central Jordan. Interim Report on the Limes Arabicus Project, 1980–1985. BAR-IS 340. Oxford: Archaeopress, pp. 11–40. Luckenbill, D.D. (1927). Ancient Records of Assyria and Babylonia Volume. 2. Chicago: University of Chicago Press. Mattingly, G.L. and J.H. Pace. (2007). Crossing Jordan: By Way of the Karak Plateau. In: T.E. Levy, P.M.M. Daviau, R.W. Younker, and M. Shaer, eds, Crossing Jordan: North American Contributions to the Archaeology of Jordan. London and Oakville, CT: Equinox Publishing Ltd, pp. 153–59 Miller, J.M. ed. (1991). Archaeological Survey of the Kerak Plateau. American Schools of Oriental Research Archaeological Reports, 1. Atlanta: Scholars Press. Parker, S.T. ed. (1987). The Roman Frontier in Central Jordan. Interim Report on the Limes Arabicus Project, 1980–1985. BAR-IS, 340. Oxford: Archaeopress. Parker, S.T. ed., (2006). The Roman Frontier in Central Jordan. Final Report on the Limes Arabicus Project, 1980–1989. Washington D.C.: Dumbarton Oaks. Porter, B.W., B.E. Routledge, E.M. Simmons, and J.S.E. Lev-Tov. (2014). Extensification in a Mediterranean Semi-Arid Marginal Zone: An Archaeological Case Study from Early Iron Age Jordan’s Eastern Karak Plateau. Journal of Arid Environments 104, 132–48. Routledge, B. (1996). Intermittent Agriculture and the Political Economy of Iron Age Moab. (unpubl. diss., University of Toronto). Routledge, B. (2000). Seeing through Walls: Interpreting Iron Age I Architecture at Khirbat al-Mudayna al ʿAliya. BASOR 319, 37–70.

IRON AGE FORTIFIED SITES ON THE KARAK PLATEAU

61

Routledge, B. (2004). Moab in the Iron Age: Hegemony, Polity, Archaeology. Philadelphia: University of Pennsylvania Press. Virtual Karak Resources Project: Environmental Study. www.vkrp.org

THE ROADS LEADING TO ZOARA IN THE ROMAN AND BYZANTINE PERIODS CHAIM BEN-DAVID 1

Ancient Zoara (modern Ghor es-Safi), on the southeastern side of the Dead Sea, was one of the major transit hubs between the Roman province of Palaestina to Arabia. This paper seeks to understand the road network which connected Zoara in the Roman and Byzantine periods to these two provinces. The paper describes the three ancient routes climbing up from Zoara to the Moabite and Edomite plateaus and the routes coming down from Judea and the Negev and then emphasizes the importance of Zoara as a major junction in the Roman and Byzantine period. Apart from its special geographical setting on the south-eastern side of the Dead sea, Zoara was an important date growing area as can be learned from the Babatha archive, the site of a Roman army garrison, and in the Byzantine period a potential Christian pilgrimage site – Lot’s monastery, long known from the Madaba map, which has been discovered and excavated in the last decades.

THE ANCIENT ROUTES FROM SOUTHERN MOAB AND NORTHERN EDOM TO ZOARA

For many years, the common view in all studies concerning the Roman road system east of the Dead Sea was that only one road, This article is dedicated to Tom Parker, a great scholar and a good friend, and wishing him many years of good health and fruitful research. 1

63

64

CHAIM BEN-DAVID

the Luhit ascent, ascended from the southeastern end of the Dead Sea to the Moabite plateau (Mittmann 1982; Roll 1994, 1999; Dearman 1997). However, the recent discovery of the Zoar Ascent, a well-built Roman road running straight from Zoara to the Moabite plateau, demonstrates the importance of Zoara as a major junction on the Roman road network in the southern Levant. Further investigations show that late-nineteenth and early twentieth-century travelers were acquainted with two additional ancient routes ascending from Zoara, both with Roman watchtowers and one with clear traces of an another, previously unknown, built Roman road. Any commercial or military link between Zoara and northern Edom or southern Moab had to overcome cliffs and steep slopes, some as high as c. 1,500 m above sea level. The major wadis that drain the mountainous plateau of northern Edom and southern Moab into the Dead Sea valley, such as the Wadi al-Hasa, contain many segments that are difficult to traverse, especially for animal caravans. Among them are narrow canyons with high, dry waterfalls or segments blocked by rockslides of large boulders. The logic behind road construction in the region mandates maximum avoidance of valley floors because of their frequently changing forms due to flooding. Therefore, trails descending to Zoara should be sought on spurs between river valleys. The descent from the upper portions of the Moabite mountains would be by passages found through marine sedimentary rock, which is relatively easy to traverse, with the main challenge in finding passage through the steep sandstone cliffs, and the sometimes even harder-to-traverse igneous rock. In all past studies concerning the Roman road system east of the Dead Sea (Mittmann 1982; Roll 1994, 1999; Dearman 1997), only one ancient road was marked as ascending from the southeastern end of the Dead Sea to the Moabite plateau. This road proceeds north along the Dead Sea coastline from Zoara for c. 12 km, and then climbs eastward south of Wadi Isal. A built Roman road, it was first documented at the beginning of the twentieth century by Musil (1907:167–70) and by Nelson Glueck (1939:148). More recently, it was also documented by Mittmann (1982) and Jacobs (1983). Mittmann proposed that this was the biblical “Ascent of Luhith” (Isa. 15:5); however, in my opinion (Ben David 2001) this remains unproven.

ROADS LEADING TO ZOARA

65

During an excursion in 1998 east of the Dead Sea, impressive remains of a hitherto unknown, well-built Roman road were discovered on the northern bank of Wadi Hasa, west of Khanzireh, descending as far as Safi-Zoara and then ascending from Zoara to the Moabite plateau. 2 The discovery of this road has important implications for the understanding of the road system in the area and makes an interesting contribution to the understanding of the Madaba map. The route itself is known today by the Bedouin as Naqb Sarmuj, and is marked as a path on topographical maps. It is mentioned by Musil (1907:63–67) and appears on Turkish and British maps from the period of World War 1. The road connecting the Safi-Zoara area to the Moabite plateau (Fig. 1) has to overcome a vertical elevation difference of some 1,500 m across a linear distance of less than 10 km. On its way, the road crosses geological formations typical of southern Jordan: igneous rock at the bottom, sandstone cliffs in the middle and steep limestone slopes and a plateau at the top. The road has been traced some 12 km and is usually 4–5 m wide. It is bounded on both sides by a 0.5–0.6 m wide, built margin. Retaining walls, up to 2 m high at some points, were constructed as dictated by the topography. One of the remarkable features of the road is the scores of built The road was first seen accidentally during a visit to the Kerak plateau when we came to a beautiful observation point called Medinet erRas on the southwestern end of the plateau. Apart from the fine view of the Dead Sea area, we spied an unpaved road descending the steep cliffs toward Safi. The road, not marked on the maps we had, plunged 800 meters through spectacular scenery, but did not reach the Dead Sea. Climbing back was not easy, but with the afternoon sun at our backs we saw the remains of a well-built ancient road along sections of the modern unpaved track. We informed the Jordanian Department of Antiquities and were more then happy to be invited back by the former director-general, Dr. Razi Bishe. Together with Dr. Fawzi Zayadine, we returned to the area and showed them remains of the road. They confirmed that this Roman road was indeed a new discovery. The participants in the excursions along the ancient road included G. Pel’i, A. Kloner, Z. Meshel and Y. Fixler. 2

66

CHAIM BEN-DAVID

steps that span it. These steps are found in both the limestone and the sandstone zones. The margins, retaining walls, as well as the steps, were constructed using the local rock in each zone. In the sandstone zone, some sections of the road were cut into the rock. The course of the descending road can be divided into three main sections, according to the geological formations it crosses.

Figure 1. The Zoar ascent – remains of the built road in the lower section (Photo by Roi Porat)

THE KUNIYEH ASCENT 3

A built Roman road, very poorly preserved and completely different from the Zoar Ascent, descends from the Moabite plateau c. Participants in the first tour of the Kuniyeh Ascent included Gilad Pel’i and Tuvia Smotritch, and on the second tour – Roi Porat, Uri Sha‘anan and Ori Ben David. In May 2009, we presented this road in the field to the scholar of ancient Zoar, Dino Politis, at which time the upper watchtower of the ascent was discovered. Amos Kloner, Gilad Pel’i and Robert Shick also took part in this trip. 3

ROADS LEADING TO ZOARA

67

2.5 km north of the top of the Zoar Ascent (Fig. 2). The road, the Kuniyeh Ascent, dips sharply into a relatively moderate segment of Wadi Kuniyeh (Ravek and Shmida 2000:148–51), later to emerge from it and continue down to the Dead Sea coast 6 km north of Zoara.

Figure 2. The ascents in the Zoara area on a 1:100,000 scale map (Map by Dina Shalem)

In 1924, members of the joint expedition of the Xenia Theological Seminary from Philadelphia and the American School of Oriental Research in Jerusalem descended this road, stressing its steep nature (Albright 1924:3; Mallon 1924:416–19; Kyle 1928:46–50). Other than that expedition, I have found no other descriptions of the ascent in travelers’ literature and research on the area. The clear and broad remains of the road, which was also spanned by numerous steps (albeit fewer than the Zoar Ascent),

68

CHAIM BEN-DAVID

can be seen from the top of the steep slope above Wadi Kuniyeh (map coordinates 7444/4413) at 800 m above sea level. Above and at the end of the spur overlooking these remains (map coordinates 7444/4412) the remains of a watchtower (fig. 3) can be seen which had not previously been documented in any archaeological research of the area. Measuring 4.20 x 6.80 m, it is built of limestone and overlooks the entire ascent up to the lower watchtower (see below).

Figure 3. The watchtower above the Kuniyeh ascent (Photo by Gilad Peli)

We dated the abundance of pottery we saw, mainly on the slopes at its foot, to the first and second centuries CE. The road, whose width is irregular, like that of the Zoar Ascent, descends on a spur which, while steep, is the only one in the area without a cliff. The road descends in a series of dozens of switchbacks whose constructed retaining walls were frequently preserved to a height of 2 m. The upper, limestone portion of the road is quite well preserved. At the point where the spur becomes sandstone, rock-falls from the steep slope cover long segments of the wide road. Passage today though this narrow, boulder-filled naqeb is by donkey. From time to time the broad remains (3–4 m) of the ancient road can be seen. The upper part of the road, which plunges from 880 m above sea level to 50 m above sea level, is clearly marked on the topo-

ROADS LEADING TO ZOARA

69

graphical map. The road descends toward the Wadi Kuniyeh streambed near a tributary in which a spring emerges (map coordinates 7441/4431). From there, the road continues on the east bank of Wadi Kuniyeh down into the streambed before the junction with a large ravine to the southeast. The ravine is shot through with sandstone and easy to cross, but no remains of the road can be seen in this section, which is about 2.5 m long, up to the beginning of black igneous rock. At that point, the streambed cannot be crossed and the road ascends to the west bank. This portion is currently in use as an unpaved road, appearing as a trail on maps since 1995. At map coordinates 7415/4418, the unpaved road turns north and the ancient road, 4 m wide at this point, can be clearly seen, sometimes with its built margins (Fig. 4) and sometimes flanked only with stones. The road turns west-southwest toward the lower terrace above the Dead Sea.

Figure 4. The watchtower in the lower part of the Kuniyeh ascent and the remains of the built road. (Photo by Gilad Peli)

At map coordinates 7409/4416, at the top of the hill overlooking the Dead Sea, is a watchtower (fig. 4), resembling the abovementioned upper watchtower. It, too, had not been previously documented by archaeological research. Measuring 5 x 8 m, it is built of light sandstone brought from about 0.5 km away, which

70

CHAIM BEN-DAVID

stands out against the surrounding black rock. Most of its stones are ashlars with dressed margins. We identified most of the sherds at the foot of the watchtower (where they were unearthed by the numerous robber excavations the site has suffered) as dating to the first and second centuries CE. Near and below the fort, the road is wide and clearly visible. Farther down the steep slope, the ancient route can be followed as far as the alluvial terrace at the foot of the cliff.

THE SOUTHERN ZOAR ASCENT – NAQB ABRASH 4

The southern ascent is a camel trail, descending from the northwestern corner of the Moabite mountains, from the top of the cliff above Wadi Hasa to Zoara (Fig. 2). It goes from 700 m down to 350 m above sea level to the area of the water system at the outlet of Wadi Hasa, a linear distance of 7 km. The upper portion of the ascent is marked on the topographical map as proceeding from the upper cliff via a chalky, sloping rock-fall toward a wide sandstone plateau between 400 and 250 m above sea level. From there, it continues along a steep sandstone spur continuing its descent via a well-built trail (Fig. 5) 600 m toward Zoara. In 1858, Johannes Rudolf Roth described the trail as the “Labrasch Ascent” during a three-week journey through southern Moab and northern Edom (Goren 2003:131–132). On his way back from Tafila to Safi, he wrote that he took this trail (Petermann 1858:272). In 1896, Gray Hill came this way (1897:143) and complained about the time it took (13 and a half hours) and the steepness, noting that they finished it by moonlight. In 1902, Libby and Hoskins (1905:272–79) also traveled this road, which they called “al-Abrash.” They wrote that the road from Tafila to Safi took them 13 hours and that the ascent itself, from the moment they went from the limestone to the sandstone zone, was the most frightening of their entire journey.

Participants on the trips to this ascent included, among others, Avraham Izdarechet, Barak Verker, Roi Porat, Eli Raz and Shmuel Raikin. 4

ROADS LEADING TO ZOARA

Figure 5. Naqeb Abrash – the camel trail in the lower part of the Naqeb (Photo by Roi Porat)

71

Interestingly, none of these travelers noted the impressive fortified site of Labrush, which was discovered in 1883 at the bottom of the ascent by the Irish naturalist Henry Hart while chasing ibexes (1885:266). The following day he brought the rest of his expedition back to see it, including the geologist Edward Hull (1889:121). Hull described it as a series of stone circles, surrounded by a perimeter wall extending for hundreds of meters and a hillock with an internal fortification containing a water cistern. Hull reported a discussion among the members of the expedition about the essence of the site; he proposed that it was an ancient military compound. Kitchener prepared a plan of the site, sketched by Armstrong and published in Hull’s report (1889, opposite p. 121) and Kitchener’s report (1884:216–217). Kitchener himself described the stone circles as nawamis, with which he was familiar from Sinai, but noted that some were outside the perimeter wall and seemed to him to be later than the main site. Kitchener proposed that one of the rectangular structures was an ancient temple, with remains of a tower on the southern side of the site. Kitchener also noted that the ruin was on an old roadway ascending eastward; this roadway is marked as a path in the above-mentioned plan of the site. In the winter of 1924,

72

CHAIM BEN-DAVID

the expedition headed by Albright, mentioned above in the description of Kuniyeh Ascent, visited the site. Mallon (1924:435) spelled the name of the site “el-Ebrosh,” and dated it to the Arab period. Kyle (1928:59) called it “Lubrous” or “El Ubrous,” and viewed it as “a fortified place to protect the caravan road.” In a survey of Byzantine and Islamic sites, King et al. (1987:449) noted the same site under the name Umm et-Tuwabin. 5 According to this source, the site was discovered by the Bronze Age scholar McCreey who had heard that name from the locals (MacDonald 1992:95 n. 1). In a survey by MacDonald (1992:86) the site was called “Umm al-Tawabin;” he, of course, linked it to the previous discoveries of Kitchener and others. MacDonald defined the site as a “very large Nabataean fortress.” He notes that the site had two parts, a lower compound including walls with a total length of 2.5 km, and an upper fortification or fortress. He proposed that the stone circles – most of which have a diameter of 2–3 m and some of which rise to a height of four courses – were for soldiers’ tents, but notes that they cannot be definitively linked to the rest of the finds at the site, i.e., the wall and the upper fortification. MacDonald’s survey (1992: 58) found sherds from the Chalcolithic/Early Bronze Age, which he said could not be securely associated to the main compound. Since most of the pottery at the site was Nabataean, he believed the entire site was Nabataean. This site is unique mainly because of its long walls (Fig. 6), which appear very impressively in the color aerial photographs published by Kennedy and Bewley (2004:140–41). Kennedy wrote that from the air, the walls recalled the Roman ramp at Masada. On a recent visit to the site, we walked the line of the walls and discovered that contrary to what had been published previously – that the walls did not form an enclosure – the walls appear to abut the southern end of the upper fortification. Though it is unclear whether the walls and the upper fortification are contemporaneDuring January 2015 Alexandra Arioti surveyed and mapped the site (Ariotti 2017), and she is planning an excavation season in the site in the winter of 2017 (http://www.pef.org.uk/blog/surveying-umm-at-tawa bin-a-roman-military-site/; http://www.pef.org.uk/blog/2016–grantabstracts/) 5

ROADS LEADING TO ZOARA

73

ous. 6 We noted a considerable difference between the western and the eastern walls; the western wall was megalithic in its construction and consists of two parallel walls in an area where the topography was moderate – a kind of wide casemate wall – while the eastern wall was built as a single, two-faced wall over steep slopes and cliffs.

Figure 6. Umm Tawabin – view from South East, the inner ridge with the fort can be seen in the center of the figure. (Photo by Roi Porat)

The upper fortification that was built at the northern end of the steep slope was severely robbed and its exact form is difficult to determine. But it seems to have been a fortress measuring 35 x 45 m with protruding towers at least on its northern side. Most of the ceramic vessels seen at the site belong to the Hellenistic and Early The visit to this site and to the lower portions of the Kuniyeh and Kathrabba ascents took place in April 2009. Participants were Benny Arubas, Uri Davidovich and Roi Porat. 6

74

CHAIM BEN-DAVID

Roman periods, including painted Nabatean vessels. One special feature of this fortification, which has a commanding view of the entire area of Zoara, is the series of five artificial moats dug on its southern side which provides access to the fortress (Fig. 6), possible a result of a battle during the Hellenistic or Early Roman period. In any case, it is noteworthy that the structure is on the ascent itself.

Figure 7. The watchtower above the lower part of Naqeb Abrash (Photo by Roi Porat)

In addition to the above site, three watchtowers were found that had not been documented in earlier research or surveys, one at the top of the ascent and two at the top of the lowest terrace.

1. Top of the ascent (map coordinates 7424/4303; 701 m above sea level): A square structure, measuring 9 x 9 m, overlooking the entire upper part of the ascent with a line of site to the forts on the lower terrace. The walls are 1 m thick and it is divided into two equal parts. 2. The top of the cliff overlooking the lower part of the ascent (map coordinates 7398/4318; 272 m above sea level): A square structure measuring 8 x 8 m, built of light sand-

ROADS LEADING TO ZOARA

75

stone blocks between which are smooth ashlars with dressed margins (Fig. 7). Roman and Byzantine pottery (but not Nabatean) was found. The place overlooks the entire lower portion of the ascent and the entire Zoara area. Notably, the watchtower is located about 0.50 km off the road itself and was apparently built there because of the special view. 3. The top of the lower part of the ascent (map coordinates 7398/4312; 306 m above sea level). A square structure measuring 10 x 10 m, also built of light sandstone between which are numerous ashlars. Plentiful pottery, a great deal from the Roman and Nabatean periods.

THE SIGNIFICANCE OF MULTIPLE ROADS NORTH OF ZOARA

Until recently, only one road was known to scholars north of Zoara from the Roman period – Kathrabba Ascent. Now we know of two more: Zoar Ascent and Kuniyeh Ascent. Could all these roads have led to Zoara and/or were they contemporaneous? The dating of roads is no easy task. In fact, until excavations are undertaken at points or structures along the road, we will have no clear answers. Still, we will raise a few preliminary suggestions. 7 First, we will discuss the purpose of the road going down the Kathrabba Ascent to the Dead Sea coast near the outlet of Wadi Isal. In modern surveys, segments of that road could only be seen as far as the slope south of Wadi Isal, and none could be seen along the Dead Sea. The first to describe this ascent, as noted, was Musil. His map clearly shows that the road crosses Wadi Isal and continues north. No parallel road southward to Zoara is marked. However, a road shown as being of similar width, although not its direct continuation, is marked from the area of Mazra‘a toward the Dead Sea. This road, which was also documented by other scholars, ends, as noted, at the Dead Sea coast. As mentioned above,

trip.

7

Most of these ideas came out of the discussions we held during our

76

CHAIM BEN-DAVID

Kyle described a Roman road that reached the Lisan. Regarding the road between Ghawar Safi (Zoara) and the Lisan, he writes: One of the interesting things here in the Orient, is that every now and then one finds himself in the footsteps of the Romans, actually going along an old Roman Road ….. suddenly we became conscious that we were again on a made road. There 8 on the right hand, still holding its head up, was an old Roman milestone. 9 The old Roman road! It ran up and down this plain at the edge of the mountain desert land. …a branch ran of to the left, right down to the center of the Lisan…

As for the question of where the road led and what the possibilities were, Kyle said, “the Romans did not build roads to nowhere. That road led down to a bridge or a ford or a ferry, or simply to a continuation of the road itself to the western side of the valley.” It is unclear from Kyle’s description from what point Kyle began to realize that he was on a Roman road. Was it only north of Wadi Isal and the continuation of Wadi Kathrabba, or perhaps he first noticed south of Wadi Isal, a segment that Glueck noted in his aireal surveys (below). This lack of clarity makes it impossible to determine the precise location of the milestone Kyle found or to attribute it to the Kathrabba Ascent. 9 It is interesting that the other members of that expedition, Albright and Mallon, did not mention anything about the milestone in their reports. In fact, after Kyle, it was never mentioned again. To the best of our knowledge, the only person who discussed this find was Beno Rothenberg, in his book, Negev – Archaeology in the Negev and the Arabah (1967: 171, 273 n. 267). Rothenberg wrote that Albright had discovered a Roman road and a milestone east of the Dead Sea. In a note, Rothenberg stated: “It is possible that remains of the Roman road, with the milestone, in the section near the Lisan, are not connected to this road (the north-south road on the east side of the Dead Sea), but rather to a road that came down from the Mountains of Moab to the Dead Sea and continued westward to the end of the Lisan. The continuation of the road from there, and the reason for its construction to that point, have not yet been sufficiently researched...” 8

ROADS LEADING TO ZOARA

77

Although it is very tempting to propose that passage by land to Masada, was via the Lisan, apparently there was no such passage in the Roman period. It should be noted that on the western coast of the Dead Sea, in the entire area of Masada, and in fact throughout the entire area between Ein Gedi and the ʿArava Junction no remains have been found of a parallel Roman road. Of course, the idea of a bridge is implausible, but the possibility of ferries or a port is certainly reasonable. In Glueck’s aerial reconnaissance of the Kathrabba Ascent (1937) he noted that remains of the road disappear near Hirbet Isal, remarking that Musil had documented the road from there to the Lisan. Glueck also wrote that from the air they could not see segments leading to the Lisan. To the south, however, segments of the road, albeit fragmented, could be seen, which in his opinion lead to Ghor es-Safi, i.e., Zoara. In a survey conducted by Jacobs along the road, she noted the paucity of finds from the Early Roman period and the lack of milestones as support for her argument that the road dates not to the Roman period but to the Late Roman or the Byzantine period. The published pottery devoted to the Nabatean, Roman and Byzantine periods does not permit a critique of these distinctions, and it should be recalled that scholars who mention the Late Roman period also mean the second century CE. Moreover, a careful study of the arguments Jacobs presented do not support her theory, since the fact that more sites were found from the Byzantine period in the vicinity of the road (and not necessarily directly associated with it) is irrelevant to the question of the date of its construction. This specific issue will only be resolved by a focused survey and/or excavations of installations connected to the road, such as forts. In the second decade of the second century CE, the Via Nova Triana was built from Bostra to Aila. Presumably, east-west roads were also constructed at that time as part of that project, either to link up with the Roman road network in Provincia IudaeaPalaestina or to the western reaches of the new province of Arabia. The Katthrabba Ascent connected the Via Nova Traiana to the area of the Lisan opposite Masada, where military presence dating to 106–113 CE has been discerned (Arubas and Goldfus 2008). That presence was not a continuation of the Roman presence following the siege and conquest of Masada. Rather, it was apparently connected to the establishment and organization of Provincia Ara-

78

CHAIM BEN-DAVID

bia, which may have included the transfer of Masada and the surrounding region from Judea to Arabia (Arubas pers. comm.). It is not impossible that during the Bar Kokhba revolt, this route also served the Roman army on its way to the rebel camps on the Dead Sea coast and to the hideout caves (Porat 2009: 371–76). In any case, and without reference to the above historical hypotheses, it seems that this road is not directly connected to the road system to Zoara. This argument is based on the three main points: 1) the testimony of Kyle and Musil; 2) the direction of the road from the southern Moabite plateau northwest, unusual for travelers on their way to Zoara, considering that the route clearly heads for the southern end of the Lisan (northwest of Ghawar Isal); 3) the fact that there are not less than two good “imperial” period ascents to Zoara.

BETWEEN THE KUNIYEH AND THE ZOAR ASCENTS

Unlike the Kathrabba Ascent, the logic behind the Kuniyeh and the Zoar ascents is clear, and in any case, they both lead to Zoara, and not to any other point along the Dead Sea coast. Based on our impression from the sherds at the fort at Kuniyeh and the meager use of the ascent, it may be proposed at this stage that the Kuniyeh Ascent was built first, perhaps also at the beginning of the second century as part of the paving of roads linking to the Via Nova Traiana. Table 1: The ascents in the Zoara area – general data

Ascent

Height differential

Linear distance

Type of road in the past

Type of road today

Kuniyeh Zoar

1150 1280

5 8

Built road Built road

Southern Zoar

1050

7

Camel ascent

Donkey trail Upper portion, unpaved road; lower portion, donkey trail Camel ascent

Number of watch towers along the road 2 1

4

After some time, apparently soon thereafter, this road went completely out of use, perhaps due to an earthquake that caused the large rock-falls that are typical of the middle, steep portion of the

ROADS LEADING TO ZOARA

79

road today. It was then that the Zoar Ascent was built, which remained in use for the rest of the Roman period, the Byzantine period and even into the Middle Ages.

THE ANCIENT ROUTES FROM JUDAEA-PALAESTINA TO ZOARA

As mentioned above, ancient travelers passed through the geographical setting of ancient Zoara (Safi) on the southeastern side of the Dead Sea as they crossed from the southern part of the Roman province of Judaea-Palaestina to Arabia. The routes leading from Judaea-Palaestina to Zoara will be described from north to south (fig.8).

Figure 8. General map of the roads and trails leading to Zoara (Photo by Dina Shalem)

THE ZERON ASCENT – NAQEB EZ-ZUWEIRAH

This is the most convenient topographic ascent from the Dead Sea up to the Judean Desert plateau due to the absence in this area of the steep cliff that is otherwise prominent along the western side of the Dead Sea. The modern road from Arad to the Dead Sea was built along this ancient route. Its use in periods as early as the Early

80

CHAIM BEN-DAVID

Bronze Age until recent times has been noted by a number of scholars.

THE LOWER SECTION

Yekutieli (2004, 2006) claimed that the lower section of the Zuweirah Ascent was used as early as the Early Bronze Age and that it connected Arad with Bab ed-Drah on the east. Based on the identification of an Iron Age tower, Aharoni and Rothenberg suggested that the biblical route from Judah to Edom passed via the lower section of the Zuweirah Pass along the route (Aharoni and Rothenberg 1960: 13–47; Rothenberg 1967: 10–112). However, a recent study by Berman (2008), based on a survey of the lower section of the pass, revealed no evidence dated to the Iron Age. By comparison, in 12 sites dated to the Hellenistic period, 5 were found to have building remains. This data led Berman to suggest that that the main phase of the Zuweirah Ascent was constructed in the Hellenistic period by the Hasmonean Kingdom in Judea, and that it may have therefore served as the road between Jerusalem and Zoara. Another interpretation to the Hellenistic remains along this route was suggested lately by Kloner (2015: 366–367), in his opinion it was part of an Idumaean trade route from Idumaea to Nabatea. On the shore of the large artificial pool that currently constitutes the southern part of the Dead Sea, and 2 km south of the Zuweirah Pass, stands Mezad Gozal. This is a massive square tower (16 x 16 m) surrounded by an enclosure wall whose thickness ranges between 2.5 and 3.5 m. Following Aharoni’s excavation of the site, he, and later Rothenberg, identified it as an Edomite roadside fort (Aharoni 1964: 112–13; Rothenberg 1967: 167–69). Hirschfeld posited that it was built as a Nabataean fort along the road that led to Zoar (Hirschfeld 2006: 167–69). However, no Nabataean finds have been reported from the site in the past or in recent excavations in 2010 (Talis 2013). The pottery found in this excavation was dated to the Hellenistic period and to the first century BCE. As for later periods, the remains of an ancient fort, Mezad Zohar, are present on the lower section of the ascent. Frank reported this to be a Roman site; however, in 1936, the British inspector of the Palestine Department of Antiquities discovered sherds only of the medieval period (Frank 1934: 201, 256; Petersen 2001). An excavation of the site in 2004 revealed that it was built

ROADS LEADING TO ZOARA

81

toward the end of the twelfth century CE and was occupied primarily in the Mamluk period, up to the fourteenth century (EricksonGini et al., 2016). Apparently it served as a guard station along the Mamluk postal route, the Barid, from Cairo via Gaza and Hebron to Zoar and Kerak.

THE UPPER SECTION

The upper section of this pass, located ca. 10 kms south east of ancient Arad and called the “Pass of ez-Zuweirah” by Robinson (1841:2.477) was traversed by many nineteenth-century travelers on their way to Moab and Edom, including Seetzen, Irby and Mangles and Tristram. At the head of this section is a casemate fort that was excavated by Harper (1995), who dated it from the end of the fifth century CE to the beginning of the seventh century CE. Based on the published pottery, Magness suggested a later date for the site’s construction, the mid-sixth century (Magness 1999: 195–99).

THE AMAZ ASCENT – NAQEB AMAZ

Robinson mentioned the existence of a camel trail along this ascent (Robinson 1841: 2.493). Until the last decades of the twentieth century, the remains of a watchtower could be seen at the head of the pass; however, no archaeological data concerning the site was recorded (Marcus 1984: 51).

THE TSURIM ASCENT – NAQEB BUWIEB

Robinson mentions that Naqeb Buwieb was a donkey trail that he viewed from below (Robinson 1841: 2.493). According to Frank, the pass was difficult to negotiate and mainly used by smugglers (Frank 1934: 278). Rothenberg surveyed some small sites along this pass, including a small square building, measuring 6.5 x 6.5 m at the head of the pass, which was apparently a watchtower (Rothenberg 1967: 120–21). Vilnai refers to that site as Qaser Haned in Arabic and Mezad Tzurim in Hebrew (Vilnai 1977: 4675).

THE TAMAR PASS – THE ROMAN ROAD FROM MAMPSIS TO ZOARA

The Tamar Pass was a road with built sections, hewn steps, watchtowers and a large fort on the upper part of the pass. It was first noted by Robinson, who referred to it as es-Sleisel when he passed

82

CHAIM BEN-DAVID

below it (Robinson 1844: 2.493). In the nineteenth century, travelers did not use this pass, preferring the abovementioned Zuweirah Pass. However, since the beginning of the twentieth century this route has become the major one, possibly due to changes in the level of the Dead Sea (Ben David 2002a: 159). On British maps from World War I, the pass is marked as Naqeb Jahniya and Naqeb Misel ibit el Saudri. Frank was the first to draw attention to the distinct remains of a Roman road along it (Frank 1934: 257–259, 279–80). According to Frank, this pass was the main route for loaded camels coming from Safi on their way to Beer Sheba and Hebron. Frank, who recorded the remains of the ancient road, was also the first to note the remains of the large Roman fort of Qasr Jahaniya, later excavated by Gichon, who identified it as ancient Thamaro (Gichon 1993). Following Frank’s description, Alt marked this route as a major Roman road (Alt 1935: 24) and so did all other scholars of the Negev. The road was surveyed in the early 1970s by Marcus (1973) and again in 2000 and 2001 by Fixler (2001), who mapped sections of it on aerial photographs by means of a computerized stereoplotter on aerial photographs. A topographic map was produced with vertical spaces of 1 m between the contours as well as graphs of the road’s inclines. This is the only built Roman road west of the Dead Sea that leads to Zoara. Thus, unlike the eastern side of the southern region of the Dead Sea, where three built roads were constructed in the Roman period, only one was constructed on the western side. This Roman road, coming from Beer Sheba and Mampsis, was the main connection between coastal cities like Gaza and Ascalon as well as inland centers like Jerusalem Eleuthropolis, Haluza and Zoara. As suggested in past studies (Ben David 2002b: 104), this road may be the one marked on the Tabula Peutingeriana as the connection, south of the Dead Sea, between the road system of Arabia and that of Judaea-Palaestina.

THE CONNECTION BETWEEN THE TAMAR PASS AND ZOARA

Clear signs of the Roman road can be seen from the modern road near the turnoff to the Small Crater south of Nahal Peres (Wadi elFarsh), just above the modern Arava Junction. From that point until Zoara no clear archaeological remnants of the road were

ROADS LEADING TO ZOARA

83

found. A tower (11.00 x 8.90 m), similar to that of Mezad Gozal that is located east of the modern Arava Junction, was apparently situated along the route to Zoara. It was noted by Frank as Kasr Musk el-Ghawarne and by Rothenberg as Mezad Mazal (Frank 1934: 256; Rothenberg 1967: 118). Rothenberg identified pottery from the Iron Age and the Hellenistic and Roman periods. The tower was excavated by Aharoni; however, no final report on the site was published. The site of ʿEn Tamar, located on the southern edge of the Dead Sea, has been described by Hirschfeld as “the second most important settlement after Zoara in the northern Wadi Arabah” (Hirschfeld 2006: 170). Based on the topography and on nineteenth-century maps, it appears that the main road passed the site of ‘En Tamar. The Nabataeans, who controlled the area in the Hellenistic and Early Roman periods, built a small fort on a hill near the spring. In the first century CE, the Nabataeans built a station on lower ground farther to the west. Excavations at this site revealed that it was occupied between the first century CE and the early third century CE (Cohen 1983; Hirschfeld 2006; Erickson-Gini, forthcoming). It was reoccupied by monks in the Late Byzantine period (Hirschfeld 2006). While Rothenberg (1967: 115) and Cohen (1983) identified some buildings as guard stations along the Roman road to Zoara, Hirschfeld viewed them as residential buildings of a settlement that he suggested may be ancient Thamaro (Rothenberg 1967:115; Cohen 1983; Hirschfeld 2006: 172–73). From ʿEn Tamar, Roll and Hirschfeld marked on their maps a straight 12 kilometer-route northeast to Zoara through the salt marsh known as the Sabkha, (Roll 1994; Hirschfeld 2006: 168, Fig. 13.2). But Rothenberg suggested that the road was intended to bypass the salt marsh, because the marsh is passable only occasionally during the summer (Rothenberg 1967: 264, n. 35, see also pg. 115, map 8). Along this proposed route, some 7 kms southeast of ‘En Tamar, Rothenberg identified a square structure (30 x 30 m), that he suggested was a late Roman fort based on the pottery of that period collected at the site (Rothenberg 1967: 117). The proposed route suggested by Rothenberg lengthens the route by 8kms.

THE ROUTE FROM THE WADI ARABAH TO ZOARA

A potential connection between Aila, Petra, the copper mines in Feinan and Zoara passed through Wadi Arabah (Smith 2010: 92–

84

CHAIM BEN-DAVID

101). In the northern section of Wadi Arabah two routes can be taken in consideration, a western and an eastern one. The western route comes from ʿEn Husub, the site that is the best candidate for ancient Thamaro. The route proceeds 3 km in a northerly direction and then northeast via Wadi el Qseib to ʿEn Tamar and Zoara. The eastern route proceeded via Qasr et-Tlah, identified with ancient Toloha (Kennedy 2004: 214; Smith 2010: 50), then via Qasr elFeifeh, identified with ancient Praesidium (Kennedy 2004: 214) to Zoara.

BIBLIOGRAPHY

Aharoni Y. (1964). Mesad Gozal. IEJ, 15,112–13. Aharoni Y. and Rothenberg B. (1960). In the Footsteps of Kings and Rebels in the Judean Desert. Raman Gan: Masada. (Hebrew) Albright F.W. (1924). The Archaeological Results of an Expedition to Moab and the Dead Sea. BASOR 14, 1–12. Alt A. (1935). Romische Kastelle und Strassen, Inschriften und Feldzeeichnungen. ZPDV 58, 1–59. Ariotti, A. (2017). A survey of the Nabataean/Roman Site, Umm at-Tawabin, Ghor as-Safi, Jordan. PEQ 149, 44–86. Arubas, B. and Goldfus, H. (2008). Masada – The Roman Siege Works and the Assault Ramp. NEAEHL– Supplementary Volume, pp. 1937–39. Ben David, C. (2001). The Ascent of Luhith and the Road to Horonaim, New Evidence for their Identification. PEQ, 133, 136–44. Ben David C. (2002a). The Roads from Judea to Zoar and Moab in the Roman Period.Judea and Samaria Studies, 11, 151–64. (Hebrew). Ben David, C. (2002b). “The Zoar Ascent – a Newly Discovered Roman Road Connecting Zoar-Safi and the Moabite Plateau., In: P. Freeman, J. Bennett, Z.T. Fiema and B. Hoffmann, eds, Limes XVIII – Proceedings of the XVIIIth International Congress of Roman Frontier Studies held in Amman, Jordan. BAR-IS1084, Oxford: Archeopress, pp. 103–12. Berman M. (2008) Zeon Ascent, Hellenistic Road in South Judean Desert. Unpublished thesis. Ben-Gurion University of the Negev.

ROADS LEADING TO ZOARA

85

Cohen R. (1983.) En Tamar. Excavations and Surveys in Israel 2, 31. Dearman A.J. (1997). Roads and Settlement in Moab. BA 60, 205–13. Erickson-Gini, T., Nahlieli, D. and Kool, R. (2016). Mezad Zohar: A Medieval Fort near the Dead Sea. Strata: Bulletin of the AngloIsrael Archaeological Society 34, 125–149. Erickson-Gini, T. (forthcoming) Excavations in ʿEn Tamar. ʿAtiqot Fixler Y. (2001). Mapping and Analysis of Roman Roads in Palestine Based on Aerial Photographs. Ph. D. Thesis. Bar Ilan University (Hebrew). Frank F. (1934). Aus der ʿAraba I. ZDPV 57, 191–280. Gichon M. (1993) Mezad Tamar. NEAEHL pp, 1437–40. Glueck, N. (1937). An Aerial Reconnaissance in Southern TransJordan. BASOR 67, 19–26. Glueck N. (1939). The Annual of the American Schools of Oriental Research, Vol. 18/19, Explorations in Eastern Palestine, III (1937– 1939). New Haven: The American Schools of Oriental Research. Goren H. (2003) Zieht hin und erforscht das Land: Die deutsche Palästinaforschung im 19. Jahrhundert (Schriftenreihe des Instituts für deutsche Geschichte der Universutät Tel Aviv, 23). Gőttingen: Wallstein. Harper, R.P. (1995). Upper Zohar, An Early Byzantine Fort in Palaestina Tertia, Final Report of Excavations in 1985–1986. (British Academy Monographs in Archaeology No. 9). Oxford: The British School of Archaeology in Jerusalem. Hart, H.C. (1885). A Naturalist’s Journey to Sinai, Petra and South Palestine. PEQ 17, 231–86. Hill, G. (1897). A Journey East of the Jordan and the Dead Sea. PEQ 29, 24–46. Hirschfeld Y. ( 2006). The Nabataean presence south of the Dea Sea: new evidence. In: P. Bienkowski and K. Galor eds, Crossing The Rift – Resources, Routes, Settlement Patterns and Interaction in the Wadi Arabah. Oxford:Oxbow Books, pp. 167–90. Hull, E. (1889). Mount Seir, Sinai and Western Palestine. London: Richard Bentley and Son.

86

CHAIM BEN-DAVID

Jacobs, L.K.. (1983). Survey of the South Ridge of Wadi Isal 1981. ADAJ 27, 245–74. Kennedy D.L. ( 2004). The Roman Army in Jordan, 2nd rev. ed. London: Council for British Research in the Levant. Kennedy D. and Bewley R. (2004). Ancient Jordan from the Air. London: Council for British Research in the Levant. King, G.R.D., Lenzen, C.J., Newhall, A., King, J.L., and J.D. Deemer. (1987). Survey of Byzantine and Islamic Sites in Jordan, Third Season Preliminary Report. ADAJ 31, 439–59. Kitchener, H.H. (1884). Major Kitchener’s Report. PEFQS 16(4), 202–21. Kloner, A. (2015). Idumaea during the Persian and Early Hellenistic Periods: Territory and Caravan Trade Routes. Eretz-Israel – Ehud Netzer 31, 359–73. (Hebrew) Kyle, M.G. (1928). Explorations at Sodom. New York: The Religious Tract Society. Libbey, M., and F.E. Hoskins. (1905). The Jordan Valley and Petra. New York: G.P. Putnam’s Sons. MacDonald, B.(1992). The Southern Ghors and Northeast ʿArabah Archeological Survey. Sheffield: J.R. Collis. Magness, J. (1999). Redating the forts at Ein Boqeq, Upper Zohar and other sites in SE Judea, and the Implications for the Nature of the Limes Palaestinae. In: J.H. Humphrey, ed., The Roman and Byzantine Near East, Volume 2. Portsmouth: JRA, pp. 189–206. Mallon, A. (1924). Voyage d’exploration au sud-est de la Mer Morte. Biblica, 5, 413–55. Marcus M.(1973). Maʾale Tamar-Peres: an Ancient Road on the Hatzera Ridge. Tel Aviv: Hazeva Field School – Israel Nature Protection Society. (Hebrew). Marcus M. (1984). The Sothern Judean Desert, Landscape Survey. Tel Aviv: Jerusalem Nature protection authority (Hebrew). Mittmann S. (1982). The Ascent of Luhith. SHAJ 1, 175–80. Musil A.(1907). Arabia Petraea-Moab. Vienna: Hölder. Petermann A. (1858). Prof. Dr. J.B. Roth’s Reisen in Palästina. PM 4, 267–72.

ROADS LEADING TO ZOARA

87

Petersen, A. D. ( 2001). Two Medieval castles and their position in the military architecture of Muslim Palestine. In: U. Vermeulen, ed., Egypt and Syria in the Fatimid, Ayyubid and Mamluk eras III. Leuven: Peeters, pp. 383–406. Porat R.(2009). The Area Between Ein Gedi and Qumran Towards the End of the Bar Kokhba Revolt. In: H. Eshel and R. Porat, eds, Refuge Caves of the Bar Kokhca Revolt, Second Volume. Jerusalem: Israel Exploration Society, pp. 365–93. (Hebrew). Ravek, S. and A. Shmida. (2000). To the Canyons of Moab and Edom. Tel Aviv: Teva Hadvarim. (Hebrew). Robinson, E.( 1841). Biblical Researches in Palestine, Mount Sinai and Arabia Petraea, Vol II. Boston: Crocker & Brewster Roll I. (1994). Roman Roads. In : Y. Tsafrir et al., eds, Tabula Imperii Romani – Iudea Palaestina. Jerusalem: Israel Academy of Sciences and Humanities, pp. 21–22. Roll I. (1999). The Roads in Roman-Byzantine Palestina and Arabia. In: M. Piccirillo and E. Alliata, eds, The Madaba Map Centenary. Jerusalem: Franciscan Printing Press, pp. 108–113. Rothenberg B. (1967). Negev – Archaeology in the Negev and the Arabah Ramat Gan: Masada (Hebrew.) Smith II A. M. (2010). Wadi Araba in Classical and Late Antiquity: an Historical Geography. BAR-IS 2173. Oxford: Archaeopress. Talis S. (2013). Mezad Gozal. [online]. Hadasot Arkheologiot Excavations and Surveys in Israel 125. Available at: www.hadashot-esi.org.il/Report_Detail_Eng.aspx?id=5451& mag_id=120 Accessed 23 July, 2015]. Vilnai Z. (1977). Mezad Zurim. Ariel – Encyclopedia of the Land of Israel. Tel Aviv: Am Oved – Tarbut Vechinuch, p. 4675. (Hebrew). Yekuteli Y. ( 2004). Landscape of Control: An Early Bronze Age Ascent in the Southern Judean Desert. ANES 41, 5–37. Yekuteli Y.(2006). Is Somebody watching you? Ancient Surveillance Systems in the Southern Judean Desert. Journal of Mediterranean Archaeology 19, 65–89.

AN ASSESSMENT AND RE-EXAMINATION OF THE AMERICAN EXPEDITION IN PETRA EXCAVATION IN THE RESIDENTIAL AREA (AREA I), 1974–1977: THE EARLY HOUSE AND RELATED CERAMIC ASSEMBLAGES TALI ERICKSON-GINI & CHRISTOPHER A. TUTTLE 1 INTRODUCTION

The excavation of Area I of the Temple of the Winged Lions project conducted by the American Expedition to Petra (AEP) was supervised by the late Kenneth W. Russell under the direction of Prof. Philip C. Hammond (University of Utah) between 1974 and 1978 (Fig. 1). The excavation of Area I is one of the most extensive excavations of a residential area in the Petra city center to date. The untimely death of Dr. Russell at age 42 in 1992 prevented him from finishing the final publication of the excavation of Area I, which he planned to entitle, “Household Excavations at Petra.” The work carried out by Russell produced important evidence of earthquake activity in Petra during the first half of the first millennium CE, forming the basis of some of his published research (1980, 1985) and influencing Hammond’s interpretation of the We extend our best wishes to Dr. S.T. Parker on the occasion of the publication of his festschrift, as well as the editors. Dr. Parker participated in the excavation of Area I in 1975. 1

91

92

TALI ERICKSON-GINI & CHRISTOPHER A. TUTTLE

stratigraphical record discovered in the Temple of the Winged Lions complex in Areas II–IX (hereafter Area II).

Figure 1. Location of the Temple of the Winged Lions and Area I

The demise of Philip C. Hammond, professor emeritus of anthropology in the University of Utah in 2008, cut short work on the final publication of his excavations in the Temple of the Winged Lions and its ancillary structures (Area II). 2 In 2010, Dr. Christopher A. Tuttle, the former Associate Director of the American Center of Oriental Research (ACOR) in Amman, Jordan, organized the Petra Temple of the Winged Lions Cultural Resource Management Initiative (TWLCRM), (Tuttle 2012, 2013). The objective of the initiative is the conservation and presentation of the Temple of the Winged Lions to visitors in Petra. The initiative includes the online publication of the Hammond archives and the final publication of the results of the excavations in Areas I and II-IX with a reevaluation of the excavated material in light of recent archeological discoveries in Petra and beyond. This paper has been prepared by the writers following an extensive and detailed study of the AEP archive and the finds discovered in the residential areas and the Temple complex. A particular Hammond published preliminary findings about the architecture of the temple and a number of artifacts uncovered there and in the ancillary rooms (Area II), Hammond (1996) and Hammond and Johnson (2000), as well as in various articles: Hammond (1977; 1978; 1980; 1981; 1986; 1987; 1992; 2000; 2002; 2003; 2005. 2

THE AMERICAN EXPEDITION IN PETRA EXCAVATION

93

emphasis is placed on the presentation of ceramic assemblages and associated finds uncovered in the excavation and a re-evaluation of the phasing of Area I in light of current understanding of Nabataean ceramic typology and dating. The conclusions to be presented here include a revision of the dating of the Early House in Area I and the ceramic assemblages uncovered its antechamber and the upper and lower levels of the structure to the late 2nd and early 3rd c. CE when the structure was abandoned. This revised dating is supported by evidence from other parts of the AEP excavations such as the Painters’ Workshop and important find spots near the temple that are presented in this paper as well as material from other parts of the Provincia Arabia in the post-annexation period. Over the past generation, researchers have largely been compelled to compress the dates of Nabataean material culture into a very narrow historical framework that effectively ends with the Roman annexation in 106 CE; this in spite of the fact that Petra continued to be an important center into the early 3rd c. Like other cities in the Roman East, Petra enjoyed a renaissance in construction and culture under Roman rule. The use of a revised ceramic chronology in dating these assemblages will undoubtedly prove to be controversial, however we believe that such a revision is long overdue and is in itself an important tool for the re-examination of the phasing of structures and occupational layers in Petra and other sites in the Provincia Arabia, the vast majority of which have been erroneously dated to the later 1st to early 2nd c. CE.

THE EXCAVATION OF AREA I

The Temple of the Winged Lions lies on a terrace above the north bank of Wadi Musa facing the cardo maximus of Petra. Prior to its excavation this area was called the “Upper and Lower Gymnasia”, (Bachmann et al. 1921). The results of a combined protonmagnetometer and soil-resistivity survey carried out by Hammond in this area in 1973 indicated a high potential for the discovery of architectural remains in Areas I and II. Area I is located ca. 50 meters due east of the Temple of the Winged Lions in Area II. A grid of nine 10 x 10 meter squares was laid out on a north-south axis in Area I. Two central squares, Sites I.2 and I.5, were excavated to a depth of ca. four meters (Fig. 2). In 1977, Russell prepared a tentative phasing of the stratigraphy in

94

TALI ERICKSON-GINI & CHRISTOPHER A. TUTTLE

Area I. The final phasing prepared by him in 1978 indicates the presence of twenty archaeological phases (Phases XX–I) and the remains of successive domestic structures of the Early Roman (preannexation, i.e., the Roman annexation of Nabataea in 106 CE), Middle Roman (post-annexation) and Byzantine periods. He designated these structures the “Early House”, the “Middle House”, and the “Late House”. The Early House is located on the western side of the excavated squares (I.2 and I.5) and it partially underlies the Middle House and Late House of later periods. The Middle House is located in the northern half of I.5. Its southern perimeter is covered by the Late House. The Late House is located in I.2 and the southern half of I.5. The cemetery was found directly below the surface of the excavation squares.

Figure 2. Plan of the Early House in Area I

The earliest archaeological material discovered in Area I, uncovered below the earliest architectural remains and in ancient fills, dates to

THE AMERICAN EXPEDITION IN PETRA EXCAVATION

95

the Hellenistic period. The latest material belongs to an overlying cemetery that Russell dated to the Late Byzantine or Early Islamic periods. Area I was partially back-filled during the excavation of the Petra Church in the 1990s.

A REVIEW OF AEP METHODOLOGY

According to the excavators, the expedition followed a modified version of the ‘Wheeler-Kenyon’ method of excavation (Hammond and Johnson 2000). In both Areas I and II, 10 m. x 10 m. excavation squares were laid out and numbered by transit survey on the basis of an electronic survey grid carried out in 1973. Soundings were carried out in areas prior to excavation. The north and west sides of the unexcavated baulks of the squares were used as controls for the internal stratigraphy. Secondary baulks were established inside the squares in order to maintain sequential control when the horizontal stratigraphy was interrupted by features encountered, e.g. walls, pits, and other installations. Permanent reference points were established for the internal triangulation of features and find spots. Discernible layers of deposition and all features, e.g., walls, installations, floors, ect, were assigned stratigraphic unit (SU) numbers, which were used in place of locus numbers. Assemblages of artifacts recovered in the excavation were also assigned SU numbers in place of basket numbers. Walls uncovered in the excavation were assigned local numbers according to their squares as well as SU numbers. The SU designations were recorded numerically sequential according to their order in the excavations but not according to the order of their actual SU depositions. The individual SUs were marked on baulks by tagging horizontal lines of deposition at 0.50 m. intervals. At the end of each season, baulk sections and top plans of the square were drawn to a scale of 1:20. The progress of the excavation was recorded daily by area supervisors in small notebooks with pre-numbered pages and they were requested to fill out daily form cards with abbreviated versions of the notebook entries. The cards were to be filed on a Daily Progress Chart in order to provide visual reference of sequential details. The field notebooks were to include descriptions of the soil color and composition, a sequential description of SUs and their relationship to adjacent SUs, and summaries of the ‘close-out’ of each square and SU correlations with adjoining squares.

96

TALI ERICKSON-GINI & CHRISTOPHER A. TUTTLE

Artifacts uncovered in the excavations were marked according to their square number and SU number. Pottery sherds and vessels were washed and sorted daily. Artifacts, including small finds, coins, ceramic vessels, glass vessels and architectural features such as carved stone objects and stucco affixes were photographed in the field laboratory at the site. Comprehensive registries of the finds were compiled for each year of the excavation between 1974 and 2005. These have proved to be highly useful in synthesizing the results of the excavation. The ceramic sherds of each particular SU were photographed in black and white in earlier years and in color in later years. It must be stated that in spite of the intentions of the excavators, the recording system proved to be problematic due to the lack of comprehensive master lists of SUs, printed excavation forms, and daily graphic logs. Moreover, absolute elevations are lacking throughout the excavation process. Due to the large size of the excavation squares (10 x 10 m.), section drawings of the baulks often fail to show critical data revealed in other parts of the squares. According to the excavators, this situation could be partially remedied by relating drawn SUs to schematic tables but in reality this was rarely the case. Generally, only half of any one square was excavated each season and the system of recording did not facilitate comprehensive documentation from one year to the next. The daily recording in the small field notebooks were handwritten and they often lack critical details with regard to find spots of important assemblages and architectural features. The area supervisors occasionally drew small top plans or sections (not to scale) where they deemed something to be of particular importance. Hammond produced yearly printed summaries of the results of the excavations that were submitted to the Jordan Department of Antiquities. These summaries are brief and they provide information concerning important features and significant artifacts and their identity numbers. The accompanying top plans are lowresolution with a lack of detail, and they do not show the locations of the drawn sections. The field photography is generally lowresolution, and it is not fully comprehensive though a number of in situ ceramic assemblages and architectural details are wellrepresented. The compilation of finds registries in each excavation

THE AMERICAN EXPEDITION IN PETRA EXCAVATION

97

season and the method of photographing pottery sherds on trays according to area and SUs have been very helpful. The AEP excavation guidelines issued each season did not include instructions concerning sifting the soil removed from the squares. Thus, valuable material evidence, such as smaller sherds and coins, was lost during the excavation process. A major part of the work carried out in the TWLCRM project is to screen soil dumps deposited by the AEP in order to remove them and rejuvenate the landscape (Tuttle 2013: 16). One particular dump (Dump 1) that appears to have been used for soil from Area I was found to contain well over one hundred coins and large amounts of discarded sherds and some small finds. The range of dates of the ceramic and numismatic finds from Dump 1 extends from the Early Hellenistic through Early Byzantine periods mirroring the recorded finds from the AEP excavation of Area I. The tightly controlled excavation in the two squares of Area I carried out by Russell between 1974 and 1977 proved to be more successfully documented than the much larger area of the temple and its ancillary structures in Areas II–V located further west. Russell eventually made several revisions in the numbering of the archaeological phases in Area I and he also revised the dating of some of the phases (see Appendices 1 and 2). He expressed dissatisfaction with the excavation methods that he was required to use in Area I, as well as surveying problems, stating: I did the best of what I could with what existed, but sounding techniques, I am convinced, are poor techniques of excavation. They caused & created confusion. Also, the west sidewall drawing of ’74 was poorly executed with stones misdrawn and layers not matched with the other sidewalls. It was also drawn as if the excavation unit was 9 meters, 20 centimeters in length, which it was not. I’ve done what I can. 3

This major error in measuring the two excavation squares prior to the excavation resulted in the misplacement of the central east-west baulk southwards. The mistake was later rectified but it created Field Notebook Area I.2W 1975 (Notebook for West of Wall 3A,B,C), P. 113. 3

98

TALI ERICKSON-GINI & CHRISTOPHER A. TUTTLE

confusion in recording finds from the most critical location in the Early House: the antechamber. The antechamber contained numerous complete vessels of the post-annexation phase that were discovered in situ.

EXCAVATION OF THE EARLY HOUSE – RUSSELL’S STRATIGRAPHICAL PHASES XVII-XVI

The Original Form of the Early House: The earliest evidence of architectural remains in Area I appeared with the construction of the Early House in Phase XVII sometime in the 1st c. CE. The original construction of the Early House was of a high quality with ashlar walls. The original walls of the structure are W3c (13c) and W7 (62) in Areas I.2 and I.5 respectively, both of which were built along a north-south axis. 4 A third wall of this phase was revealed in Area I.5N along an east-west axis: W5 (155). The ashlar-built north-south wall, W7 (62) survived to height of at least seven courses, 1.75 m. in height. In the Byzantine period, the western wall of the Later House, W3a (13a), was built directly over W7 (Section 1). According to Russell, the fate of the original ashlar-built structure was unclear. In the reconstruction phase (Phase XVI), new walls built of rough, irregular stones (rubble) and bonded with a mud mortar were built up against the earlier ashlar walls of Phase XVII. The thickness of the walls average between 0.25 and 0.30 m. The Final Form of the Early House: The final form of the Early House (Phase XVI) was revealed in the western side of Areas I.2 and I.5. Walls from this phase include, in Area I.2: W8 (65), W9 (82), W12 (125), and in Area I.5N: W7 (164), W9 (183) 5 and W8 (52). The final form consisted of at least two interior rooms (Rooms 1 and 2), with Room 1, situated directly south of Room 2, During the excavation, each wall was initially assigned a number and a SU number, both of which appear in the plans and sections. In this paper, the SU number of the walls will appear in parentheses next to the original wall number. Features and stratigraphic layers will appear with the SU designation. 5 Wall 9 (183 = I.2 Wall 30 (804). 4

THE AMERICAN EXPEDITION IN PETRA EXCAVATION

99

and a hearth located outside the southern wall of Room 1 (SU 738). Sets of stairs present in each of the interior rooms indicated that the building had an upper floor. Field photographs reveal the presence of a ‘double’ wall in Room 1 (Fig. 3) that appears to be a kind of support or revetment to the main north-south wall, W3c (13c), running through squares, Area I.5W and I.2.

Figure 3. Room 1 facing south

Upper Floor of Room 1: The upper layers above Room 1 were not disturbed by later burials to the extent of those located in Area I.5 W to the north. The plan of the later burials shows two (Burials

100

TALI ERICKSON-GINI & CHRISTOPHER A. TUTTLE

6 and 7) located above Room 1 and none over the area of the hearth. The presence of the upper floor debris was not correctly identified by Russell, who considered it to be an occupation layer dated by him as “transition Byz/Naba.” 6 A coin described as “Imperial – pre-Constantine, post 106 A.D.” (’74 AEP RI no. 85) was discovered in SU 81 together with other finds: “Finds incl [include] lrg [large] quantity of near complete, but crushed Late Nab & rouletted, cream-slipped wares.” 7 However, the remains of an upper floor with intact vessels in a number of find spots were present in the upper layers above Room 1, ca. 1.50 to 2 meters above the ground floor. A “cache” of intact vessels, including ceramic unguentaria, was discovered in this layer in SUs 91 and 115 (Figs. 7: 1–7). These and other vessels from the upper layers of Room 1 were the same types as those found on the ground floors of the Early House, particularly from the antechamber (Room 2) as will be described below. In SU 714, Russell described a “doorstep of W8 (65)”, i.e., a possible doorway at the top of the wall dividing Rooms 1 and 2 that belonged to the upper floor. A restorable plain ware bowl was found in this locus (’75 AEP RI no. 183). Parts of metal latch were discovered in the upper floor debris in SU 724 that may have originated from this upper level doorway (no access was found along the lower level of W8 (65) between Rooms 1 and 2 that could have accounted for a door latch), (’75 AEP RI no. 93). 8 The remains of a “fire pit,” possibly a brazier, was uncovered in the upper layers of Room 1 in SU 713, which was described as containing “lots of pottery” 9 and also a fragment of a miniature incense altar decorated with red painted rectangles and an eight-pointed star (’75 AEP RI 30). An intact lamp of the Middle Roman period (’75 AEP RI no. 181) was discovered in SU 719 under SU 714, a layer that is once again described by Russell as Field Notebook pg. 42: comments on SU 714. Also pg. 56. 1974 Field Notebook Site 1.2 Book 1, SU 81. 8 A carved limestone figurine, 11.19 cm. high, depicting a woman with her knees drawn up to her chin was also discovered in SU 724 (’75 AEP RI no. 150). The figurine appears to be Egyptian in style, possibly a depiction of mourning Isis. 9 1975 Field Notebook I.2W, West of Wall 3 A, B, C. pg. 39. 6 7

THE AMERICAN EXPEDITION IN PETRA EXCAVATION 101 “Byz/Naba” 10 Likewise, an intact cup (’75 AEP RI no. 185) was revealed nearby in SU 715, located under another part of SU 714, described by Russell as a “clay sand living surface with pottery on its surface,” and dated “Byz/Naba.” 11 Lower Floor of Room 1 (Fig. 3): The eastern part of the room revealed in the excavation measures ca. 6 x 9 m. The western part runs under the western baulk of Area I.2. The ground floor of the room appears to have had an earthen floor (SU 737). As noted above, Room 1 shares its northern wall, W8 (65), with the antechamber (Room 2). In the northeastern corner of the room a raised platform (SU 742), 0.25 x 0.80 m., abuts wall W9 (82). The platform is situated directly below a small niche (SU 120), 0.25 in width that was built into the back of Wall 9. Traces of plaster were revealed on the east face of W9 (82). The staircase in Room 1 (SU 740) is located in its southeast corner where it bonds with wall W10 (102). It is made of three sandstone blocks and the front faces of each block were plastered. To the south, the room was connected with an exterior space where a hearth was located (SU 738) by way of a doorway, 0.40 m. in width, constructed between walls W12 (125) and W10 (107). Russell notes that wall W12 (125) had thin sandstone slabs at its base on both sides. This wall, along with W10 (107), was plastered above the slabs. The floor of the northern half of floor of Room 1 (SU 737) was covered with a layer of soil, ca. 8 cm. in depth, over which a layer of collapsed stones (SU 724), only one course in height, was discovered. These stones appear to have been derived from the upper courses of the upper floor. A round grinding slab stone (51 cm in diameter and 11 cm thick) with short legs was found standing against the northern side of the stairway (SU 740), (Fig. 4). 12 A second stone object, described as a “tripod mortar” (51.5 cm in diameter and 10.9 cm in height) with rectangular feet and a concave top, the border of which was broken off, was revealed in SU 737 (1975 AEP RI no. 209). A third grinding stone, rectangular in shape (28 cm x 34 cm and .04 cm thick, was discovered on the Ibid. pg. 50. Ibid. pgs. 44–45. 12 Ibid. pg. 104. 10 11

102

TALI ERICKSON-GINI & CHRISTOPHER A. TUTTLE

platform of SU 742. 13 Finally, part of another oil press, round and spouted with ‘runnels’ (55 cm. in diameter and 37 cm. thick), was revealed along the north face of W12 (125). 14 The discovery of an oil press and mortars in the house and a large cache of ceramic unguentaria discovered in the antechamber (Room 2) suggests that perfumed oils were produced in the Early House. Other finds include a ceramic jug of the Middle Roman period discovered on the floor next to the southern side of the lowest stair in the corner of wall W10 (102), that was not included in the finds registry (Fig. 8:5).

Figure 4. Room with grinding stone resting against the staircase, facing south 13 14

1975 Field Notebook I.2W, West of Wall 3 A, B, C. pg. 104. Ibid.

THE AMERICAN EXPEDITION IN PETRA EXCAVATION 103 The Hearth: South of Room 1, a hearth (SU 738) was uncovered in the southwest corner of the excavation square in an area that was open to the sky. With regard to this, Russell noted that the presence of thick plaster (4–5 cm) on the southern faces of walls W12 (125) and 10 (107) as opposed the plaster on their interior faces (1 to 3 cm thick). 15 The hearth was bounded by a low partition wall (SU 739), ca. 0.30 m. high, on the north and east with the northern wall running westward into the west baulk of Site I.2. An intact oil lamp of the Middle Roman period (1975 AEP RI no. 190), (Fig. 8:12), described as “Byzantine” in the finds registry, was revealed on the floor next to the hearth in SU 738. The space eastwards, between the hearth and walls W7 (62), which underlies W3a (13a), and W3c (13c), was heavily disturbed in the Byzantine period with the construction of the cistern (SU 732) of the Later House. No burials were found above the hearth and here the hillside slopes steeply to the south, apparently due to the presence of the open space where it was constructed. The Antechamber, (Room 2), (Fig. 5): This room is generally referred to as an antechamber, but in the field notes it also appears as a storage room or closet. 16 Its small size, ca. 0.70 x 1.98 m., and its position between the staircase (SU 206) and the ashlar-built eastern wall of the building, Wall 7 (62), indicates its use as the house pantry and indeed this was the find spot for an assemblage of largely intact ceramic vessels that will be described below (Figs. 10–11). The assemblage was discovered at floor level and slightly above. Notably, the excavators recorded that: “vessels were stored in this antechamber, some on top of others (e.g., plates on plates ect.), all originally facing upright.” 17 Traces of plaster (SU 77) were extant along on the ashlar-built wall, Wall 7 (62) and on the north face of Wall 8 (65) in Room 2. Wall W8 (65), which forms the southern wall of the antechamber, survived to a height of six courses, 1.60 m. high. The wall runs west under the western baulk of Area I.2. A wide niche (SU 743), ca. 0.40 cm. wide, was constructed into north face of the wall opposite Ibid., Entry: Closing Comments: pg. 111. 1977 Field Notes, (no page number), July 11, 1977. 17 1977 Field Notes, I.2 Room II (803) continued. No page number. 15 16

104

TALI ERICKSON-GINI & CHRISTOPHER A. TUTTLE

the stairs. The back wall of the niche was made up of three stone slabs, possibly of marble, and the inner walls were plastered. A second, smaller niche, 11 cm. wide and 32 cm high was constructed into the wall to the left of the central niche. The niche is stepped with a shelf penetrating a few centimeters back into the wall. 18 An intact juglet, described by Russell as “Early Byzantine (?)” was discovered in the area of the niche (1975 AEP RI no. 188), (Fig. 8:1). 19 The excavation did not reveal any opening between Rooms 1 and 2 but one may have existed further to the west. The stairs (SU 206) present in Room 2, extend from the west baulk eastward where it bonds with wall W9 (183). The unpaved floor surface at the foot of the stairs (SU 802) and wall W9 (183), was located at a depth of ca. 3 m. below the surface of the excavation square. A similar floor surface (SU 803) was present at slightly lower level between the stairwell and the eastern wall, W7 (62). A low, narrow wall, W30 (804), formed from flat slabs of stone, bound a space between wall W9 (183) and the eastern wall W7 (62). An installation described as stone-lined pit discovered in the northwest corner of the room, was found to be a repository of a number of ceramic vessels (SU 124; SU 126), (Figs. 8:18–25). 20 Wall 7 (164), 1.75 m. long and ca. 0.52 wide, formed the northern wall of the antechamber. The top of the eastern end of the wall was damaged by a later pit. Traces of two layers of plaster were found on the southern face of the wall. The outer layer was apparently decorated with black lines, ca.0.35 wide, that were in-

1975 Field Notebook I.2W, West of Wall 3 A, B, C. Entry 7– 9.8.75, page 57. 19 Ibid. pg. 56. 20 Part of a figurine, 4.2 cm in length, was also discovered in the installation (’76 AEP RI no. 173). The object was misidentified by the excavators as an ‘Atargatis’ figurine, whereas this is actually the figurine of a young male with a raised hand (probably Harpocrates) similar to the ‘Nude Youths, Standing’ found elsewhere at Petra and described by Tuttle 2009: 400–402. 18

THE AMERICAN EXPEDITION IN PETRA EXCAVATION 105 tended to produce the semblance of a stone-built wall. The same decorated plaster was evident along the eastern wall, W7 (62). 21

Figure 5. Excavation of the antechamber, Room 2 facing southeast Masonry style decoration has since been discovered at Zantur IV (Kolb 2001). The preparation of black paint, apparently used for architectural decoration, was discovered in the ‘Painters’ Workshop’ in Areas III.7 / III.8. 21

106

TALI ERICKSON-GINI & CHRISTOPHER A. TUTTLE

Figure 6. Location of the Marble Workshop, the Painters’ Workshop and the room in Area II.IW, in the western rooms of the temple

THE AMERICAN EXPEDITION IN PETRA EXCAVATION 107

Figure 7. Top – Ceramic hoard in Su 91 and Su 115; Bottom – Finds from the Upper Floor of the Early House

108

TALI ERICKSON-GINI & CHRISTOPHER A. TUTTLE

Figure 8. Top – Finds from the upper and lower layer of the Lower Floor of the Early House; Bottom – Finds from SU 124/126 in the Lower Floor of the Early House

THE AMERICAN EXPEDITION IN PETRA EXCAVATION 109

Figure 9. Top – SU 162 I.5N; Bottom – Area I.5W SU 162 and 169 – under W2 (149)

110

TALI ERICKSON-GINI & CHRISTOPHER A. TUTTLE

Figure 10. SU 176 – Lower Floor ceramic hoard

THE AMERICAN EXPEDITION IN PETRA EXCAVATION 111

Figure 11. Top – SU 800 – Ceramic hoard on the floor of the antechamber; Bottom – SU 803 – Ceramic hoard on the floor of the antechamber

112

TALI ERICKSON-GINI & CHRISTOPHER A. TUTTLE

Figure 12. Vessels from the Painters’ Workshop

THE AMERICAN EXPEDITION IN PETRA EXCAVATION 113

Figure 13. Top – Vessels from the Marble Workshop; Bottom – Vessels and part of the unguentaria hoard from the room in Area II.1W

114

TALI ERICKSON-GINI & CHRISTOPHER A. TUTTLE

Flooring slabs were uncovered in the southeast corner of walls W7 (62) and 8 (65). The slabs had apparently been removed from the rest of the antechamber, leaving a poorly defined earthen surface (SU 800). Several intact or nearly intact ceramic vessels were found on this surface and penetrating into the layer of soil below (SU 803) which Russell considered to be the sub-floor of the room. However, some confusion in its identification is evident, for instance, an in situ jug photographed in SU 800 (July 11, 1977), (Fig. 11:6) was included in SU 803 in the finds registry (’77 AEP RI no. 103). This particular jug and eight other intact vessels collected in SU 803 were all designated as “(Late) Nab” in the 1977 finds registry (’77 AEP nos. 94–95, 97–99, 100–103). Pieces of charcoal in one spot at the bottom of SU 803 below ceramic vessels evinced the presence of a small fire pit. Although most of the broken vessels were restored, the excavators reported the presence of at least one large, four-handled storage jar, “same as 2 large store jars from the vaulted chamber of the temple,” i.e., the Painters’ Workshop. 22 This particular jar and the large jars from the Painters’ Workshop were not restored or drawn although their position in the workshop were noted. At the beginning of the excavation in Area I, the central baulk between the northern square (Area I.5W) and the southern square (Area I.2) divided this part of the antechamber from its southern side. Here, resting on the intact floor slabs, an assemblage of vessels, mainly in the form of ceramic unguentaria was discovered together in SU 176 with a single coin of Trajan (99–112 CE), (Figs. 10:26–28). 23 The ceramic vessels discovered with the hoard of unguenataria were of the same type found in the rest of the antechamber in SU 800, SU 803 and as well as in the remains of the upper floor such as the stone-lined installation (SU 124, 126) and near the surface in SU 91, SU115. The unguentaria in SU 176 were discovered with several other intact vessels that will be described below and presented in Figure 10. A layer of collapsed stones like that found near the floor level of Room 1 in SU 724 was not apparent in the antechamber although some collapse was evident in 22 23

Field Notes, July 11, 1977: Pot Removal (no page number). ’76 AEP RI no. 247, I.5W (SU 176).

THE AMERICAN EXPEDITION IN PETRA EXCAVATION 115 the layer above SU 800 in SU 144 near the northern wall of the room W7 (64). In the corresponding layer in Area I.5W SU 173, a type of molded capital was uncovered that has been discovered elsewhere in Petra (McKenzie 2005: 190, Diagram 14: L). 24 Russell proposed that following the abandonment of the Early House, a layer of debris accumulated or was dumped over the remains of Phase XIV. He noted occupational remains and pottery north of the antechamber in Area I.5W which he designated Phase XIII and dated to the mid-2nd to 3rd c. CE. This phase is represented by remnants of floors and one lone wall, Wall 2 (SU 149) located north of Room 2 of the Early House and west of the Middle House. According to Russell, these scanty remains were all that were left of the 2nd and 3rd centuries in wake of the construction of the Middle House in Phase XII, and the dismantlement of its western wall, Wall 14 (SU 228) in later periods. List of Important Loci (SUs) in the Early House SUs 91, 115, 114, 123 (Area I.2, Sondage C) Above Early House (upper floor) and (Wall 12 (SU 125)). SU 45 (Area I.2) Above the Early House (upper floor), under SU 32&51 and over SU 88&55 – Middle Roman pottery. SU 144 (Area I.5S = I.5W SU 173) Early House, layer above SU 176=SU 800 in the antechamber (Room 2). SUs 124, 126 (Area I.5W) Early House, stone-lined pit abutting the south face of W7 (164) present in SUs 144, 146 and 151 (=Area I.2 SU 722). SU 176 (Area I.5S) Early House, over the floor of the antechamber (Room 2), (=SU 800). SU 800 (Area I.2) Early House, over the earthen floor of the antechamber (Room 2), (=SU 176) SU 802 (Area I.2) Early House, earthen floor in front of stairs (SU 206) and wall niche (SU 743) next to the antechamber (Room 2). SU 803 (Area I.2) Early House, earthen floor/sub-floor of the antechamber (Room 2). This particular capital was visible on the surface outside the northwest corner of Area I in October 2014. 24

116

TALI ERICKSON-GINI & CHRISTOPHER A. TUTTLE

SU 738 (Area I.2) Early House, floor of hearth area south of Room 1. SU 714, 715, 722, 723, 724 (Area I.2) Early House, layers of fill or remains of an upper floor.

EXCURSUS: NABATAEAN CERAMIC WARES OF THE POSTANNEXATION PERIOD (TALI ERICKSON-GINI)

At this juncture a discussion about Nabataean pottery of the postannexation period (after 106 CE) is in order. The compression of the dates assigned to the material culture in Petra has produced an artificial gap that is not justifiable on either historical or archaeological grounds. This compression has resulted from an uncritical acceptance of the fineware chronology of the Zantur excavations (Schmid 2000: Abb. 97–98). By endorsing the Zantur chronology, researchers are required to accept an extremely short range of 20 years for a very large amount of ubiquitous types of pottery (designated Phase 3b) that make up the majority of the Nabataean ceramic corpus in Petra and other Nabataean sites. In addition, they are required to accept the appearance of a completely new and extensive ceramic assemblage in the space of approximately a decade and its abrupt termination twenty years later. A re-examination of the Zantur fineware chronology by the writer has revealed that it contains a number of serious difficulties. 25 The main difficulties in the Zantur chronology center on Phase 3, which covers most of the 1st through 3rd c. CE. Zantur Phase 3 is divided into three sub-phases: 3a (20–80 CE), 3b (80–100 CE) and 3c (100–150 CE). The dating of Phase 3 is based on a very small amount of datable material, for example, the main table showing the datable material (Schmid 2000: Abb. 420) shows that no coins were available to date either Phase 3a or Phase 3c. Moreover, the earliest sub-phase, 3a, with its purported range of 60 years “Problems and Solutions in the Dating of Nabataean Pottery of the Roman Period,” presented on February 20, 2014 in the 2nd Roundtable “Roman Pottery in the Near East” in Amman, Jordan on the premises of the American Center of Oriental Research (ACOR). 25

THE AMERICAN EXPEDITION IN PETRA EXCAVATION 117 was vastly underrepresented. 26 At Zantur, there appears to be little justification for the beginning dates for either Phase 3b (80 CE) or 3c (100 CE) or their terminal dates (100 CE and 150 CE respectively). No ‘clean’ loci, i.e., sealed contexts, were offered to prove the dating of Phases 3b and 3c and the contexts are mixed with both earlier (3b) and later (3c) material (ibid., 184). This raises the question as to why a terminal date of 100 CE was fixed for Sub-phase 3b. The coin evidence for Sub-phase 3b is scanty and some of the coins could date as late as 106 CE while there is a discrepancy between the dates of the coins and the imported wares, many of which date later than 100 or 106 CE. In order to date Phase 3 in Zantur, there was a heavy dependence on a very small quantity of imported fineware sherds, mainly ESA. Of the forms used, Hayes 56 is listed in both Phase 3b and 3c (ibid.) and since this particular form dates later than 150 CE (Hayes 1985: 39) the majority of the forms and motifs of both sub-phases 3b and 3c should be assigned to the later 2nd and early 3rd centuries. Painted Fineware Bowls: Two key types of Nabataean painted fineware bowls found in Petra and other Nabataean sites are of supreme importance in the dating ceramic assemblages. These are Schmid’s Typ E 2a 37 (Abb. 89 – lattice-work background) and Typ E 1b 10 (Abb. 90 – background of parallel lines). Typ E 2a 37 is the most firmly dated of the Nabatean painted fine ware tradition for it appears in the Roman siege camp (Camp F) of 73 CE at Masada (Magness 2005: Figs. 3: 1–5, 6:11, 9: 9, 10:5, 11). Contrary to the Zantur report (Schmid 2000: 100–101, n. 502–504), only the first type (Typ E 2a 37) has been reported in 1st c. contexts at Masada (see also: Yadin 1966: 224–225 and Bar Nathan’s Type M-Nab PBL2, Nos. 38–43 2006: 295–299, 302). 27 Evidence In the words of the report: “Unfortunately, so far only a few homogeneous FKs (find spots/loci) have been registered with fineware exclusively from Phase 3a. After all, if the Western Terra Sigillata form, Conspectus 20, 4 from FK 1122 (Abb. 420, 421 Nr. 43) accurately reflects the duration of Phase 3a, we can thus estimate [the period] as from 20 to 70/80 CE” (Schmid 2000: 38). 27 A case of circular dating occurred in Bar Nathan’s volume when she cited Schmid’s terminal date for his Typ E 2a 378 as the terminus ad 26

118

TALI ERICKSON-GINI & CHRISTOPHER A. TUTTLE

from sites damaged or destroyed in wake of the early 2nd c. earthquake highlight the appearance of Typ E 2a 37 (the lattice-work background) compared to the terminal date of 80 CE given for Sub-phase 3a. These sites include ʿEn Yotvata (Erickson-Gini 2012s: Fig. 5:7–8), Horvat Dafit (Dolinka 2006: 123, 126), Horvat Hazaza (Erickson-Gini forth.), and Mezad ʿEn Rahel (Erickson-Gini et al. forth.). 28 It is important to note that both the ʿEn Yotvata site and Mezad ʿEn Rahel, both located in the Wadi Arabah, were abandoned in wake of the early 2nd c. earthquake. On the other hand, the most ubiquitous type of Nabataean fineware painted bowl, Typ E 1b 10 (with the background of parallel lines), which at Zantur is confined to a very narrow range of twenty (!) years, does not appear at all in Masada but does appear consistently in assemblages throughout the 2nd c. and as late as the early 3rd c. Research into other Nabataean sites and Petra itself provide evidence that Nabataean fineware production continued robustly throughout the 2nd and early 3rd c. and that the repertoire of finewares was broad and impressive, actually surpassing that of the 1st c. CE. Whereas Nabataean fineware production is typically described as peaking in the 1st c. CE and declining in quality thereafter, Petra potters continued to produce high quality wares in even larger quantities in the later 2nd and early 3rd c. (Erickson-Gini 2010: 101–02). In this later period, the production of painted fineware bowls bifurcated with different workshops producing different qualities of wares, some of a superior quality and others of an inferior quality well into the 3rd c. This fact was particularly apparent in a pantry abandoned in the early 3rd c. that was discovered in Oboda (Avdat), (Erickson-Gini 2010: 91, Fig. 1.76c). There, painted fineware bowls of a high quality with bold designs in the form of six-pointed stars, leaves, fruit, palmettes ect., were stored with what is considered to be painted bowls of a lower quality (ibid., quem of Garrison 2 (sub-phase 2) at Masada (Bar Nathan 2006: 280). However, Schmid based the terminal date of sub-phase 3a at Zantur on the presence of Typ E 2a 378 on preliminary information and photos published by Y. Yadin (1967) and Y. Patrich (1985 and 1990), none of which, incidentally, included Typ E 1b 10. 28 ʿEn Rahel fort, A-1012, Locus 20A (courtyard), Basket 226/7.

THE AMERICAN EXPEDITION IN PETRA EXCAVATION 119 Figs. 2: 1–3). The same bowl and six-pointed star pattern was found on sherds of bowls together with the SU 176 unguentaria hoard in the antechamber of the Early House in Area I (Figs. 10:2– 3). Similar bold decoration appears on a bowl from Khirbet Tannur assigned to Phase 3c by Schmid (2013: 242–245, Fig. 18:5.21). Bowls with this kind of decoration were also found in abandoned forts and stations of the early third c. along the Incense Road, for example at Moyat ʿAwad (Erickson-Gini and Israel 2013: Figs. 27– 28). These kinds of painted fineware bowls with boldly executed motifs are categorized as Typ E 1b 10 in the Zantur report with little or no discussion (Schmid 2000: Farbtafel 4:9–12). An entire stack of six bowls of this type were discovered in an abandoned state in the ‘Marble Workshop’ below the Temple of the Winged Lions (Fig. 13:4). 29 Hammond designated this decoration as “dark red-painted,” (Hammond and Johnson 2000: 37). One bowl in the stack displays the ‘bird-eating-grapes’ motif that appears close in time to the termination of fineware production at Petra in the 3rd c. The same motif continues to appear on coarse ware bowls soon after. In spite of Hammond’s proposal that the dark red-painted ware be confined to the end of the 1st c. CE and prior to the appearance of the black-painted ware, he admitted to its presence in later phases: “The presence of dark red-painted Nabataean wares on the occupational floors also indicates that the black-painted variety did not entirely supplant the earlier type,” (ibid., 37). In this Hammond was correct, both the higher quality and lower quality painted wares were produced in the same time period, but evidence from other sites and in Petra itself has shown that this occurred in the first half of the 3rd c. and not as early as proposed by Hammond and Schmid. “Debased” Painted Fineware and Unguentaria: What has been described as “debased” painted fine ware, which corresponds to Schmid’s Dekorphase 3c, also appears in contexts of the 3rd c., fully one hundred years later than the proposed Zantur chronology of 100–150 CE. Hammond also ascribed the appearance of this kind of painted ware, which he termed “black-painted,” to the early 2nd c. (Hammond and Johnson 2000: 56). However, he also con29

’81 AEP RI nos. 39–44.

120

TALI ERICKSON-GINI & CHRISTOPHER A. TUTTLE

tended that it continued much later than the date proposed by Schmid, i.e., well into the middle of the 3rd c. as attested by its presence together with a rare Palmyrean coin of Vabalathus (271– 272), (’83 AEP RI c. 35): “black-painted Nabataean wares occurring there indicates their continuance from an earlier period.” (ibid., 77). Both Hammond and Russell were aware of this fact, for example with regards to the discovery of a coin of Philip I (244–249) and the unguentaria hoard from the room in Area II.1W (Hammond's Phases 18–19): “Occupational ceramic remains included a most homogeneous sherd assemblage of: store jars, cooking pots, jars, jugs, juglets, bowls, dark-red-painted and black painted Nabataean plates, terra sigillata, a lamp (disc: rosette and slash), unguentaria, teat-based pots. Decorations included dark-red, pink, light red, yellow, tan, brown, and black slips; wavy combing, ridging, piecrust rims; interior bowl painting, red-drip, painting and rouletting,” (2000: 67). He continues: “the occupational strata produced a large number of registered items: a Nabataean black-painted redware bowl with a palmette, zig-zag and seed design (AEP ’83 RI #20); a small Arretine (?) ware, terra sigillata bowl with inverted, rounded rim, sharply carinated body, ring base, with interior and exterior slip poorly applied and fabric not well levigated” (AEP ’83 RI #21), (ibid., 68). And finally: “the horde of unguentaria suggests revitalization of the unguent trade. The number of recovered coins also indicated increased wealth and prosperity. Nabataean blackpainted ware finally predominates this phase” (ibid., 70). Thus Hammond proposes that the range of the production of the blackpainted ware (Schmid’s Dekorphase 3c) extended from ca. 100 CE to the mid-3rd c., one hundred and fifty years in length (!). It also precludes that specific types of ceramic vessels, lamps and glass vessels associated with the black-painted ware were produced over an extremely long period of time although in actuality, this is not the case. A “Massive Increase” in Particular Types: In addition to painted wares, in the 2nd and 3rd centuries rouletting became a very popular decorative element and this element appears either alone or together with painted motifs that appear more and more frequently on closed fineware and also plain ware vessels. Three

THE AMERICAN EXPEDITION IN PETRA EXCAVATION 121 particularly ubiquitous types of fineware vessels, either lacking decoration or decorated with rouletting, correspond to Schmid’s ‘key forms’ of unpainted fineware: Gruppes 7, 8 and 9 (ibid., 7). 30 Gruppe 7 is a fineware bowl with a carinated rim and rouletted base, the rims of which Schmid noted a ‘massive increase’ in the Zantur assemblage (ibid., 10). Gruppe 8 is a small fineware carinated cup with a ring base that is often rouletted. A complete example of this cup with a rouletted base was discovered in the Oboda pantry of the late 2nd/early 3rd c. described above (Erickson-Gini 2010: Fig. 2: 9). Gruppe 9 is a larger carinated bowl with a grooved rim, a ring base and upright walls that are frequently rouletted. Although Gruppes 8 and 9 appeared less frequently in the assemblage, they were reported as appearing ‘only’ in conjunction with the Gruppe 7 vessels (Schmid 2000: 10). In spite of reports of a ‘massive increase’ in the appearance of these three groups in Sub-phase 3b (with its mere 20 year range), the excavators noted that they continue to appear in Sub-phase 3c (ibid.). 31 Notably, an ‘overwhelming preponderance’ of Gruppe 7 bowls and associated vessels of Gruppes 8–9 were discovered in EZ III (ibid., 21). Although these vessels lack the high-profile attention given to the painted fineware bowls, they are far more ubiquitous and are probably a more accurate diagnostic indicator of the post-annexation period between 150 and 225 CE. All three groups appear in impressive quantities with other vessel types described below in abandoned contexts (early 3rd c.) in sites along the Incense Road (Erickson-Gini and Hirschfeld, forthcoming). In the AEP excavations, all three groups were found in the both the upper and lower floors of the Early House as well as the concentration of Middle Roman pottery found in Area I.5N SU 162 (Figs. Gruppe 7: Typ E 1c 8 (Abb. 52–53), Typ E 1c 7 (Abb. 54–56); Gruppe 8: Typ E 3a 2 (Abb. 57–59), Typ 3b 188 (Abb. 60); Gruppe 9: Typ E 8a 94, 95, 97 (Abb. 62–65), Typ E 8a 96 (Abb. 61), (Schmid 2000: 9). 31 Examples of these all three groups are represented in the report as having been derived from FK 3, Schicht 15, which among other things, was the source of the Dekorphase 3c bowl, E 1b 10 (Abb. 92). For section see Abb. 412. 30

122

TALI ERICKSON-GINI & CHRISTOPHER A. TUTTLE

9:1–27). Many of these were designated as “late” Nabataean or even Byzantine by Hammond and Russell in their preliminary analyses of the Area I assemblages as can be seen in the card catalogs, field notes and finds registries between 1974 and 1977. Other Diagnostic Vessels in Post-Annexation Contexts: Important diagnostic vessels of this period include globular decanters, which appear variously as undecorated, with rouletted decoration, painted decoration or both together. These decanters have either narrow or very wide necks, like the one discovered in the Oboda pantry, and they generally share a kind of flattened, triangular rim that often survive in assemblages of the late 2nd and early 3rd c., making it a particularly important dating tool (Cohen 1987: 29; Erickson-Gini 2010: 2:43). In the AEP excavations, three examples of this kind of rim was revealed in a locus containing Middle Roman wares of the late 2nd/early 3rd c. in Area I.5N in SU 162 (Figs. 9:12–14). Although the majority are lack decoration, they sometimes appear with drippy, dark decoration that is considered a hallmark of Schmid’s Sub-phase 3c. 32 In spite of their size and fragility, intact and partially intact globular decanters have been found in abandoned assemblages elsewhere in Petra, such as the three examples from the ceramic assemblage in the Painters’ Workshop (Figs. 12:15–17) and an example from the room in Area II.1W (Fig. 13:18). One decanter from the Painters’ Workshop (Fig. 12:16) is heavily decorated with rouletting, which is a hallmark of finewares of the later 2nd and early 3rd c. CE. Yet another globular decanter with traces of tan and black paint was discovered in the AEP 2000 season in Area III.10, SU 11 (RI. 37). In this instance, the decanter was dated to 363 CE in the finds registry. A second type of decanter or jug is barreled-shaped with thick walls. Multiple complete examples of this type of jug were discovered at Moyat ʿAwad and Shaʿar Ramon in abandoned contexts of the early 3rd c. (Cohen 1987: 29). It appears in various sizes and it is often burnished and sometimes burnished and rouletted (Erickson-Gini and Israel 2013: Fig. 29). Indeed, decanters with variations of this type of rim appear at Zantur as Typ G13b 53 and G13 b 54 (Abb. 310–312), which appeared in both sub-phases 3b and 3c (Schmid 2000: 74). 32

THE AMERICAN EXPEDITION IN PETRA EXCAVATION 123 Other important diagnostic vessels are decorated or undecorated fineware cups, a complete example of which was discovered in the Oboda pantry (Erickson-Gini 2010: 2:25). The painted examples are usually decorated with a running wreath and /or ‘stylized dolphin’ motifs, the former being a common motif on painted vessels at the very end of the painted ware tradition in the 3rd c. (Negev 1971: 167, right). One of the latest decorated cups was discovered in a caravanserai of the 2nd–3rd c. at Oboda (Cohen 1980: 46). Both motifs show up on heavy plain ware bowls in contexts of the latest phase at Zantur (Phase 4) and in the 363 destruction horizon of Mezad Hazeva (Fellmann-Brogli 1996: Abb. 844–846; Erickson-Gini 2010: Fig. 4:1). A number of painted ware cups were discovered in the assemblage of Area I.5N SU 162 (Figs. 9:1–4), one of which was also rouletted. Oil Lamps: A further consequence of the compression of the Sub-phase 3b range is reflected in the dating of lamps produced in the post-annexation period at Petra. The most ubiquitous type of lamp in this period is Grawehr’s Typ I.1: a round-lamp with a small filler hole and egg motif that is categorized into three sub-types that are virtually identical to one another, and which supposedly appear in 80 CE and as late as 363 CE, a range of nearly 300 years(!). The appearance of this type of lamp in the late 1st and early 2nd c. CE has not been confirmed in other excavations but they do appear frequently in later contexts, particularly in sites along the Incense Road. 33 Like many of the vessel types described above, the Typ I.1 lamp in all its sub-types is a firm diagnostic feature of the later 2nd and the 3rd c. Middle Roman lamps of this type were discovered in the debris of the upper floor of the Early House (Fig. 7:13) and on the floors of its lower floor (Fig. 8:12; 11:24) as well as in Area I.5N SU 162 (Fig. 9:23). A very similar type of lamp, but lacking decoration along the rim, Grawehr’s Typ I.2, also appears in 3rd c. contexts as attested by its presence in Area II.1W ('85 AEP RI no. 23; dated with a coin of Philip I (244–249), (Hammond and Johnson 2000: 67; Grawehr 2006: 315, nos. 288–289). A The earliest sub-type, I.1a was found at Petra in Tomb 64b, Loculus 5 in proximity with a coin of Septimius Severus (Zayadine 1982: 371). 33

124

TALI ERICKSON-GINI & CHRISTOPHER A. TUTTLE

similar case can be made for the dating of the glass vessels at Zantur, the majority of types of which appear to cluster towards the end of the 1st and early 2nd c. CE, due to the dependence on the fineware chronology. Once again, an artificial gap with a paucity of vessels was produced in the Middle Roman period (2nd– early 3rd c. CE) due to the fineware chronology (for an example see Keller 2006: 54, Typ VII.52, no. 1740).

A PREPONDERANCE OF ABANDONED VESSEL HOARDS – EPIDEMIC?

One of the most intriguing aspects of the ceramic assemblages in the latest phase of fineware production at Petra is the frequency that these same types and more appear in vessel hoards in abandoned contexts. Such contexts are particularly noticeable sites abandoned in the early 3rd c. along the Petra-Gaza road, also called the Incense Road, such as Moyat ʿAwad (formerly called Moa), Qasra, Mezad Neqarot, Shaʿar Ramon, Mezad Mahmal and Oboda. However, assemblages of complete or nearly complete vessels of this exact same horizon continue to be uncovered in Petra, for example the hoard of vessels in the Painters’ Workshop and the stack of painted fineware bowls referred to above in the adjoining Marble Workshop, as well as the Petra North Ridge Tomb 2. Indeed, the excavators of Tomb 2 themselves raised the possibility that the tomb was hastily hewn in order to accommodate victims of an epidemic (Bikai and Perry 2001: 65). In recent years, two assemblages of complete vessels have been uncovered in mysterious contexts inside disused water channels. The first is vessel hoard dated to the 3rd c. CE in the C4 Building outside the Qasr al-Bint as reported by Francois Renal and presented in the 2nd Roundtable “Roman Pottery in the Near East” in Amman, Jordan in 2014 (Renal, pers.com.). A molded ‘Silenus’ lamp was discovered in the assemblage that is identical to a lamp uncovered in contexts of the same period at Moyat ʿAwad (Erickson-Gini and Hirschfeld, forthcoming). 34 A second vessel A third molded ‘Silenus’ lamp, which appears to have been made from the same mold, was discovered in Aila (Aqaba), (Dolinka 2003: No. 34

THE AMERICAN EXPEDITION IN PETRA EXCAVATION 125 hoard was discovered in a disused, monumental water channel in the Upper Market in an excavation by Brown University conducted by Megan Perry in 2010. The hoard was discovered above nearly a meter of deposition over the floor of the channel and it included vessels identical to those found in the Early House and the Painters’ Workshop including painted ware bowls (Figs. 7:7; 8:10, 18, 21), fine ware bowls with carinated rims and rouletted bases (Schmids’ Gruppe 7), (Figs. 10:13, 15), carinated bowls and cups (Figs. 7:5, 8; 8:9; 10:4–7, 24; 12:1–9), fine ware jugs with rouletted decoration (Fig. 11:29), a cooking pot (Fig. 10:30), a fineware, incurved bowl with a low ring base identical to that found in SU 176 (Fig. 10:23), a globular jug with a trefoil rim and ridged neck of a type found in Middle Roman and 363 CE contexts (Erickson-Gini 2010: Fig. 36, 38) and the base of unguentarium identical to one revealed in the 3rd c. hoard in the room in Area II.1W described below (Fig. 13:11). 35 As we shall see below, the abandonment of the Early House in Area I and abandoned hoards in rooms of the Temple of the Winged Lions complex were probably the result of an epidemic that occurred sometime in the 3rd c. rather than the early 2nd c. earthquake as claimed by Russell.

VESSEL HOARDS FROM THE TEMPLE OF THE WINGED LIONS COMPLEX

Important hoards of ceramic vessels and associated objects of the Middle Roman period were discovered in the area west of the temple cella. The first hoard is that of the well-known Painters’ Workshop, located in Area III.8W, a basement room situated below the western external corridor of the temple cella. In the room opposite 29). The presence of the lamp at Aila suggests that a revision of the date of the context there is necessary. 35 The ceramic hoard was uncovered in Loci A10/A15 by M. Perry in the Petra Upper Market on behalf of Brown University 2010 under the direction of S.A. Alcock and C.A. Tuttle. Preliminary processing of the vessels was carried out by T. Erickson-Gini, however the 2010 excavation has not yet been published.

126

TALI ERICKSON-GINI & CHRISTOPHER A. TUTTLE

the workshop, the Marble Workshop was revealed, containing a hoard of painted ware bowls. The third hoard was not well publicized. It was located in a room in Area II.1W that was apparently added to the western side of the temple complex opposite the wide staircase in the later 2nd or early 3rd c. A hoard of small ceramic unguentaria were uncovered together with coins of the 2nd and 3rd c. and other vessels and objects over a stone floor (SU 70). Although the excavator assigned a 1st to early 2nd c. CE dated to the finds in the Painters’ Workshop, he admitted that the hoard included vessels that could be dated as late as the 3rd c. (Hammond 1987: 134). With regard to the Marble Workshop, the painted ware bowls stacked in the room were regarded by him as dated to the reign of Malichus II (40–70 CE), (ibid., 136). A detailed analysis of the excavation records and photographs of the sherds from each of the find spots reveals firm evidence that all three hoards were abandoned in the 3rd c. as will be shown here. The Painters’ Workshop (Figs. 12:1–24): This important group of vessels, many of which have parallels from both floors of the Early Houses, is included here in order to put forward detailed information from the AEP archives that sheds important light on the date of the hoard. The Painters’ Workshop is a covered basement room located west of the Temple, which Hammond described as part of a complex of rooms used for maintenance and repairs of the Temple (Hammond 1987: 129). It was located in Area III.8W and it shared a wall with the Marble Workshop (Area III.7). In Hammond’s latest evaluation he posited that the ceramic hoard dated to the re-modeling stage of the Temple, between 50 CE and the reign of Trajan (Hammond and Johnson 2000: 36). The hoard consisted of a pile of intact and partially intact ceramic vessels discovered sitting on the floor surface of the room (SU 104), (Fig. 12:24). These were described by Hammond as ‘standard’ Nabataean cups and bowls, some of which contained traces of paint. The majority of vessels are made up of ten cups and twelve bowls that have carinated forms like those found in the antechamber in Area I. The hoard also included five cooking pots, three storage pots, three funnels and a strainer, three large globular jugs (one rouletted) and a ribbed jug (ibid., 132–134). Large storage jars were found further away, near the entrance to the room in SU 103. The field notes report the discovery of an oil lamp in SU 105 in the same corner that does not appear in the finds registries or

THE AMERICAN EXPEDITION IN PETRA EXCAVATION 127 sherd photos. 36 Notably, Hammond refers to the fact that “flanged rim bowls with strap handles,” (casseroles?) and “black exterior painted Nabataean wares,” were discovered among the hoard (ibid., 36). It is important to note that in the Painters’ Workshop the vessels already appeared above the floor (SU 104) in a layer designated SU 101 and SU 102. Thus the excavator’s designation of the hoard as SU 104, the floor of the room, is disingenuous. The pile of vessels actually lies in the matrix of SU 101 and SU 102, the locus located directly over the floor (ibid., 130, table) and under SU 101. 37 SU 101 was described as a layer of charcoal lying throughout the room. SU 102 also contains “large bunches of charcoal,” throughout the room that apparently were derived from the burning of eaves of a wooden beams supporting a roof above this part of the room since the room was found to be lacking the vaulted ceiling in back (northwards), (Ibid., 19). The pottery vessels were initially identified as “early and Late Nabataean ware.” A coin of the Emperor Caracalla (198–217) was discovered in SU 102 along with the hoard (’77 AEP RI c.16). Similarly, a coin dated to the last quarter of the 3rd c. CE was revealed in the level above the charcoal scatter (SU 101) in SU 43 (no. 43). A third coin, apparently an imperial Roman coin, was found directly on the floor (SU 104), (’77 AEP RI c. 28). Decorative elements in the form of molded plaster faces were discovered over the floor the south in SU 103 (’77 AEP RI nos.184–191), however these same faces were also revealed in the charcoal scatter (SU 101) above the floor (1977 AEP RI 23 a/b, 112–114). Alternatively, a small, Roman period lamp appears in the 1977 finds registry (’77 AEP RI no. 180), which corresponds to Grawehr’s Typ I.9 (2006: 320, no. 315) that is listed as uncovered in Area II.8W in SU 30, no mention of which is made in the field notes of that day in spite of other details provided about the finds: 1977 Field Notes, Area II.8 (W), page 70, “Black-painted, bowl with splotchy slip. Poor Man’s Terra Sig. Store jars, cooking wares, yellow wares, tan slips.” Thus the possibility arises that the provenance of the lamp from SU 105 Area III.8 W was confused. 37 1977 Field Notes, Area III.8 (W), pg. 20. 36

128

TALI ERICKSON-GINI & CHRISTOPHER A. TUTTLE

In spite of Hammond’s dating of the Painters’ Workshop to the later 1st c. CE, the internal evidence found its excavation points to the abandonment of the assemblage and the room itself in the 3rd c. CE. The Marble Workshop (Figs. 13:1–4): This room is located at the foot of the monumental staircase to the north. The excavator describes it as “another basement-level room, adjoining that of the Painters’ Workshop” (Hammond 1987: 134) and it shares a wall with the Painters’ Workshop (W6 in Area III.7E, designated W1 in Area III.8W). Photographs taken during the excavation of Area III.7E (originally designated Area III.8W), reveal the presence of a heavy layer of collapse that evidently tumbled downward from the monumental staircase and walls above. Although the Marble Workshop shared the same occupational level as the nearby Painters’ Workshop (see Hammond 1987: 130, table), the excavator posited an earlier date (ca. 70 CE) for the contents of the room apparently on the basis of 1st c. Nabataean coins found in the room and inscriptions found on the marble slabs ('81 AEP RI nos. 16, 21, 37). 38 However, at least three other coins of the Early Byzantine period were discovered in the debris of the room (C#4, 11, 13), some of which were found together with Nabataean coins in loci both in the upper and lower layers. The lower part of a glass jug (no. 27), (Fig. 13:3) of the 2nd and early 3rd c. discovered in SU 302, which is dated “Later Roman Period (post A.D. 551)” in the 1981 finds registry. 39 Nabataean inscriptions found on some of the marble slabs were of particular importance because Hammond used them to date the entire temple complex in spite of their presence in a collection of marble fragments that could have been derived from anywhere in Petra (Hammond 1987: 136). 39 It is described thus in the ’81 finds registry: “Base of a rectangular bottle with raised design: square with rounded corners, cross lines extending through the corners.” No information was provided as to the color of the vessel, which measures 6.95 cm. by 6.67 cm. According to Keller, in Zantur this type of vessel, which corresponds to his Typ IV.3 (Vierkantflasche) of the 3rd c., indeed has rounder corners, as opposed to the sharp corners on earlier examples (2006: 52). 38

THE AMERICAN EXPEDITION IN PETRA EXCAVATION 129 Two piles of marble fragments were discovered in the room, one along the east side, which contained more than 1,000 pieces (SU 313). The stack of painted ware bowls (’81 AEP RI 39–44), (Fig. 13:4) was discovered sitting on a second pile of marble slabs in SU 318 in the southwest corner of the room. Other, nearly intact vessels were found in the room. These include a Middle Roman lantern (no. 29), (Fig. 13:1) and two plain ware bowls including a carinated bowl (Fig. 13:2), like those found in the antechamber in Area I (nos. 22–23), all of which were uncovered in SU 310. The unguentaria hoard in a store-room in Area II.1 West (Figs. 13:5–18): The room was apparently attached to the western wall of the monumental staircase after the beginning of the 2nd c. CE. This ancillary structure was made up of a room divided into two parts, both of which have a stone floor (SU 70). 40 Photographs show that the western wall of the structure survived to a height of six courses. The door frame survived to a height of three courses. Some stones in the pavement in the southern, outer room were missing and the southern end of the floor appears to have been missing or demolished in the 363 CE collapse. In 1983, a number of unique artifacts were uncovered in SU 69 overlying the stone floor (SU 70), south of the doorway. These included a scarab (’85 AEP RI 1) and part of a marble goddess figurine (no. 133) 41 together with an important hoard of ceramic unguentaria (’83 AEP RI 107– 108, 135–142) and coins of the 2nd and 3rd c., the latest of which The 1985 field notes of Area II.1W, pg. 8, note that some of the pavement stones of SU 70 were pieces of worked stone and the frieze of the temple. One example was photographed: 6/25 (1985). 41 Described as follows: “torso only; left side only. Left arm bent forward and elevated across chest; left breast shown; navel marked by hole; possible tracing of draping about hip; legs broken and missing.” No dimensions are provided but the scale shows it to be small, ca. 5 by 5 cm. The description conforms to a ceramic goddess figurines discovered in late 2nd–early 3rd c. contexts at Oboda (Erickson-Gini 2010: Fig. 3:2). Parts of similar figurines were also found in the Nabataean / Roman sites of Horvat Hazaza and Mezad Beʿer Menuha (Tuttle 2009: 435–426, Cat. # 063–064). 40

130

TALI ERICKSON-GINI & CHRISTOPHER A. TUTTLE

is that of Philip I (244–249), (Figs. 13:7–12). 42 The unguentaria are small, ranging between 7–8 and 10 cm. in height. They have noticeable variations in body and rim shape. This well-dated assemblage formed the basis for dating of the latest type of Nabataean piriform unguentarium, Form XII (Johnson 1987: 64–65). However, none appear to have been used as the ‘prototype’ of the latest form of Nabataean unguentaria, whereas according to the 1974 finds registry, vessels nos. 59–64 (ibid., 64) were all bought in Petra and were not derived from stratigraphic contexts. The unguentaria from SU 69 were found in clusters in different parts of the room and also in the adjoining space south of the floor that was excavated in 1985. In 1985, excavation in the adjoining space south of the floor (SU70) revealed a continuation of the hoards with unguentaria and coins uncovered in loci SU 73 and 74 together with bits of gold foil and a fragment of a gold chain (’85 AEP RI 4,6,8). Although the excavator dated this assemblage to the 2nd c. CE, the three unguentaria (nos. 15–16, 19) are the same as those discovered in the rest of the room on the stone pavement. 43 Other intact vessels dis’83 AEP RI #C44–C62. Coins #C54–C56, #C61–C62: Septimus Severus (193–211). Coin #C56 was minted in Rabbath-Moaba. Coin #C58: Elagabalus (218–222). Coin #C59: Philip I (244–249) minted in Antioch on the Orontes. Coin #C60: Julia Domna (173–217), minted in Rabbath-Moaba. A Palmyrean coin of Vabalathus (271–272) was recovered in the layers above the hoard in SU 67 (’83 AEP RI #C35) while a coin of Hadrian was recovered in the layer directly above SU 69 in SU 68: ’83 AEP RI #C41. 43 Two of the three unguentaria (’85 AEP RI 15 and 19) were presented as Forms X and XI respectively in Johnson's chronology (1987: 62–64). However the drawing of Form XI is not that which appears in the 1985 finds registry – apparently a mix-up with an unguentarium that appears in the 1974 finds registry (’74 AEP RI no. 31), which according to the registry, was also bought in Petra and was not derived from a controlled excavation. What is not presented is no. 16, the drawing of which attests to its close resemblance to Johnson’s latest form, found in the same room in SU 69, Form XII. The writers were unable to find the original drawings of these three vessels for presentation here. 42

THE AMERICAN EXPEDITION IN PETRA EXCAVATION 131 covered alongside include a carinated cup and bowl (nos 31, 11), a globular decanter (no. 12), (Fig. 13:18) and three plain ware bowls (nos. 13, 20–21), (Fig. 13:5) all of which have parallels in late 2nd and early 3rd c. contexts as described above. This is confirmed by the presence of sherds of Middle Roman painted fine ware bowls and a painted ware cup that were photographed from SU 73 and SU 74 (24–25.6.85), (Figs. 13:13–15). In spite of claims by the excavator, these were clearly part of the assemblage discovered in the same room in the 1983 season that was dated correctly to the 3rd c. CE.

THE DATE OF THE ABANDONMENT OF THE EARLY HOUSE AND ASSOCIATED CERAMIC ASSEMBLAGES

A detailed examination of the field notebooks, card catalogs and vessels points to the abandonment of the Early House towards the end of the 2nd or early 3rd c. CE. As noted before, Russell posited that the layers above the Early House in both Rooms 1 and the antechamber (Room 2) post-dated its abandonment and were deposited in the form of dumps and fills. However, these layers contained intact vessels and sherds of vessels of the same type of those found on the ground floor of the Early House. Rather than dumps and fills, these layers appear to have belonged to the upper floor of the Early House. Russell noted the fragments of painted plaster and intact vessels in one of the upper layers of Room 1 in SU 722. Here, ‘black painted’ Nabataean ware was found together with rouletted fine ware vessels containing burn marks. 44 The burials of later periods penetrated the upper layers of the Site I, particularly in its northwest corner (Area I.5W), creating some confusion concerning the nature of the deposits. The northwest quadrant (Area I.5W) was also junction of later wall constructions and an interface between the walls of the Early, the Middle and the Late Houses. However, one wall, W2 (149), dated by Russell to the late 2nd to late 3rd c. CE, actually sealed vessel sherds of the Middle Roman period (Area I.5W SU162, 169) like those of the Card catalog entry, AEP ’75, 9.8.75, Site I.2, SU 722: Note – “Periods: E. Byz[antine] / L[ate] Nab[ataean]?” 44

132

TALI ERICKSON-GINI & CHRISTOPHER A. TUTTLE

Early House. Similar vessels of this period was discovered in the northeast quadrant in Area I.5N in SU 169, (Figs. 9:1–27). 45 Ceramic vessels and related finds from the ground floor of the Early House Room 1: The floor of the Room 1 was largely void of in situ vessels with the exception of a globular jug with a triangular rim over the lower floor in SU 724 (Fig. 8:5), which was not included in the 1975 AEP finds registry. The Antechamber (Room 2) – SU 176 (Figs. 10:1–30): The amount of intact and nearly intact vessels discovered in Room 2 (the antechamber) was prolific. The assemblage in SU 176 included the six ceramic unguentaria (1976 AEP RI 216–221), (Figs. 10:26– 28) 46 found on the floor slabs with the Trajanic coin and other intact vessels such as two plain fine ware bowls (one deformed in the firing process), (1976 AEP RI 226–227) and three carinated fine ware cups (1976 AEP RI 222–224), (Figs. 10:10–12) that correspond to Schmid’s Gruppe 8 (see n. 26). 47 Russell initially identified the vessels as dating to the Byzantine or Roman periods. 48 An examination of the photographed and drawn sherds (SU 176, Aug. 10, 1976) reveals the presence of a fine ware bowl with a rouletted base (Fig. 10:13) and a plain carinated bowl an everted rim (Fig. 10:24) as well as a number of fineware bowls with carinated rims (Schmid’s Gruppe 7, see n. 26), (Figs. 10:13, 15, 18) and three carinated bowls with rouletted decoration (Figs. 10:8–9, 14) that correspond to Schmid’s Gruppe 8 (see n. 26). Notably, casserole with horizontal handles of the Middle Roman period (Fig. 10:25) is present in the assemblage. It appears in the 1976 photograph (Aug. 10 1976) but was not included in Russell’s 113/114 CE drawn vessels. This particular type of casserole is diagnostic of a time period that is later than what Russell claimed the assemblage to be and like the Middle Roman lamp in SU 800 (Fig. 11:24) it was also left out of the later, unfinished analysis of Not to be confused with SU 169 in I.5W located directly below wall W2 (149). 46 These were all of the same type and three are presented here. 47 All described by Russell as Late Nab[ataen] in AEP Card catalog, Nos.222–224, 10.8.76, Site I.5W, SU 176. 48 SU 176 Area I.5W in AEP ’76 Pottery Notes dated 9.8.76. 45

THE AMERICAN EXPEDITION IN PETRA EXCAVATION 133 the pottery from the Early House. This type of casserole also appears among the assemblages revealed elsewhere in the Early House such as I.5W SU 124/SU 126 (Figs. 8:24–25) and in Area I.5N SU 162 (Fig. 9:27), both of which date undisputedly to the Middle Roman period. Also from SU 176 was a globular juglet (Fig. 8:1) discovered below the niche in the antechamber (1975 AEP RI 188), which was initially dated to the 3rd c. or possibly the Early Byzantine period. 49 Eventually, Russell appears to have included it in his drawings of “113/114 AD” pottery. With regard to the cache of unguentaria (Figs. 10:26–28) they all belong to the most ubiquitous type found in contexts of the 2nd to early 3rd c. corresponding to Johnson’s Form IX and indeed, the cache was a primary source for Johnson’s dating of this particular form dated as it was on the basis of the Trajanic coin in SU 176 to “later than circa A.D. 100” (Johnson 1987: 62–63). 50 Johnson’s dating is essentially correct, however, as this type consistently shows up in in assemblages that date toward to the end of the 2nd or early 3rd c. such as in SU 176. An unguentarium with a similar body shape and neck (Fig. 7:10) was discovered nearly 2 meters above the floor of Room 1 in SU 91 (see below) and this type of elegant, elongated neck was also present in the assemblage of Middle Roman pottery in Area I.5W SU 162/SU 169 (Fig. 9:44) under W2 (149). The Antechamber – SU 800 and SU 803 (Figs. 11:1–35): Similar vessels to those discovered in SU 176 were found further north in the antechamber in SU 800 and SU 803. Vessels from SU 800 include at least four cooking pots (Figs. 11:17–18, 21–22) and eight plain fine ware bowls, three fine ware cups (one with a rouletted base, 1977AEP RI 93), (Figs. 11:1–3), one rouletted fine ware carinated bowl (Fig. 11:11) and a plain carinated bowl (Fig. 11:27). Other items include, among other things, the rim of a Middle Roman barrel-shaped juglet (Fig. 11:4), storage jar rims (Figs. 1975 Card catalog, Phase X, Room III, No. 188. Site 1.2 (SU 722), 7.8.75; Hammond 1973 #13: 28, 41 3rd c. A.D. 50 One parallel used to date Form IX, listed from the 1974 AEP finds registry is no. 46, which according to the registry was not found in a stratigraphic context but was bought at Petra for “J.D. 500”. 49

134

TALI ERICKSON-GINI & CHRISTOPHER A. TUTTLE

11:16, 19, 20, 23) and a well-preserved bone stopper (1977 AEP RI 116). A continuation of the same hoard of vessels was found in what was considered to be the sub-floor, SU 803. Here were revealed a rouletted fine ware jug (1977 AEP RI 94), (Fig. 11:29) and a closely ribbed, plain ware jug (1977 AEP RI 103), (Fig. 11:35), both of which Russell described as L[ate] Nab[ataen]. 51 Also uncovered were a ceramic unguentarium (1977 AEP RI 95), (Fig. 11:32), three plain fine ware bowls (1977 AEP RI 97–99), (Fig. 11:25–26), three fine ware cups, one of which had a rouletted base (1977 AEP RI 100–102), and two cooking pots with pointed bases and fire marks (1977 AEP RI 140, 147), (Fig. 11:34), all of which are described as Late Nabataean in the 1974 finds registry. The unguentarium (Fig. 11:32) corresponds to Johnson’s Form IX, a cache of which was discovered in the near vicinity in SU 176. According to the catalog card pottery notes, the types of wares found in SU 800 and SU 803 included some that Russell recognized as possibly belonging to the Late Roman period. 52 Russell noted that the two cooking pots from SU 803 had parallels with those of the 3rd c. from other sites. 53 The pottery notes also verify the discovery of the Middle Roman lamp present in the sherd photograph of July 7, 1977, related to above, which he records as dated possibly to the Byzantine period. 54 The Upper Floor Debris over the Antechamber: With regard to the antechamber, Russell recognized the finds from the upper layers (SUs 144 and 173) as belonging to the same phase. A small ceramic unguentarium (Fig. 8:2) discovered in SU 173, I.5W Card catalog Nos. 94 and 103, Phase X, Room II, antechamber. Site 1.2, SU 803 (803 = rubble of SU 800), 11.7.11. 52 1977 AEP Card catalog pottery notes. Area I.2 SU 800, 6.7.77. 1977 AEP card catalog pottery notes Site I.2 SU 803 (= SU 268), 11.7.77. 53 Cooking pot No. 147 – Cleveland 1960: 68–69, Fig. 6, Pl. 16 64– 70 generally for cooking pots. Early 3rd c. Parker 1987: 571 Fig. 100 #83, 534–535, mid 3rd to mid 4th. Harding 1950: 86 Pl. 26 # 105, mid-late 3rd c. CE. Cooking pot No. 140 – Harding 1950: 85, Pl 26 #9, mid-late 3rd c. CE. 54 1977 AEP Card catalog pottery notes. Area I.2 SU 800, 7.7.77: Lamp discus w/split handle, Byz? 51

THE AMERICAN EXPEDITION IN PETRA EXCAVATION 135 (= I.5W 144) belongs to the latest type of Johnson’s chronology, Form XII, dated to the period between 225 and 250 CE (Johnson 1987: 66–67). This type was discovered in the Oboda pantry (Erickson-Gini 2010: Fig. 2:38) and Russell refers to it as a ‘Late Nabataean’ type in his phasing cards. 55 Ceramic vessels in SU 144 included a painted ware bowl (Fig. 8:10) corresponding to Schmid’s 3c (2nd c. CE), a fine ware bowl with a carinated rim (Fig. 8:4), carinated plain ware bowl with an everted rim (Fig. 8:9) quite similar to a specimen from the Painters’ Workshop (Fig. 12:8), the rim of a barrel-shaped fineware jug (Fig. 8:6) and ‘Nabataean’ type cooking pots (Figs. 8:14–16). Ceramic vessels from the northern half of the antechamber (Figs. 8:18–25): A large number of intact ceramic vessels were discovered along the northern edge of the antechamber but do not seem to have been included as part of the antechamber assemblage, probably due to the manner in which the room was excavated, which was complicated by the presence of the central baulk. The remains of a stone built installation (probably a closet) along the northern side of the room in wall W7 (164) was misinterpreted as a stone-lined pit (SU 127) that contained a large amount of intact pottery vessels (SUs 124 and 126). Among the vessels found in this spot was part of a Harpocrates figurine misidentified as Atargatis (1976 AEP RI 173). Photographs of the sherds reveal the presence of a Middle Roman casserole with horizontal handles (Figs. 8:24– 25) along with painted fine ware bowls (Figs. 8:18–21) and a painted ware jug (Fig. 8:22). In a find spot next to and below the installation on the floor of the room (SU 146, I.5W) was revealed a small, bag-shaped jar with a string-cut base (1976 AEP RI 231). Russell describe the vessel as “Ro[man] / Nab[ataean] although in the 1976 AEP registry it is listed as Byzantine.” 56 Card catalog No. 215, Phase X, Room II, ante[chamber], Site I.5W SU 173 = I.5 SU 144. 56 Card catalog, Phase IX, trash dump, No. 231. Site I.5W SU 146, 7.8.76. Parallels on back of card include: Hammond 1973: #7, 28, 40 – 2nd half of the 3rd c. CE, Baramki 1935 Pl. 3 #5, base and ribbing parallel – 3rd c. CE. 55

136

TALI ERICKSON-GINI & CHRISTOPHER A. TUTTLE

Ceramic vessels and related finds from the upper floor of the Early House: Several intact vessels and other finds were discovered in the upper layers of Rooms 1 and the antechamber (Room 2). These include the same types as those found on the floor of the antechamber. Russell ascribed much of this material to the deposition of Phase IX, middle trash dumps. However, these vessels appear to have originated in the upper floors of the Early House. A plain ware bowl (1975 AEP RI 183) was uncovered in SU 714, Site I.2 in the antechamber (Fig. 7:12). This vessel is identical to the type of bowls discovered in the antechamber described above. A fine ware cup (1975 AEP RI 186), (Fig. 7:8), of the type discovered in the antechamber was revealed in SU 80, Site I.2 in an upper layer of Room 1. Similarly, a small, bag-shaped jar with a string-cut base, described as a juglet (1975 AEP RI 185), (Fig. 7:11) was found in SU 715, Site I.2. 57 It is identical to a vessel (1976 AEP RI 231) found in SU 146, I.5W (the floor of the antechamber) described above. A cache of intact vessels was discovered in Sondage C (SU 91 and 115, (Figs. 7:1–7) in levels that correspond to the upper floor of the Early House, nearly 2 meters above the floor of Room 1 (SU 737) and 0.30 m. above the remains of W12 (125). 58 The assemblage in SU 115 (Area I.2) contained the same type of wares as those discovered on the floor of the antechamber inside the Early House. They include a fine ware bowl with a carinated rim (1974 AEP RI 68), (Fig. 7:7), a rouletted, carinated fine ware bowl (1974 Parallels from card catalog, 7.8.75: Baramki 1935: Pl. 3 #5, base and ribbing parallel – 3rd c. CE; Hammond 1973: #7: 28, 40, 2nd half of the 3rd c. CE. 58 SU 115 in Area I.2 is not to be confused with a second SU, also designated 115, recorded north of the Early House in Area I.5W. These loci (both marked as SU 115) can be seen in the AEP section drawing of Area I, Site 2, 5W, West Baulk/East Face. SU 91 in Area I.2 is not pictured in this section drawing due to its location in Sondage C. However, the in situ photos of the cache of vessels (Roll 5, Aug. 1974) clearly shows intact vessels in these adjoining loci, located close to the surface of Site I.2, as described in the field notes of July 30, 1974. 57

THE AMERICAN EXPEDITION IN PETRA EXCAVATION 137 AEP RI 69), (Fig. 7:6), a carinated bowl with an everted rim (1974 AEP RI 72), (Fig. 7:5), and a fine ware beaker with a horizontal rim (1974 AEP RI 71), (Fig. 7:2). Likewise, a nearly intact unguentarium (lacking the rim), (1974 AEP RI 3) was discovered in SU 52, 0.60 m. above W8 (65) between Rooms 1 and 2 (Fig. 7:10). All of these vessels have solid parallels with vessels found in contexts dated to the late 2nd through early 3rd c. The fine ware beaker (Fig. 7:2) is a type that Hammond dated to the late 2nd c. CE as noted by Russell himself. 59 An intact cooking pot, described as Nabataean (1974 AEP RI 84), (Fig. 7:14) was uncovered next to Wall 3 (13) in SU 123 in Sondage C. Middle Roman assemblages outside of the Early House in Area I: In the northeast quadrant in SU 162, 1.5N were uncovered a number of intact vessels similar to those found in the antechamber (Figs. 9:1–27). They include a carinated bowl with a horizontal rim (1977 AEP 92 I.5N), (Fig. 9:11) that Russell described as a ‘Late’ Nabataean vessel from a Phase IX floor as well as a Nabataean lamp (1977 AEP 109 I.5N) that he ascribed (correctly) to the 3rd CE CE. The assemblage includes four examples of a wide-neck, fine ware jug with a triangular rim (Figs. 9:12–14) and painted fine ware cups (Figs. 9:1–4) that are diagnostic vessels of the late 2nd and early 3rd c. Other vessels include small jars (Figs. 9:5, 9–10), one of which exhibits painted decoration typical of the Middle Roman period. The assemblage also includes a type of thin-walled, wide-mouth jar or vase (Fig. 9:15), intact examples of which have been discovered in early 3rd c. contexts at Moyat ʿAwad (EricksonGini and Hirschfeld, forthcoming). Plain ware vessels include ribbed-neck storage jars (Figs. 9:16–17), a large storage jar (Fig. 9:18), serving and cooking pots (Figs. 9:19, 24–26), a cooking ware jug with a trefoil rim (Fig. 9:20) and part of a casserole with a horizontal handle (Fig. 9:27), both of which are clearly dated to the later 2nd and early 3rd c CE. The layers directly underlying wall W2 (149) located north of the antechamber, designated Area I.5W SUs 162 and 169, produced vessel sherds, some of which are like those found on the 1974 Card catalog, Phase IX, So. Trash Dump? No. 71, Site I.2, SU 115, 30.7.74. 59

138

TALI ERICKSON-GINI & CHRISTOPHER A. TUTTLE

floor of the antechamber. A number of these are painted fine ware vessels, bowls and cups of the Middle Roman period (Figs. 9:28– 34, 38–41) as well as lamps of that period (Figs. 9:36–37, 43). Photographs of the sherds reveal the presence of ribbed-neck storage jars and “Nabataean” type cooking pots (Fig. 9:42), all of which were produced long past the Roman annexation. A heavy concentration of pottery was reported in SU 147, I.5W (under SUs 128 / 138 and over SUs 146 / 161), close to SU 115 described above. Here was revealed a cooking pot (1976 AEP RI 243) that Russell initially dated to the Byzantine period but later changed to the Nab[ataean] / Ro[man] period. 60 Pottery assemblages of the postannexation period were also discovered in the remains of the upper floor of the Early House in SU 45 and 88. Once again, these wares are the same as the greater part of the assemblage found on the ground floor of the Early House in Rooms 1 and 2. 61 An examination of the documentation of the excavation of the area in and around the Early House provides exhaustive examples of discrepancies between the early date assigned to the destruction of the structure and the finds from upper levels generally, but not always

Card Catalog no. 243, SU 147, Site I.5W, 3.8.76. Parallels on back: Hammond 1973: 27, #1; 39 #1 – 1st – 3rd c. CE. 61 AEP 1974 card catalog entry for Area I.2: “SU 45 (SU 88) NW, Common w[are] / high con[centration] [of] Nab[ataean] & Roman fine ware. 1/2 black painted bowl with leaf pattern (stolen). Heavy store jars, scant-ribbed and combed, cream slips. Most red ware – vertical loop handles including thin strap – ring bases – rouletted base (no string cut). Terra Sig[latta] – unguentarim neck – Nab[ataean] early & late. Rom-NabByz? NB Byz probably contaminated from burials – should be clean Rom[an] – Nab[ataean].” AEP 1974 card catalog entry for Area I.2, 20.7.74: “SU 45&88. Common ware – Byzantine storage jars, ribbed & plain – flat ribbed handles – incised handles, flat-wavy-combed – pie crust bowl decor – most red wares with cream slips – rouletted Nab plain fine, early and late Nab painted [ware] all ring bases. Nab Rom Byz. Possibly contaminated by Byz fill of burials along with T.C.” 60

THE AMERICAN EXPEDITION IN PETRA EXCAVATION 139 assigned to later periods or the result of dumping. 62 Several types of pottery vessels found on the floors of the Early House continue to show up with vessels considered to be of a much later date. 63

EVIDENCE OF AN EARTHQUAKE EVENT IN THE EARLY SECOND CENTURY CE

Russell’s misreading of the archaeological evidence led him to attribute the end of the occupation of the Early House in Phase XV to earthquake destruction that he dated to 113/114 CE based on the discovery of the single coin found in the antechamber, a brass sestertius commemorating Trajan’s alimenta italiae endowment dated to the period between 103 and 117 CE, together with the hoard of unguentaria and other ceramic vessels (Russell 1985: 40–41). Although the Early House was not destroyed and abandoned by an earthquake in the early 2nd c., evidence of earthquake damage is discernable with the renovations that took place in its final occupation in Phase XVI. Subsequent research carried out in several sites, 64 including Petra itself, indicate that an early 2nd c. earthquake did indeed take For example, the juglet from SU 722 (collapse above floor), (75RINo. 188) is included in the card catalog under heading: Phase X CE 113/114 Collapse Lowest House. 63 For example: a note on a catalog card: KR AEP ’75 SU 722 – 9/8/75 bl ptd Nbtaen Periods: E. Byz / L. Nbta ? [also rouletted ware with fire marks, painted plaster frags. bldg and wall face plaster). 64 Evidence of an earthquake at Petra in the late first or early 2nd c. CE has been uncovered by Kirkbride and Parr at Petra (Kirkbride 1960: 118–19; Parr 1960: 129; Joukowsky and Basile 2001: 50) and more recently in the az-Zantur excavations (Kolb and Keller 2002: 286; Grawehr 2007: 399). Evidence of the event has also been uncovered in sites in the surrounding region at Aqaba (Dolinka 2003: 30–32, Fig. 14), ʿEn Yotvata (Erickson-Gini 2012a), Moyat ʿAwad and Shaʿar Ramon (Cohen 1982: 243–44; Erickson-Gini and Israel 2013: 45), ʿEn Rahel (Korjenkov and Erickson-Gini 2003), Mezad Mahmal (Erickson-Gini 2011), Mampsis (Negev 1971: 166; Erickson-Gini 2010: 47), Oboda (Erickson-Gini, in press) and Horvat Hazaza (Erickson-Gini, in press). However, with regard to Khirbet Tannur, in light of the final publication and re-evaluation of 62

140

TALI ERICKSON-GINI & CHRISTOPHER A. TUTTLE

place (Erickson-Gini 2010: 47). 65 An examination of the records and photographs of the western side of the Temple of the Winged Lions complex also reveals evidence of earthquake damage that precedes that of the 363 CE earthquake. This evidence includes the blockage of doorways with architectural fragments that appear to have been derived from the temple, for instance in Area III.8 (SU 113; W2; Aug. 2, 1977), that were also used in the construction of the pavement in WII.1W. Revetments adding support to walls were photographed in Area III.7 (AEP 83900). In addition, a hoard of vessels of the late 1st c. BCE and first half of the 1st c. CE was discovered in the AEP 2000 season in a spot next to the eastern corridor in Area III.10 (SU 19). This assemblage of restorable vessels included several plain fineware, carinated bowls that correspond to later forms of Schmid’s Gruppe 5 dated to the second half of the 1st c. BCE (2000 AEP RI. 41), (2000: Abb. 41) together with early forms of his Gruppe 6 dated to the 1st c. CE (2000 AEP RI. 11), (2000: Abb. 50) and two early painted ware bowls (2000 AEP RI. 42) corresponding to Schmid’s Dekorgruppe 2a (2000: Abbs. 80– 81) dated to the end of the 1st c. BCE and early 1st c. CE. 66 In spite of the presence of these early vessels, the AEP 2000 season finds registries records nearly all of them as dating to 363 CE. Russell was correct in dating the early form of the Early House (Phase XVII) to the 1st c. CE: ceramic vessels of that period, inNelson Glueck’s excavation by J.S. McKenzie et. al. (2013), his claim that Altar 3 was built in wake of earthquake damage of the early 2nd c. (termed the 113–114 CE earthquake) appears to be untenable due to the re-dating of Period 2 at the site to the first half of the 2nd c. CE (Mckenzie 2013: 137). 65 The occurrence of an early 2nd c. earthquake has been disputed by S.G. Schmid who proposed that destruction contexts indicated in Nabataean sites were the result of Roman military actions in wake of the annexation of Nabataea in 106 CE (1997: 418–420). 66 A lamp in found in close proximity in SU 18 (2000 AEP RI. 23) corresponds to Grawhr’s Typ 3.D (75–125 CE), (2006: 293). A complete unguentarium of a type with a flat base like that found in the Nabataean fort at ʿEn Rahel was also present in SU 19 (2000 AEP RI. 44).

THE AMERICAN EXPEDITION IN PETRA EXCAVATION 141 cluding part of a Nabataean painted ware bowl, 67 Schmid’s Typ E 2a 37 (2000: 38; Abb. 89) and a Nabataean lamp (1977 AEP RI no. 100), (Figs. 11:28, 33) were discovered in SU 803 in the antechamber, which Russell describes as the ‘rubble of 800’. 68 These vessels were mixed with intact vessels of a later period that were found in both SU 803 and SU 800 and they include a Middle Roman lamp (Fig. 11:24) that for unknown reasons was not registered in the 1977 finds registry but does appear in the black and white photograph No. 2 of sherds from SU 800 in Area I.2, dated July 7, 1977. A second Nabataean lamp of the Middle Roman period (1975 RI #190), (Fig. 8:12), was discovered on the floor surface next to the hearth (SU 738) that Russell attributed to the occupation of the Early House. The Early House was obviously renovated, prior to its final form in Phase XVI, similar to other buildings discovered in Petra. Some Nabataean communities, such as Mampsis and Oboda, underwent a wave of new construction in the newly-organized Roman Province of Arabia while sites such as ʿEn Rahel and ʿEn Yotvata were destroyed and never re-built. Renovations in wake of structural damage evident in structures in many sites in the years following the annexation, as well as the construction of new buildings, point to a widespread earthquake event in southern Transjordan and the Negev in the early 2nd c. CE. The event may have influenced or even prompted the Roman annexation that occurred soon afterwards. At Petra, the Early House was not destroyed at that time but rather it was renovated and occupied until the early 3rd c. when it was abandoned, possibly in the wake of an epidemic.

Entry in 1977 Field Notebook Area I: “I.2 Room II (803) continued [no page number] vessel #27.. 40 cm from E wall, 85 cm from N wall...painted thin bowl frags...under #1 frags”. 68 Card catalog, No. 100, 11.7.77, Area I.2, SU 803. With regard to the Nabataean lamp, these continued to be produced well into the 2nd c. CE and are not to be confined solely to the 1st c. CE. 67

142

TALI ERICKSON-GINI & CHRISTOPHER A. TUTTLE

COMPARISONS WITH OBODA

The rather crude architecture of the Early House with its antechamber (pantry) and narrow stairways leading to upper floors is strikingly similar to the Middle Roman house at Oboda. The assemblage of stacked vessels in the antechamber of the Early House, situated in a space behind the stairs suggests that this room served as pantry like that uncovered in a house of the late 2nd – early 3rd c. in Oboda (Erickson-Gini 2010:101–28). Moreover, the assemblage contains ceramic vessels that match those discovered in the Oboda pantry (Room 6) as described in detail above. Large animal bones were found in both the antechamber 69 and the Oboda pantry. 70 No definitive structural damage to the Early House was reported unlike Russell’s description of the 363 CE destruction of the Middle House, The walls of the Early House were preserved to remarkable heights, suggesting that they were not robbed out for later construction material but rather that both floors of the building was abandoned and eventually collapsed. At Oboda, the rooms of the abandoned structure of the early 3rd c. CE were intentionally filled and later construction took place above the fill and along the exterior of the abandoned building. In Area I, the main northsouth, ashlar-built wall of the earliest structure, Wall 7 (SU 62), was preserved to a height of seven courses, nearly two meters high. Other walls such as Wall 8 (SU 65) and Wall 9 (SU 82), survived to heights that allowed for the preservation of wall niches and decorated plaster. Russell's contention that the vessels discovered in 1977 AEP Card catalog – Pottery Notes Site I.2, SU 800. 6.7.77. Also: 7.7.77 – ‘Lrg animal bone – teeth frags (goat – sheep?). 70 A examination of the large animal bones from the Oboda pantry has revealed that they were covered with ink inscriptions mainly in Greek or Latin but also in Nabataean script. One scapula from the Oboda pantry had a perforated corner apparently used for hanging the bone from a string like a notebook in the pantry and inked lines of script were crossed out with incised lines. An exact replica of this kind of ‘bone notebook’ was discovered in contexts of the 3rd–4th c. in a nearby farmhouse in Oboda excavated by Cohen and Negev (see Erickson-Gini 2012b: 53, Fig. 5). 69

THE AMERICAN EXPEDITION IN PETRA EXCAVATION 143 higher layers above the ground floor were deposited in “middle trash dumps,” (his Phase IX), is belied by the fact that so many vessels were intact and were the same types that were discovered on the ground floor as well. The possibility that these vessels rested on the surfaces of an upper floor was never raised, in the spite of the presence of two staircases in the building: in Room 1 and next to the antechamber (Room 2). The vessel and lamp types match those found in other abandoned hoards in Petra and elsewhere. The excavators recognized that many of these types had been uncovered in the Painters’ Workshop next to the temple, the assemblage of which is largely presented in this paper (Figs. 12:1–24).

CONCLUSIONS

The primary issue concerning the Early House is the date and manner of its abandonment. An outstanding difficulty in Russell’s phasing in Area I is the two hundred year period between the renovations that supposedly took place in the Early House in the early 2nd c. CE (Phase XVI) and the construction of the Middle House in the early 4th c. CE (Phase XII). This gap in the archaeological record is largely artificial and can be attributed to the fact that a single coin was used to date the critical ceramic assemblage found in Room 2 (antechamber) of the Early House (SU 176, 800, 803) to the very beginning of the 2nd c. 71 Rather than its destruction by earthquake in the early 2nd c., the body of evidence points to its abandonment sometime in the early 3rd c. similar to sites along the Petra–Gaza road. This detailed study of the excavation of the Early House has wider implications for dating vessel types, strata and structures in Petra and other Nabataean sites, much of which has been confined to a very narrow and untenable dating range of the 1st c. CE. In fact, Petra continued to flourish and the Nabataeans maintained their status as international traders and perfume producers under According to the 1975 fieldnotes (Area I.5S), Russell originally recorded the possibility that Room 2, termed the ‘antechamber’ (including SUs 146, 144, 800, 805 and 803) may have collapsed in the 2nd or 3rd c. CE 71

144

TALI ERICKSON-GINI & CHRISTOPHER A. TUTTLE

the protection of the Roman army long after their loss of sovereignty. The Roman annexation led to a renewed economic boom and an increase in demand for perfumed oils that was filled by households such as the Early House. This phase of prosperity was eventually impacted by crises in the Roman East in the 3rd c. that signaled the collapse of international trade and a fall in demand for perfumed oils. The many complete vessels discovered in the Early House and in vessel hoards elsewhere in Petra and sites along the Petra–Gaza, appear to have been abandoned and often covered up, probably in the wake of an epidemic in the region in the 3rd c. This event is mirrored by the evidence from the hastily hewn tomb, Tomb 2, on the Petra North Ridge (Bikai and Perry 2001: 65), which was filled with many of the same types of ceramic vessels as those presented here. An epidemic may account for the abrupt discontinuation of the production of fineware pottery and unguentaria in Petra as well as the discontinuation of locally minted city coins there and in other cities in central and southern Transjordan in the early 3rd c. (Kindler 1983:78).

POST SCRIPT – THE LATEST PHASE IN AREA I AND THE EARTHQUAKE OF 551

With regard to the fate of the latest phase of occupation in Area I (the Upper or Later House), Russell and Hammond's supposition of destruction in both Areas I and II as result of the 551 CE earthquake is not supported by subsequent archaeological in Petra nor is it supported by the material evidence found in their excavations. As noted above, Russell’s initial phasing did not include this event, but he does refer to it in a 1990 report to ACOR on the Household Excavations. 72 Hammond and Johnson (2000) later published the supposed 551 earthquake in their phasing of the Temple of the According to the 1976 field notes, the Area I members of the excavation team were already referring to the 551 earthquake as the cause of the destruction of the first building remains uncovered, i.e., the Later House during the course of the excavation (1976 FN #3, Area I.2 North / Area I.5 cont., P. 179). 72

THE AMERICAN EXPEDITION IN PETRA EXCAVATION 145 Winged Lions. In spite of this, the finds from the Later House do not include vessels or coins of the 6th c. CE and findings from the Petra Church excavations do not support the occurrence of an earthquake event in that period. 73 A 551 earthquake event is also elusive with regard to the temple in Area II. Hammond and Johnson provide a list of coins discovered there, the latest of which dates to Constantius II (337–361 CE), (2000).

ACKNOWLEDGMENTS

The writers wish to thank the anonymous readers for their comments as well as a number of people who, without their efforts, this paper would not have been possible. These include, first and foremost, Barbara Porter, the director of the American Schools of Oriental Research in Amman, who has supported, encouraged and enabled the documentation, compilation and study of the archive of the late Philip C. Hammond in ACOR and in the TWLCRM headquarters in Wadi Musa. The archive was very kindly given to ACOR by Mrs. Lin Hammond for which we are grateful. We would also like to thank the many students and volunteers who took part in the AEP excavations and carried out much of the documentation of the excavation of Area I between 1974 and 1977. We would like to especially thank William Glanzman and David Johnson for providing us with insights about its excavation. Thanks is due to Qais Twessi and Anna Dudin for their help in providing photographs and preparing the original pencil drawings of some of the pottery for this publication. Noé Michael was very helpful in translating German texts for which we are grateful. We also wish to thank many of our colleagues for their direct or indirect assistance in completing this article. They include Elena Ronza, Glenn Corbett, Stephan G. Schmid, Bernhard Kolb, François Renel, Susan E. Alcock, Alex R. Knodell, Megan Perry, Robert Wenning, Laurent Gorgerat, Benjamin J. Dolinka, Judith S. McKenzie and Jodi Magness. A special thanks is due to the editor of this volume, Walter Ward, for his initiative, dedication and patience. Russell discovered the Petra Church in the vicinity of Area I and was to have conducted its excavation before his untimely death in 1992. 73

146

TALI ERICKSON-GINI & CHRISTOPHER A. TUTTLE

BIBLIOGRAPHY

Bachmann,W., von Watzinger, C., and T. Wiegand. (1921). Wissenschaftlich Veroffentlichungen des Deutsch-turkischen DenkmalschutzKommandos. Berlin/Leipzig: de Gruyter. Bar-Nathan, R. (2006). Masada VII. The Yigael Yadin Excavations 1963–1965 Final Reports: The Pottery of Masada. Jerusalem: Israel Exploration Society. Baramki, D.C. (1936). Two Roman Cisterns at Beit Nattif. QDAP 5, 3–10. Bikai, P.M. and Perry, M.A. (2001). Petra North Ridge Tombs 1 and 2, Preliminary Report. BASOR 324, 59–78. Cohen R. (1980). Excavations in Avdat, 1977. Qadmoniot. Jerusalem: Israel Exploration Society. Cohen R. (1982). New Light on the Date of the Petra-Gaza Road. BA45, 240–47. Cohen R. (1987). Excavations at Moa. Qadmoniot XX, Nos 1–2, pp. 77–78. Jerusalem: Israel Exploration Society. (Hebrew). Dolinka B.J. (2003). Nabataean Aila (Aqaba, Jordan) from a Ceramic Perspective. Local and Intra-regional trade in Aqaba Ware during the First and Second Centuries AD. Evidence from the Roman Aqaba Project. BAR-IS 1116. Oxford: Archaeopress. Dolinka B.J. (2006). ARABIA ADQUISITA? Ceramic Evidence for Nabataean Cultural Continuity During the Antonine and Severan Periods: The Aqaba Ware from Horvat Dafit. Unpublished Ph.D. Dissertation. University of Liverpool. Erickson-Gini T. (2010). Nabataean Settlement and Self-organized Economy in the Central Negev Highlands: Crisis and Renewal. BAR-IS. 2054. Oxford: Archaepress. Erickson-Gini T. (2011). Mezad Mahmal. http://www.hadashotesi.org.il/report_detail_eng.aspx?id=1675&mag_id=118 Erickson-Gini T. (2012a). ʿEn Yotvata. http://www.hadashotesi.org.il/report_detail_eng.aspx?id=1960&mag_id=119 Erickson-Gini T. (2012b). Nabataean Agriculture: Myth and Reality. JAE 86, 50–54. Erickson-Gini T. (forthcoming). A Nabataean Roadside Temple in the Negev Highlands. Excavations in Horvat Hazaza. ʿAtiqot.

THE AMERICAN EXPEDITION IN PETRA EXCAVATION 147 Erickson-Gini T. (forthcoming). A Late Roman and Early Byzantine Residential Quarter in Oboda. ʿAtiqot. Erickson-Gini T. and Israel I. (2013). Excavating the Nabataean Incense Road. JEMAHS, 1, pp. 24–53. Erickson-Gini T. and Hirschfeld Y. (forthcoming). Rudolph Cohen’s Excavations in the Nabataean-Roman Sites along the Incense Road in the Negev Desert, 1978–1988, Final Report. IAA Monograph Series. Jerusalem: IAA. Erickson-Gini T., Israel Y., and Nahlieli D. (forthcoming). Early Nabataean Forts in the Central Negev: ʿEn Erga, ʿEn Rahel and ʿEn Ziq. ʿAtiqot. Fellmann-Brogli R. (1996). Chapter VII. Die Keramik aus den Spatromischen Bauten. In: A. Bignasca et. al. Petra – Zantur I. Ergebnisse der Schweizerisch-Liechtensteinischen Ausgrabungen 1988– 1992. Mainz: Verlag Phillip von Zabern, pp. 232–41. Grawehr M. (2006). Die Lampen der Grabungen auf ez-Zantur in Petra. Petra – Ez Zantur III.2: Ergebnisse der SchweizerischLiechtensteinischen Ausgrabungen. Mainz: Verlag Phillip von Zabern. Grawehr M. (2007). Production of Bronze Works in the Nabataean Kingdom. SHAJ 9, 397–403. Hammond P.C. (1977). The Capitals from ‘The Temple of the Winged Lions’, Petra. BASOR 226, 47–51. Hammond P.C. (1978). Excavations at Petra, 1975–1977. ADAJ 22, 81–101, 229–46. Hammond P.C. (1980). New Evidence for the 4th-Century A.D. Destruction of Petra. BASOR 238, 65–67. Hammond P.C. (1981). Ein nabatäisches Weihrelief aus Petra. Bonner Jahrbücher 180, 137–141. Hammond P.C. (1986). A Religio-Legal Nabataean Inscription from the Atargatis/Al-ʿUzza Temple at Petra. BASOR 263, 77–80. Hammond P.C. (1987). Three Workshops at Petra (Jordan). PEFQS 192, 129–141. Hammond P.C. (1991). Nabataean Settlement inside Petra. The Ancient History Bulletin 5.1/2, 36–46.

148

TALI ERICKSON-GINI & CHRISTOPHER A. TUTTLE

Hammond P.C. (1992). The Goddess of ‘The Temple of the Winged Lions’ at Petra (Jordan). In: F. Zaydine, ed. Petra and the Caravan Cities. Amman: Department of the Antiquities of Jordan, 115–127. Hammond P.C. (1996). The Temple of the Winged Lions, Petra, Jordan, 1973–1990. Fountain Hills, Arizona: Petra Publishing. Hammond P.C. (2000a). Nabataean Mason Marks and Other Archaeological Signs. PEQ 132, 123–30. Hammond P.C. (2000b). Nabataean Metallurgy: Foundry and Fraud. In: J.A. Greene, ed. The Archaeology of Jordan and Beyond, Essays in Memory of James A. Sauer. Winona Lake, MI: Eisenbrauns, pp. 145–156. Hammond P.C. (2002). A Note on a Zodiac Lamp from Petra. PEQ 134, 165–168. Hammond P.C. (2003). The Temple of the Winged Lions. In: G. Markoe, ed. Petra Rediscovered. New York: Harry N. Abrams, pp. 223–29. Hammond P.C. (2005). Temple of the Winged Lions. Minerva 16(5), 19–24. Hammond P.C. and Johnson D.J. (2000). Survey and Excavations: “Temple of the Winged Lions”, Petra, Jordan 1973–1990. Fountain Hills, Arizona: Petra Publishing Hayes, J.W. (1985). Sigillate Orientali. Enciclopedia Dell’Arte Antica, Classica E Orientale. Roma: Istitvto Della Enciclopedia Italiana. Joukowsky M.S. and Basile J.J. (2001). More Pieces in the Petra Great Temple Puzzle. BASOR 324, 43–58. Keller D. (2006). Die Gläser aus Petra. Petra – Ez Zantur III.1: Ergebnisse der Schweizerisch \-Liechtensteinischen Ausgrabungen. Mainz. Kindler A. (1983). The Coinage of Bostra. Warminster: Aries and Phillips. Kirkbride D. (1960). A Short Account of the Excavations at Petra in 1953–1956. ADAJ, 4–5, 117–22. Kolb, B. (2001). A Nabataean mansion at Petra: Some Reflections on its Architecture and Interior Design. SHAJ 7, 437–445.

THE AMERICAN EXPEDITION IN PETRA EXCAVATION 149 Kolb B. and Keller D. (2002). Swiss-Liechtenstein Excavation at Az-Zantur / Petra: The Twelfth Season. ADA, 46, 279–294. Korjenkov A. and Erickson-Gini T. (2003). The Seismic Origin of the Destruction of the Nabataean Forts of Ein Erga and Ein Rahel, Arava Valley, Israel. Archaeologischer Anzeiger 2, 39–50. Magness J. (2005). The Roman Legionary Pottery. In: B. Arubas and H. Goldfus eds,. Excavations of the Site of the Jerusalem International Convention Center (Binyanei Haʿuma): A Settlement of the Late First to Early Second Temple Period, the Tenth Legion's Kilnworks, and a Byzantine Monastery Complex. The Pottery and Other Small Finds, JRA Supplements 60. Portsmouth: JRA, pp. 69–190 McKenzie J.S. (2005). The Architecture of Petra. Oxford: Oxford University Press. McKenzie J.S. (2013). Chapter 2: Architecture and Phases. In: J.S. McKenzie et. al., The Nabataean Temple at Khirbet Et-Tannur, Jordan. Vol. 1 – Architecture and Religion. Final Report on Nelson Glueck's 1937 Excavation. Boston: American Schools of Oriental Research, pp. 39–177. McKenzie J.S., Greene J.A., Reyes A.T., Alexander C.S., Barrett D.G. Gilmour B., Healey J.F., O’Hea M., Schibille N., Schmid S.G., Wetterstrom W., and Kansa S.W. (2013). The Nabataean Temple at Khirbet Et-Tannur, Jordan. Volumes 1 and 2. Final Report on Nelson Glueck's 1937 Excavation. Boston: American Schools of Oriental Research. Negev A. (1971). Mampsis: A Report on Excavations of a Nabataeo-Roman Town. Archaeology 24 (2), 166–71. Parr P. (1960). Excavations at Petra, 1958–1959. PEQ 92, 125–35. Russell K. (1980). The Earthquake of May 19, 363 A.D. BASOR 238, 47–64. Russell K. (1985). The Earthquake Chronology of Palestine and Northwest Arabia from the 2nd through the mid-8th century A.D. BASOR 260, 37–60. Schmid S.G. (1997). Nabataean Fine Ware Pottery and the Destruction of Petra in the Late First and Early Second Century A.D. SHAJ 6, 413–420.

150

TALI ERICKSON-GINI & CHRISTOPHER A. TUTTLE

Schmid S.G. (2000). Die Feinkeramik der Nabataer. Typologie, Chronologie und kulturhistorische Hintergrunde. Petra – Ez-Zantur II, Teil 1. Mainz: Verlag Phillip von Zabern. Schmid S.G. 2013. Chapter 18: The Pottery. In: J.S. McKenzie et. al. eds., The Nabataean Temple at Khirbet Tannur, Jordan. Vol. 2 – Cultic Offerings, Vessels and Other Specialist Reports. Final Report on Nelson Glueck’s 1937 Excavation. Boston: American Schools of Oriental Research, pp. 207–318 Tuttle C.A. (2009). The Nabataean Coroplastic Arts: A Synthetic Approach for Studying Terracotta Figurines, Plaques, Vessels, and other Clay Objects. Unpublished PhD Dissertation, Brown University. Tuttle C.A. (2012). Temple of the Winged Lions Cultural Resource Management Initiative – Some First Steps Forward (2009– 2012). ACOR Newsletter, 24.2, 1–6. Tuttle C.A. (2013). Preserving Petra Sustainably (One Step at a Time): The Temple of the Winged Lions Cultural Resource Management Initiative as a Step Forward. JEMAHS 1, 1–23. Yadin Y. (1966). Masada: Herod’s Fortress and the Zealots’ Last Stand. New York: Weidenfeld & Nicolson. Zayadine F. (1982). Recent Excavations at Petra (1979–81). ADAJ 26, pp. 365–93.

EXPLORING THE EGYPTIAN “FROG” LAMPS UNEARTHED AT THE RED SEA PORT OF ROMAN AILA (AQABA) AND THE ROMAN FORTRESS OF L EGIO IV M ARTIA AT EL-LEJJŪN, JORDAN ERIC C. LAPP 1 INTRODUCTION

Over the years a handful of Egyptian “frog” lamps have been unearthed by the archaeological excavations directed by S. Thomas Parker in Jordan. As frog lamps are uncommon in Roman Arabia and Palestine, their discovery at the Red Sea port of Roman Aila (modern Aqaba) and the fortress of Legio IV Martia at el-Lejjūn is significant. This study in honor of Professor Parker’s extraordinary scholarly career sheds light on the typology, chronology, ware, and petrography of this special group of zoomorphic lamps from the land of the Nile. It further investigates the use of frog lamps among members of the Christian population at both sites. Egyptian “frog” lamps acquire their name from the representation of the amphibian in molded high relief on the upper half of the lighting vessel (fig. 1) (for a comprehensive treatment of this lamp type, see Shier 1978: 24–30; also Bailey 1988: 226–229). Deep grooves typically outline the frog’s pointed head, squat body, and The excavation of Barrack B at el-Lejjun was so long ago now, Tom, but what fond memories of friendships and discoveries made – a link to an ancient time captured in the swirling lines of a feint finger impression on a clay lamp fragment. 1

151

152

ERIC C. LAPP

bent legs with webbed feet. In some examples, small randomly pricked dots create a pattern that imitates the skin of the slimy creature, exhibiting a kind of naturalism (fig. 1) (Bailey 1988: 259, pl. 48, Q2134 EA and Q2135 EA). A palm frond replaces the frog on further lamps (Shier 1978: pl. 22, nos. 154, 156, 158–59; pl. 23, no. 166), while bosses substitute for the muddy little croaker in still others (Bailey 1988: 262–263, Q 2163–2177 EA, pl. 49). Frog lamps have a globular yet triangular-shaped body. A central fillinghole is typically positioned at the bottom of a deeply concave discus. The discus has a small diameter and is surrounded by a wide ridge. In comparison to other clay lamps, frog lamps are heavy for their size and have thick walls (5 mm) which ensure higher durability. Their oil chambers are comparatively small. A thin slip or wash uniformly decorates the upper and lower halves.

Figure 1. A complete Egyptian frog lamp of the naturalistic style pricked with dots resembling the amphibian’s skin similar to that found on sherd no. 1 recovered at Roman Aila. Thirdfourth century CE. Coptic cemetery near Quarara. Top view. Not to scale. (After Menzel 1969: 89-89, no. 584, Abb. 75, 6).

EGYPTIAN “FROG” LAMPS

153

The Greek initial alpha is a common lampmaker’s mark found impressed on the base of this type (Shier 1978: 26). An alpha, for example, is found on the base of a frog lamp in The Archaeological Museum of the Johns Hopkins University (inv. no. 2331D). In some instances, a beta or palm branch marks the base (Shier 1978: 26). Given the magical powers attributed to the alphabet (Żurawski 1992: 91), the Greek letters alpha and beta may have served an amuletic purpose (Shier 1978: 6). The constellations and combinations of stars and crescents impressed on some lamp bases (Shier 1978: 6) may also have metaphorically linked the lamp to the lights of the celestial. Although found at sites from Alexandria to Nubia, frog lamps essentially originated from workshops located in Upper Egypt (Bailey 1988: 227). Sir Flinders Petrie recovered hundreds from the cemeteries at Roman Ehnysia (Petrie 1905: pls. LIII– LXXIV). Numerous examples were also excavated from the houses and temples at Karanis, Egypt (Shier 1978: nos. 77–299). Frog lamps are rarely found outside of the Nile Delta. Three, however, were uncovered at Roman Aqaba (nos. 1–3) and one at Lejjūn (no. 4), the only discoveries in the region.

Figure 2. Egyptian frog lamp fragments recovered at Roman Aqaba: (left) Area J, naturalistic frog style, early second to midthird century CE (no. 1); and (right) Area M, geometric frog style, early third century CE (no. 2). Photograph by E. Lapp.

154

ERIC C. LAPP

Figure 3. A nearly complete upper half of an Egyptian frog lamp recovered at Roman Aqaba (no. 3). Two deeply impressed palm branches decorate the shoulder. Area J. Early second to mid-third century CE. Photograph by E. Lapp.

Since Predynastic times the frog was a symbol of reproduction and regeneration (Shier 1978; Andrews 1998: 63). The amphibian was a popular amulet worn by Egyptians for centuries (Shier 1978: 24), and therefore, it is not surprising that it eventually occurs on Egyptian lamps The ancient Egyptians fully understood the lifecycle of the frog and the effect the flood waters of the Nile had on its prolific breeding and spontaneous regeneration. This observation was not lost on the Elder Pliny who, writing in the first century CE, comments on the use of the frog for healing purposes, especially against fever-induced chills and coughing (Natural History, 32.29, 88). In time, the Egyptian belief in the frog as synonymous with fertility and renewal found its way into Egyptian religion where the amphibian was deified as Heqt, the goddess of fertility and birth (Shier 1978: 24, n. 213). Women may have worn amulets symbolizing Heqt during childbirth to protect themselves and their children from harm (Petrie 1914: 12, nos. 18a-o). It is plausible that frog lamps, too, were lit to further invoke the powers of Heqt to protect the fetus against death in the last moment of birth. The goddess’ later connection with the late stages of childbirth is further reflected on some examples of frog lamps depicting two, probably even

EGYPTIAN “FROG” LAMPS

155

dead fetuses as suggested by their tightly closed eyelids (Bailey 1988: 264, Q2183–2185, pl. 50). Frog lamps with fetus-imagery may have accompanied the burial of infants lost in childbirth. These lamp-photoamulets were likely intended to protect infants from darkness and anonymity by invoking the rebirth powers of Heqt which granted them eternal life (Lapp 2017). Eventually rebirth or resurrection became closely linked with the frog (Jacoby and Spiegelberg 1903: 215–28; Shier 1978: 24, n. 215). The phrase ‘repeating life’ was periodically expressed by using the frog hieroglyph after a personal name (Gardiner 1951: 475, I, 7; Shier 1978: 24, n. 216). Heqt’s association with resurrection evolved with the myth of Osiris, where the goddess ensured Horus’ life in birth. It is no wonder, then, that early Christians in Egypt embraced the frog as a symbol of eternal life. One Christian frog lamp bears the Greek inscription: ΕΓω [sic] ΕΙΜΙ ΑΝΑCΤΑCΙC (‘I am the resurrection’) (LeBlant 1878: 101; Shier 1978: 24). The inscription originates from the Gospel which reads: Ἐγώεἰμι ἡ ἀνάστασις καὶ ἡ ζωή (‘I am the resurrection and the life … [it continues] Anyone who believes in me will live, even though they die’) (John 11:25). The power of the amulet as protection against mortality and anonymity is reflected in the next passage: καὶ πᾶς ὁ ζῶν καὶ πιστεύωνεἰςἐμὲοὐμὴ ἀποθάνῃεἰς τὸν αἰῶνα πιστεύειςτοῦτο (‘and whoever lives by believing in me will never die’) (John 11:26). It is no coincidence that the inscription occurs on both Egyptian Christian frog amulets and frog lamps. Not only is Jesus life, he passes on life to his believers, protecting them from death. With the addition of the cross on the nozzle-bridges of several later frog lamps at Karanis (Shier 1978: 147–48, nos. 437–40), Egyptian Christians appropriated the powers of the pagan goddess Heqt in the form of a frog. Impressed crosses with leaves on the arms – and presumably Christian – also occur on the bases of two frog lamps in the collection of the British Museum (Bailey 1988: 258–59, Q2132 and Q2133, pl. 47, fig. 145). The entrenched historic association of the frog with life and fertility, and later among Egyptian Christians as a symbol of resurrection, made the frog lamp a potent photoamulet (Lapp 2017). That the Cairo Geniza and Greek magical papyri mention the burial of frogs stuffed with aggressive magical spells in tombs to curse the intended corpse (Bohak 2017: 168–169) further bolsters the plausibility of frog lamps serving a magical function when left in graves and cemeteries.

156

ERIC C. LAPP

THE AQABA AND LEJJŪN “FROG” LAMPS

All three frog lamp fragments found at Aqaba belong to Shier Type A 5.2 (“Lamps with Triangular Shaped Bodies”) (Shier 1978: 26–8, nos. 103–213). Lamps of this group are characterized by a double convex body in profile with rounded upper and lower halves. The nozzle is typically short with a rounded end and is incorporated into the body of the lamp. The discus is small and deeply convex with a central filling-hole. The frog body preserved in high relief on lamp sherd no. 1 is fashioned in a naturalistic style (fig. 2, left). Adding to the naturalism are the pricked dots resembling the amphibian’s skin. The fragment was recovered from the mudbrick church (Area J) exposed at the port. The front and hind legs found on fragment no. 2 are more geometric in execution (fig. 2, right).

Figure 4. A complete frog lamp excavated at the Roman Fortress of el-Lejjūn (no. 4). Angle Tower (NW Room). Third to fourth century CE. Photograph by E. Lapp.

This sherd was excavated in the Nabatean and Roman domestic area (Area M) of the site. Both sherds originally belonged to lamps of Shier Type A 5.2a (Shier 1978: 27). No representation of a frog is portrayed on lamp fragment no. 3 found at Aila, but rather a deeply impressed palm leaf that extends along each side of the shoulder from the nodule handle (fig. 3). A deep groove distin-

EGYPTIAN “FROG” LAMPS

157

guishes the nozzle from each side of the shoulder. No finger imprints were observed on the exterior and interior surfaces of the frog lamp sherds as are commonly found on Classic Nabatean lamps (Lapp and Nicoli 2014: 41–3, figs. 12–13). The complete frog lamp found at el-Lejjūn (no. 4) belongs to a variation characterized by a globular body with three bosses on the upper half and an undefined base (fig. 4). Any semblance of a frog found on the earlier examples is missing on these later lamps. Examples of this group are typically molded from a red or brown clay and contain some mica flakes. They, and the Lejjūn lamp, may have been manufactured in the vicinity of Thebes sometime between the third to fourth century CE (Bailey 1988: 227, 229, 262–63, Q 2163–77). The geographic distribution of frog lamps tends to be limited to Egypt where examples have been recovered at Antinoe, Athribis, Armant, Buto, Clysma (Suez), Deir el-Medineh, Edfu, Ehnasiya, Elephantine (Aswan), Helwan, Karanis, Medinet Habu, Medinet Madi, Memphis, Qan (Antaeopolis), and Quta (fig. 5) (Bailey 1988: 227– 29). The Lejjūn lamp lacks a controlled stratigraphic context and was recovered from topsoil (C.7:001) in the northwest room of the angle tower (Parker 2006: 379, fig. 16.74, no. 379). Therefore, it is difficult to determine the original intended purpose of the lamp’s lighting whether for mundane daily illumination or for magical photoamuletic use. In the early fourth century, the church historian Eusebius of Caesarea described Aila as a port frequented by ships from Egypt (Onomasticon, 166.12–15), an observation substantiated by the abundant archaeological evidence for Egyptian artifacts at the site which includes the frog lamp fragments (nos. 1–3) (Parker and Smith 2014: 16). Frog lamps have also been excavated at ClysmaQolzoum, a port contemporaneous with Aila on the Egyptian side of the Red Sea (e.g. Bruyère 1966: pl. XXXII, top row, first and second from right; second row, first from right; pl. XXXV, bottom row, second from left). Conversely, South Jordan lamps recovered in substantial quantity at Aila further occur at Clysma (Bruyère 1966: pl. XVII, top row, second from right; second row from top, first and sixth from the right). The frog lamps unearthed at Aila may have arrived there via the ships from Egypt or the Roman road between Clysma and the port. Frog lamps were surely acquired by lamp-users of various religious affiliation. Had they appealed exclusively to Christians, a significant population existed in

158

ERIC C. LAPP

Aila by 325 CE (Parker and Smith 2014: 17) to have purchased them. The Aqaba examples (nos. 1–3) were manufactured roughly in the same period. A sizable part of the garrison at Lejjūn was likely Christian no later than the sixth century, essentially when the church located in the northwest corner of the site was built (Parker 2006: 569). Although several sixth-century lamps marked with crude crosses were found in the fortress (Parker 2006: 569), the Lejjūn frog lamp is considerably earlier and its recovery from topsoil tells us little about the context where it had initially provided lighting.

Figure 5. Distribution map of the Egyptian frog lamp type.

PETROGRAPHIC CHARACTERISTICS

Using a Nikon Labophot2–Pol polarizing microscope, limited petrographic thin-section analysis was conducted on two frog lamp sherds excavated at Aqaba (nos. 1–2). Thin sections of the fragments were cut and polished to 30 mm thickness. Sample no. 1 was

EGYPTIAN “FROG” LAMPS

159

selected because it is characterized by a pale yellow marly fabric and belongs to the frog group fashioned in a naturalistic style. Sample no. 2 was chosen because it is fashioned in a different, red ferruginous clay fabric. Petrography revealed that both fabrics exhibit similar mineralogical components and micromorphological features (fig. 6). Each fabric contains a high abundance (20% of the matrix) of well sorted, angular, coarse silt to medium sand-size quartz grains probably added by the lampmaker as temper. Keep in mind, however, that well sorted quartz inclusions can also occur naturally in clays. Less abundant in the fabrics are the angular, very fine sand to fine sand-size plagioclase feldspar grains and the coarse silt to fine sand-size muscovite and biotite mica flakes (cf. with the petrographic characteristics of the Class 52 amphora from Egypt in Peacock and Williams 1991: 204–05, fig. 123). Mica flakes are a common characteristic of frog lamp fabrics and are also visible on the surfaces of such lighting vessels (for examples containing mica flakes, see Bailey 1988: 227–28, Q 2129–2131 EA; Shier 1978: 24, 26, nos. 127–28, 131, 133–35, 139–45, 148–49, 151–58). Mica is abundant in Egypt and small glittering flakes are commonly found in Nile silt and in various Egyptian clays (Lucas and Harris 2012: 262). The Aqaba and Lejjūn matrices further contain highbirefringence fibers of an igneous or metamorphic origin. The relatively high estimated porosity (PE) for both frog lamp samples can be explained in part by the elongated pores found in each matrix (fig. 6) (Lapp 2012: 67, table 4; 14.12 for sample no. 1 and 16.77 for sample no. 2). The pores explain the “spongy” characteristic of the fabrics. Cracks follow the contours of some of the grains. No vapor veins were identified in the frog lamp fabrics unlike those occurring in the Classic Nabatean matrices (Lapp 2012: 77, fig. 5: C,E). Rounded Terra Rossa clay nodules occur in high quantity (20% of matrix) in the fabric of sample no. 1. They exhibit high sphericity. Sample no. 2 contains crystalline hematite grains. Neither matrix contains bioclastic material.

160

ERIC C. LAPP

Figure 6. (Upper) Photomicrograph of the naturalistic-style frog lamp sample with pale yellow fabric (no. 1) showing a high abundance of well sorted, angular quartz sand grains and mica flakes in a silty groundmass. Plain polarized light (10x). (Lower) Photomicrograph of the geometric-style frog lamp sample with red fabric (no. 2) showing a high abundance of moderately sorted, angular quartz sand grains and mica flakes in a silty ferruginous groundmass. Plain polarized light (10x). Photomicrographs by E. Lapp.

EGYPTIAN “FROG” LAMPS

161

CONCLUSION

The Parker excavations have recovered four examples of frog lamps. All three Aqaba fragments belong to Shier Type A 5.2. The Lejjūn lamp, however, is a bossed type. At Aqaba, the treatment of the “skin” with pricked dots found on one sherd suggests a naturalistic style, while the unmistakable, flexed front and hind frog legs portrayed on another is indicative of the geometric style. A palm frond decorates the upper half of the third frog lamp fragment. At Lejjūn, a complete frog lamp has a globular body. The three bosses on its top half, however, hardly substitute for the amphibian in this later example. The lamps range in date from the early second to fourth centuries CE. The petrographic results suggest that well sorted, angular quartz sand was likely added as temper. The petrological composition of the fabrics is certainly reflective of the lithology of the Nile region, particularly the mica flakes that commonly occur in Nile silt. The frog lamps’ association with the resurrection would have appealed especially to Christians at Aqaba and Lejjūn. The Egyptian-made lighting vessels likely found their way to both sites by traders, soldiers, or other visitors.

CATALOG

1. Inv. No. J.19:5.23.25642.RAP97. Fig. 2, left. Table 1.GD = 4.7 cm. Incomplete. Egyptian “Frog” Type – Naturalistic; Shier Type A 5.2a, “Frog.” Early second to mid-third century CE. Findspot: Aqaba, Area J. Origin: Lower Egypt. Shoulder and discus fragment. On shoulder: small (3 mm x 2 mm), consecutive, deep groove delineates leg remnant; hand-pricked dots imitate texture of frog skin. Hand-incised, deep groove delineates rounded ridge around small deeply concave discus with central filling-hole (preserved arc, 8 mm). Fabric: Munsell 10YR7/3 very pale brown. No slip. Mica flakes (low abundance, fine sand); black grits, sub-angular (high abundance, very fine sand); frosty quartz, sub-rounded (high abundance, fine sand); Fair condition. Moldmade. Parallels, shape-type and naturalistic frog skin: Clysma-Qolzoum (Suez) (Bruyère 1966: pl. XXXII, top row, first and second from right; second row, first from right); Ehnaysa, Egypt (Petrie 1904: pl. LXIII: 55–56, 66; pl. LXIXA: F57); Karanis, Egypt (Shier 1978: 101, pl. 29, no. 239, dated to the early second to mid-third century CE); Egypt (Description de l’Egypte, 2002 : no. 17, vol. 5, pl. 78); British Museum (Bailey

162

ERIC C. LAPP

1988: 258, pl. 47, Q 2129 and Q 2132); Schloessinger Collection (Rosenthal and Sivan 1978: 60–61, nos. 242 and 244). Bibliographic Reference: Unpublished.

2. Inv. No. M.3:40.05.23563.RAP96.RAP00. Fig. 2, right. Table 1. GD = 5.1 cm. T = 6–8 mm. Incomplete. Egyptian “Frog” Type – Geometric Style; Shier Type A 5.2a, “Frog.” Early third century CE. Findspot: Aqaba, Area M. Origin: Lower Egypt. Incomplete, upper half only (45% preserved). Shoulder and discus fragment. Wide shoulder. On shoulder: two frog legs in high relief (complete, arched front and hind legs with webbed feet). Small, deeply concave discus (diameter, 2 cm). Central filling-hole (diameter, 7 mm). Fabric: Munsell 2.5YR6/6 light red. No slip. Mica flakes, angular (low abundance, fine sand); black grits, sub-angular (high abundance, very fine sand). Good condition. Some flaking. No burning. Moldmade. Parallels, shape-type and legs: Ehnaysa, Egypt (Petrie 1904: pl. LXIII: 20, 28); Karanis, Egypt (Shier 1978: pl. 19, nos. 103–104, 107, 109; especially similar: no. 107 dated to the early third century CE); British Museum (Bailey 1988: 262, pl. 49, Q 2159 EA and Q 2160 MLA). Bibliographic Reference: Unpublished. 3. Inv. No. J.11: 0.356.80116.RAP00. Fig. 3. GD = 7.0 cm; W= 6.6 cm; T = 5 mm. Incomplete. Egyptian “Frog” Type; Shier Type A 5.2c, “Palm”; Petrie Class P, “Corn and Palm.” Early second to mid-third century CE. Findspot: Aqaba, Area J. Origin: Lower Egypt. Nearly complete, upper half only. Nozzle and back shoulder fragment, broken and missing. Wide body. Wide shoulder. On shoulder: palm leaf branches in high relief. Wick-hole (diameter, 7 mm). Central filling-hole, diameter 7 mm. Each side of nozzle, deep grooves. Fabric: Munsell 10YR7/3 very pale brown. No visible inclusions. Rough and “spongy” fabric under magnification. No slip. Round, degenerate nodule handle. Heavy burning around nozzle. Good condition. Moldmade. Parallels, shape-type and acanthus leaves: British Museum (Bailey 1988: 261, Q 2153 EA, fig. 135, pl. 49); Schloessinger Collection (Rosenthal and Sivan 1978: 62–63, nos. 253–255; Clysma-Qolzoum (Suez) (Bruyère 1966: pl. XXXV, bottom row, second from left); Ehnaysa, Egypt (Petrie 1904: pl. LXV: 9–10, 13–15, 86; pl. LXVI: 60); Karanis, Egypt (Shier 1978: pl. 22, nos. 154, 156, 158–159; pl. 23, no. 166; especially similar: no. 166 dated to the early second to mid-third century CE). Bibliographic Reference: Unpublished.

EGYPTIAN “FROG” LAMPS

163

4. Inv. No. 14372. Fig. 4. L = 5.5 cm; W = 4.8 cm. Complete. Egyptian “Frog” Type. Third to fourth century CE. Findspot: elLejjūn. Origin: Possibly Thebes (Lower Egypt). Three bosses on upper half. Undefined base. Fabric: Munsell 2.5YR5/8 red. Burning around wick-hole. Moldmade. Parallels, shape-type: British Museum (Bailey 1988: 262–63, Q 2163–2177 EA, pl. 49). Bibliographic Reference: el-Lejjun (Parker 2006: 379, fig. 16.74, no. 379).

BIBLIOGRAPHY

Andrews, C. (1998). Amulets of ancient Egypt. Austin: University of Texas Press. Bailey, D. (1988). Catalogue of the lamps in the British Museum III: Roman provincial lamps. London: British Museum Publications. Bohak, G. (2017). ‘Magic in the cemeteries of Late Antique Palestine’ in Expressions of cult in the Southern Levant in the Greco-Roman Period, eds. Oren Tal and Zeev Weiss, 181–196. Turnhout: Brepols. Bruyère, B. (1966). Fouilles de Clysma-Qolzoum (Suez) 1930–1932. Cairo: L’Institut Français d’Archéologie Orientale. Description de l’Egypte. (1809–1829). Köln, Paris: Taschen edn. (Reprint 2002). Eusebius.Onomasticon. (1904). Das Onomastikon der biblischen Ortsnamen. Edited by E. Klostermann. Hildsheim: G. Olms. (Reprint: 1966.) Gardiner, A. (1951). Egyptian Grammar. London: Oxford University Press. Jacoby, A. and W. Spiegelberg. (1903). Der Froschals Symbol der Auferstehungbei den Aegyptern. Sphinx 7, 215–228. Lapp, E. (2012). Water absorption analysis of Nabatean, North African and other clay lamp fabrics unearthed at the Red Sea port of Roman Aila (Aqaba, Jordan). Archaeometry 54, 1.56–79. Lapp, E.C. (2017). ‘Encountering photoamulets and the use of apotropaic light in Late Antiquity’ in The Oxford handbook on light in archaeology, eds. C. Papadopoulos and G. Earl, Oxford University Press, Oxford. Lapp, E. and J. Nicoli. (2014). Exploring 3D Modeling, Fingerprint Extraction, and Other Scanning Applications for Ancient Clay

164

ERIC C. LAPP

Oil Lamps. Digital Applications in Archaeology and Cultural Heritage 1.22, 34–44. LeBlant, E. (1878). Note sur quelques lamps égyptiennes en forme de grenouille. Mémoires de la Société nationale des antiquaires de France 39. 4th series, vol. 9. Lucas, A. and J. Harris. (2011). Ancient Egyptian Materials and Industries. Mineola: Dover. Menzel, H. (1969). Antike Lampen im Römisch-Germanischen Zentralmuseum zu Mainz. Mainz: Verlag des Römisch-Germanischen Zentralmuseums Peacock, D. and D. Williams. (1991). Amphorae and the Roman economy: an introductory guide. London: Longman. Parker, S.T. (2006). The Roman frontier in central Jordan: final report on the Limes Arabicus project, 1980–1989. Washington, D.C: Dumbarton Oaks. Parker, S.T. and Smith, A. (2014). The Roman Aqaba project final report: the regional environment and the regional survey, vol. 1, American Schools of Oriental Research Archaeological Reports 19, Boston: American Schools of Oriental Research. Petrie, W. (1905). Roman Ehnasya: Herakleopolis Magna, 1904. London: The Egypt Exploration Fund. Petrie, W. (1914). Amulets. London: Constable and Company. Pliny the Elder. (1938–1962). Natural History, translated by W. Jones, Cambridge, MA.: Harvard University Press. Rosenthal, R. and R. Sivan. (1978). Ancient lamps in the Schloessinger collection. Qedem 8. Jerusalem: The Hebrew University of Jerusalem. Shier, L. (1978). Terracotta lamps from Karanis, Egypt: excavations of the University of Michigan. Kelsey Museum of Archaeology Studies 3. Ann Arbor: The University of Michigan Press. Żurawski, B. (1992). Magica et ceramica: magic and ceramics in Christian Nubia.” Archaeologia Polona 30, 87–107.

FROM OPAQUE TO TRANSPARENT: COLORLESS GLASS AND THE BYZANTINE REVOLUTION IN INTERIOR LIGHTING JANET DUNCAN JONES 1

Most of the glass in our lives is within arm’s reach and unseen. We know it as transparent, colorless, able to take on myriad shapes – a ubiquitous material that acts as an invisible foreground to our lives. We watch our lives unfold through windows, eyeglasses, phone screens, windshields, and camera lenses. Daily life is illuminated by waves of light from electric light bulbs. We appreciate the color of a fine wine or observe chemical reactions through glass vessels molded to particular shapes and sizes. Glass is such an essential part of our lives today that it is difficult to imagine life without it. But the human/glass relationship began in a very different way. This paper reflects on moments in the evolution of that relationship – moments that led from colorful opaque glass to transparent colorless glass, from dimly lit interiors to a blaze of light, and from glass as a rarity to glass as an intimate partner in our daily lives.

For Tom, that rarest of project directors, who understands the importance of all the materials uncovered by his excavations and who demonstrates for all of us the importance of both transparency and color in work and life. And for Miles, whose light was extinguished too soon. 1

165

166

JANET DUNCAN JONES

MOMENT 1: GLASS IN NATURE

If we cast our minds back to the beginnings of the human experience of glass, we find a rare and mysterious material. Humans first came to know the dark natural glass (obsidian) that emerged from the bowels of volcanoes. This glass became a highly desirable material for the tool kits of early humans because, like flint, obsidian fractures conchoidally and was, therefore, easy to shape and held a razor sharp edge. Some of the earliest trade in human history involved obsidian. As unusual as it was as a material, obsidian was one of the first materials to shed light on the secrets of prehistoric trade allowing the sourcing of imported obsidian through trace element analysis (Renfrew et al. 1966: 30–72). As metal displaced obsidian in tool production after the fourth millennium BCE, obsidian shifted to a different niche as a prestige material for beads, pendants, seals, and small vessels, prized for its dark lustrous appearance and for the ease with which it could be formed into objects using lapidary techniques (Moorey 1994: 69–70).

MOMENT 2: MANUFACTURE OF GLASS 2

Manufactured glass most probably emerged out of Mesopotamian vitreous glaze production sometime in the late third millennium BCE. Vividly colored drips of solid glaze on the furnace floor may have led the first glassmakers to realize the potential of a material that could imitate – through the agency of mineral additives – the vivid, opaque colors of semi-precious stones like lapis lazuli, carnelian, and turquoise (Grose 1989: 45). These glasses came to be employed by secretive temple or palace monopolies to produce, through labor intensive methods, small solid inlays and objects for use at the most exclusive levels of society. Glass remained mysterious and inaccessible to most humans. Over the next millennium, this early use of glass for brightly colored inlays, beads, and pendants evolved into the production of the first glass vessels – small, opaque, vividly colored bottles made by the core-forming technique that were used primarily to hold kohl and unguents. Glass 2 For

excellent treatments of the early history of glass production see Grose 1989 and Stern 2001.

COLORLESS GLASS AND BYZANTINE INTERIOR LIGHTING 167 was valued for the first two thousand years of its production, therefore, for its ability to appear to be something else. Glass production developed, during this period, as two distinct industries: glass making and glass working. Different specialists usually did these processes at separate locations. Glass making involved production of unfinished refined glass from raw ingredients – silica from sand or crushed quartz pebbles, and flux materials (lime and soda from natron or plant ash). Color was created through the presence or addition of metallic oxides. The atmosphere of the furnace, oxidizing or reducing, could further impact the color of the final product; for example, while copper can produce red glass, if oxygen is introduced to the furnace, the final color will be green or blue. Finished glass vessels were risky trade objects, but ingots could move great distances from the original glassmaking establishment. We know that glass ingots were a traded commodity as early as the Bronze Age when Egyptian made glass was worked in places as far afield as Cyprus and the Greek mainland (Jackson and Nicholson 2010: 295–301).

MOMENT 3: VESSELS IN COLORLESS TRANSPARENT GLASS

During the centuries that followed the fall of the Bronze Age, the needle of civilization in the eastern Mediterranean and Near East was set into a different groove, and so it was with glass production. Glassmaking had ceased to be observable in the archaeological record during the ‘Dark Age’ between 1200–900 BCE. When it resumed in northern Mesopotamia and Asia Minor in the early Iron Age, the core-formed bottle returned along with something new – molded vessels produced in transparent, colorless glass. The emergence of transparency indicates several substantial shifts in both taste and technology: 1) a move from a focus on opacity to an interest in transparency; 2) a refinement in the technology of coloration to produce colorlessness rather than color; and 3) an improvement in production techniques that allowed the fashioning of sizeable transparent, colorless vessels with few visible impurities or bubbles. The driving force was most likely an interest in creating an imitation of rock crystal (pure quartz), a rare and highly valued material that was extremely hard, brittle, and, thus, difficult to work. High quality crystals of pure quartz were not unusual finds in the

168

JANET DUNCAN JONES

Near East and Asia Minor, but they were normally small, more suitable for the production of smaller objects like beads or cylinder seals (Moorey 1994: 95). At the time that colorless glass vessels come into production in the second quarter of the 1st millennium BCE, rock crystal had gained favor as a luxury material and was beginning to be used more often to make small vessels, some with ornate cut decoration (Moorey 1999: 95). Colorless glass offered an accessible and practical substitute. The colorless glass vessels produced in this early period would be the first ancestors of the ubiquitous colorless wine glasses, tablewares, and lamps of our world today – vessels that succeed, in part, by vanishing to allow the color of their contents to shine through. Ironically, the ability to create a truly colorless, transparent glass evolved out of the technology for colorizing and opacifying glass. Colorless glass can occur naturally if the batch ingredients of silica, lime, and soda are pure. This would happen, for example, if quartz pebbles were used as the source of silica in preference to sand, which contains other minerals that give the glass a pale greenish or brownish color. The rare examples of colorless glass (beads, seals) from the second millennium BCE were produced in this way. In the first millennium BCE, the use of antimony trioxide as an opacifier led to the discovery that, under certain circumstances, antimony functions as a decolorant (Reade 2012: 330). It can also serve as a fining agent, removing small bubbles from the glass. After the second century BCE, manganese oxide was the primary decolorant in Roman glass production (Reade 2012: 331). Thus, the discovery of methods of producing colorless glass using decolorants led to an upsurge in the number of transparent colorless glass objects and vessels produced. From its earliest appearance, glassworking adapted metallurgical techniques (casting in open and closed molds) and lapidary methods (cold working such as drilling, cutting, and grinding). The use of multipart molds first allowed for the production of small, solid, molded vessels, such as the Sargon Vase, into which an opening could be drilled (Grose 1989: 75–76) and later was adapted to create larger open shapes by charging multipart molds with chips of glass and heating for a long period (Stern 1994: 50–51).

COLORLESS GLASS AND BYZANTINE INTERIOR LIGHTING 169

Figure 1a and b. Interior and exterior views of molded colorless mesomphalic phiale from Tumulus P, Gordion. Inv. no. 4000 G206. D. 15.4 cm.

170

JANET DUNCAN JONES

The earliest colorless vessels are molded bowls of the eighth to the seventh centuries BCE from Nimrud and Gordion. Over one hundred colorless, mostly hemispherical, bowls from Nimrud in northern Mesopotamia, and two mesomphalic bowls from Gordion in central Anatolia, illustrate two key characteristics of early colorless transparent glass: 1) the influence of metalworking and of lapidary techniques, and 2) a concern with optical effects. The mesomphalic bowls from Nimrud and Gordion were decorated with petal decoration on both the exterior and interior in imitation of metal prototypes (Barag 1985: 67). The larger of the two Gordion examples replicates the typical Phrygian petal design well known from a group of bronze bowls found in Tumulus MM at Gordion, perhaps suggesting local production (Jones 2005: 106–108). The mesomphalic petal bowl from Tumulus P at Gordion (fig. 1a–b) also illustrates well the concern of glassmakers with the optical effects of transparent glass. The maker of the bowl appears to have paid special attention to the optical effects of the interaction between interior and exterior decoration, a particular challenge in a transparent vessel. Using cold working, the craftsman gave strong definition to transitional zones. For example, a line is incised using a lathe around the exterior of the cup at the level of the tips of the interior petals and shallow grooves have been cut to define the lower ends of the interior petals (Jones 2005: 106). In the fourth century BCE, there is a second burst of creativity in the production of colorless vessels when a renewed interest in colorless glass drinking bowls on the part of the Achaemenid Persians led to the emulation of Persian styles by subject states (Grose 1989: 81). The Greek world of the fourth century had fallen hard for the luxury wares and styles of the Achaemenid Persian overlords of Asia Minor and the use of and taste for colorless glass began to spread widely across the ancient world at this time. Persian styles of colorless drinking bowls caught the attention of the Athenian playwright, Aristophanes, who mentions Persian rock crystal cups in his Acharnians, first performed in 425 (Ignatiadou 2005: 422). Elegant drinking and serving vessels in glass of the vessel types preferred by the Persians begin to appear in greater numbers, especially in Asia Minor, in the second half of the fourth century BCE. Quantities of wide shallow phialai and deep kalyx cups decorated with combinations of rays, petals, almonds, or grooves are probably regionally produced emulations. The shapes and decora-

COLORLESS GLASS AND BYZANTINE INTERIOR LIGHTING 171 tion are, again, strongly influenced by contemporary work in metal and ceramic; they seem to be an almost conscious revival of Iron Age styles and techniques of molded glass production (Jones 2005: 108–113). The popularity of the Achaemenid style outlasted the Achaemenids themselves and flourished under Macedonian kings who worked vigorously to imitate the authority, style, and tablewares of the Persian Great Kings and of Alexander.

MOMENT 4: GLASSBLOWING

It is difficult to over-emphasize the impact of the introduction of glassblowing on the glass working industry in the first century BCE. The Romans had moved into the eastern Mediterranean in the second century BCE fueling the development of the Hellenistic glass industry with economic support in the form of political stability and wider access to resources and markets (Grose 1989, 185). Large-scale production of Hellenistic types of glass vessels continued and evolved, but an entirely new method of glassworking, glassblowing, emerged in the mid-first century BCE. The first evidence comes from Jerusalem and presages a revolution of both the creative and the economic sides of glass production. Glassblowing allowed for an entirely different economic model of glass production and consumption, far closer to mass production and mass marketing than had been possible before. The Roman Empire afforded a fertile environment for the swift development of the new industry. Glassblowing made the production of large numbers of transparent, thin-walled glass vessels possible and, by the third century CE, glass had become a commonplace in the Roman domestic sphere with production centered on the forests of northern Europe and in the Syro-Palestinian region of the eastern Empire. The most important use of glass vessels throughout the period was for drinking, but in the fourth century CE, the introduction of glass lamps, with their light-intensifying effect, served to revolutionize interior lighting.

MOMENT 5: INTERIOR LIGHTING WITH GLASS

Interior lighting was an early human technological innovation that freed human beings from the lighting cycles of the sun. The ability to illuminate interior spaces, however dimly, created new patterns of domestic life and work that extended human productive time

172

JANET DUNCAN JONES

into the night. Only forty thousand years ago, humans moved from open fires and pine torches to lamps – tallow filled cavities of limestone slabs with wicks of moss or lichen (Brox 2010: 7). Cave painters painted at Lascaux in the flickering, indistinct light of small hollowed stone lamps (Brox 2010: 9–10). Pre-ceramic societies made lamps out of what was at hand: shells, stones, even oily fish and bird carcasses (Bowers 1998: 11). In the classical world, middle and upper class Greeks and Romans fashioned small clay lamps with clay covers and wick channels; these gave a weak light and, unless carefully tended, produced more smoke than flame (Bowers 1998: 12–13). In order to cast a downward light, the wick (often dripping oil), had to be adjusted to hang over the edge of the lamp (Bowers 1998: 12–13). Oil guzzling lamps with fat wicks or multiple wicks, and lampstands for multiple lamps were used to amplify the light. Clearly, interior lighting prior to the advent of the glass lamp provided dimly lit rooms (with oily floors) filled more with smoke than with light. In the Roman world, the widespread adoption, by the third century CE, of glass tablewares had led to experimentation with the use of glass vessels as lamps. Drinking vessels with traces of oil have been found in excavations of sites like Karanis from this period onward, suggesting that cups began to be used as lamps, sometimes with suspension hooks for hanging (Stern 2001: 262). O’Hea argues that the use of the glass vessel as an oil lamp derived from the practice of using glass vessels as candle holders in liturgical contexts (O’Hea 2007: 236–239). In practical terms, where ceramic lamps were filled with olive oil, glass lamps were fueled by a layer of castor oil floating on water; a wick was floated or anchored in a wick holder. 3 Water intensified the light, which radiated in all directions with the result that the entire lamp glowed. The water kept the glass cool and extinguished the wick when the oil was consumed. Light emission studies have demonstrated that a glass lamp was nearly twice as bright as a ceramic lamp (Stern 2001: 262). This was a new experience in indoor lighting – a transparent goblet or For a detailed discussion of glass lamp typologies in the region and of the distribution of lamps with wick holders vs. lamps with loose wicks, cf. O’Hea 2012: 296. 3

COLORLESS GLASS AND BYZANTINE INTERIOR LIGHTING 173 bowl transformed into a lamp that cast light in all directions with unprecedented intensity. The effect must have been stunning. By the late fourth century, as the advantages of glass lamps over ceramic lamps became evident, glass lamps, as such, came into production in the eastern Mediterranean (Antonaras 2006: 23). The western empire seems to have lagged behind by a century or more. Quicker development in the east may have been driven by religious imperatives, both practical and theological. For example, glass lamps burned longer than ceramic lamps and thus would have proved more practical on Shabbat when lamps could not be refilled, and glass lamps allowed a person to satisfy the requirement that, in giving a Beracha (a blessing) during Havdalah (the Jewish religious ceremony marking the end of Shabbat or Jewish holidays), one could see and use the light as the blessing was being given (Stern 2001: 262). The impact of the blaze of light created by lamps in Byzantine churches is remarked upon by ancient authors such as Prudentius (Cathemerinon V) who commented upon the otherworldly effect of this illumination and pilgrims such as Egeria who, in her account of her pilgrimage to sacred places in the eastern Mediterranean in the late fourth century, remarks on the way that vast numbers of candles and glass lamps created a ‘lumen infinitum’ (Itinerarium 34.4–7). Such an investment in interior lighting, which required substantial equipment, fuel, and continual maintenance by dedicated servants, served not only to illuminate the interiors of increasingly grand structures, but also perhaps as a manifestation of Light as a metaphor for God, Christ, and the Christian message (Montserrat 1995: 438). Archaeological projects in southern Jordan – the Limes Arabicus Project (Parker 2006) and the Roman Aqaba Project (Parker 2014 and forthcoming) and the Humayma Excavation Project (Oleson and Schick 2013 and forthcoming) – have been important in illustrating the chronological range and cultural context of glass lamp usage in this region during the period in which the introduction of the glass lamp was taking place. Aila (ancient Aqaba), with its complete stratigraphic profile of domestic occupation at the site from Nabataean levels (1st century CE) through the Abbasid period (10th century), affords a glimpse of the evolution of glass in domestic contexts at the site over time (Jones forthcoming). In Roman levels at Aila, glass finds are predominately tablewares and quantities are sparse. Finds from the Byzantine period reveal an

174

JANET DUNCAN JONES

increasing range of vessel types and a nearly three-fold increase in the quantity of finds as glass became more accessible and affordable to a wider range of people (Jones forthcoming). Glassware becomes a commonplace in the domestic sphere of the eastern Empire at this time and glass lamps take a more formal place in domestic contexts. These finds give us a sense of the ways in which local populations in southern Jordan moved away from the Nabataean cultural preference for ceramics into a wider regional and more Romanized pattern of glass usage and consumption in the third century CE (Keller 2006: 183).

Figure 2a. Bell-shaped bowl or lamp with wheel-cut grooves. Transparent, colorless glass. D. 10.0 cm. Roman Aqaba Project, Area J Church.

Figure 2b. Conical drinking vessel or lamp with blue blob decoration. Transparent, colorless glass. D. 10.0 cm. Roman Aqaba Project, Area J Church.

The most concentrated finds of glass lamps of this period have come from Byzantine churches. A building at Aila, identified as an early church, reveals the kinds and quantities of lamps used in religious contexts in the fourth century CE. Typical of the period are conical, hemispherical, or bell-shaped vessels often decorated with wheel-cut grooves and/or applied blue blobs in various configurations (fig. 2a–b). Examples from the eastern empire, like those at Aila, tend to be thick-walled and colorless or very pale green. Lamps and beakers of this sort (although with regional differences in fabrics and details of production) were widely distributed throughout the eastern Empire. They are well attested in Egypt, particularly at Karanis (Harden 1936: 155–166), and throughout

COLORLESS GLASS AND BYZANTINE INTERIOR LIGHTING 175 Syro-Palestine, for example nearby at Petra ez-Zantur, where many vessels of these types were discovered, many locally made, from the fourth century onward (Keller 2006: 184), and at Jalame, where conical examples with blue blobs and hemispherical grooved bowls were manufactured in the fifth century CE (Weinberg and Goldstein 1988: 87–98, nos. 404–438 and 465–476; Slane and Magness 2005: 261). There are also a few examples of stemmed lamps from the building at Aila; such lamps were used over a long time span from the 3rd century CE to early the Islamic period (Hadad 1998: 69 Type 3). Conical, hemispherical, and stemmed lamps could not stand on their own and depended on lampstands (polycandela) (O’Hea 2007: 239) (fig. 3). The placement of multiple lamps into polycandela would create the blaze of light that would further transform the experience of interior lighting in high status structures such as churches or palaces. The heavy conical or bowl-shaped lamp type described above is the predominate type of lamp from the fourth century into the fifth centuries. It is superseded in the region by the tumbler lamp (fig. 4a), a beaker-shaped vessel in transparent natural blue green with three handles for suspension by means of bronze hangers such as those excavated at Pella in association with lamp fragments (Smith and Day 1989: 114–115, fig. 32). They often included an attached internal wick holder (fig. 4b). They are common to the many later Byzantine churches of the late fifth to sixth centuries in the region. By the sixth century, hanging lamps became a standard part of the iconography of church depiction, for example in the mosaics from the church of St. John Baptist at Gerasa (O’Hea 2012, 296). Dense layers of crushed lamps of this type have been noted in contemporary Late Byzantine/Early Islamic period churches at Lejjun (Jones 1987: 627–628; 2006, 412), Jerash (Meyer 1987: 184 and 210, fig. 12–P, 212), Khirbet al-Karak (Delougaz and Haines 1960: 49), Pella (Smith and Day 1989: 70), Humayma (Jones 2013: 529–532), Petra (O’Hea 2001: 370), and Deir ʿAin ʿAbata (O’Hea 2012: 294). Sometimes the deposit is from structural collapse as at Humayma where nearly 2,500 fragments (over a third of the total amount of glass produced by the entire site) were recovered from the floor below a cupboard in the south pastophorion (Jones 2013: 529–530). More commonly, the broken glass was deliberately buried, often in the aftermath of an earthquake like that suffered at Lejjun in 551 (Jones 2006: 412). These strata have been

176

JANET DUNCAN JONES

variously interpreted, with some proposing that such glass was destined for recycling and others wondering if, perhaps, much admired church fittings were deliberately buried to protect them from just that (O’Hea 2007: 246). What this profusion of lamps does illustrate with certainty is the increasing demand for great numbers of glass lamps to illuminate the growing number of church interiors; fragility would ensure the need for constant resupply.

Figure 3. Bronze polycandelion with glass lamps in the Collection of the Walters Art Gallery. Acc. No. 54.2466. Photo courtesy of the museum.

COLORLESS GLASS AND BYZANTINE INTERIOR LIGHTING 177

Figure 4a. Rim, handle, and upper body fragment from tumbler lamp. Transparent, blue green glass. D. 10.0 cm. Humayma Excavation Project, Area C101, Byzantine Church.

Figure 4b. Concave base fragment with attached wick holder from lamp. Transparent, blue green glass. D. base 4.0 cm. Humayma Excavation Project, Area C101, Byzantine Church.

A particularly extraordinary glass vessel from the Aila structure, a cage cup, was recovered from a room identified as the pastophorion (Jones 2003: 180–182 and forthcoming). A cage cup was a luxury vessel type most often used as a hanging lamp or drinking vessel. These vessels comprise an inner vessel and an outer cage attached to the inner vessel by thin struts. The outer cage often contained geometric shapes and an inscription in colored glass overlay. The Aila cup is preserved in three rim fragments and one body fragment (fig. 5). The main body of the vessel is extremely high quality, thick, transparent, colorless glass; the multiple colors of the outer cage – transparent blue, yellowish-amber, and green – are apparent in a horizontal band just below the shoulder level on the three joining rim fragments (Jones 2003: 181). The lamp itself would have glowed in multiple colors and the light thrown would have projected colorful patterns onto the walls, floor, and ceiling adding still another novel dimension to the experience of interior lighting. In archaeological terms, all of these successive lamp types provide useful evidence for dating and for categorizing the structures in which they are found. For example, high concentrations of glass lamps are a hallmark of public buildings, churches, and high status houses and, the fact that a large number of lamp fragments, but only a very small number of tumbler type lamp fragments were

178

JANET DUNCAN JONES

found in the Aila church helps date the structure to the fourth century. One final revolutionary dimension of lighting was the introduction of window glass in the early years of the Roman Empire, at the same time as a radical new architecture was made possible by the exploration of concrete as a building material. Window openings up to this point had served primarily as sources of light and of ventilation. Where climate demanded, windows were fitted with skins or plaster-mounted fragments of mineral like muscovite. In the middle of the first century CE, transparent glass began to appear in the windows of public and private structures. Glass windows are known from the houses of Pompeii and Herculaneum where the introduction of glass windows can be dated closely to the period between the earthquake of 62 CE and the eruption of Vesuvius in 79 CE (Dell’Acqua 2004: 111–112). Beyond merely allowing light to penetrate into interior spaces, windows could be coupled with wall decoration to create luminous effects. In addition, mirrors of obsidian were mounted on walls of houses at Pompeii (Di Pasquale 2004: 137–139) and prisms (Beretta 2004: 305) could be used to cast rainbows of color onto interior walls. In the great villas of this region, windows began to do what we in the modern world take as given – provide a link to the outside world, especially in patrician villas in the countryside (Dell’Acqua 2004: 112). Window glass is also found in quantity in the military architecture at Humayma and at Lejjun where glass window fragments were found in the principia, the barracks, the baths, and later churches (Jones 2006: 410). Windows expanded in size with the creation of large public buildings like the imperial baths where large semicircular windows, or thermal windows, at the ends of the vaulted halls kept warm rooms warm while bringing enough light in to illuminate large rooms. The continuation of this tradition of glass in architecture in areas of the Roman Empire beyond the Alps led to the rise of stained glass architecture in the Medieval period (Dell’Acqua 2004: 121). Finally, in the first century CE, transparent colorless glass fueled a different sort of revolution that is well worth noting. Magnifying (or burning) lenses and optical instruments were introduced that made observation and measurement far more precise than had previously been possible (Beretta 2004: 305). Although the idea of eyeglasses would have to wait until the Middle Ages, classical refer-

COLORLESS GLASS AND BYZANTINE INTERIOR LIGHTING 179 ences indicate that lenses were used as an aid to sight perhaps as early as Assyrian times and Pliny makes a reference to Nero’s use of an emerald to watch gladiatorial combats, perhaps for the pleasant color, but perhaps also as an aid to sight (Pliny, Natural History 37.16).

Figure 5. Four fragments of cage cup showing profile reconstuction, pattern of color overlays, and condition of fragments. Transparent colorless inner cup with overlay in D. est. 12.0 cm. Roman Aqaba Project, Area J Church.

CONCLUSION

Technology marches on. Glass has been strengthened by processes ranging from tempering to chemically strengthening to atomic reconfiguring. Silica-based glasses have been joined by a wide range of glasses with other major constituents for particular applications. In many cases, glass has been replaced with substitutes such as acrylics and synthetically produced sapphire that imitate glass the

180

JANET DUNCAN JONES

way glass began its career as a mimic to turquoise and carnelian. Glass and its transparent colorless stand-ins continue to play a central and dynamic role in daily life. As I sit in my office in the late afternoon completing this reflection on colorless glass and on light, the room is illuminated by an electric lamp with a compact fluorescent bulb bright enough that a shade is routinely necessary to soften the glare and direct the light downward onto my desk. My computer screen (no longer glass, but I can remember when it was) beams my text out at me, but I have the best possible antiglare coating on my new progressive lenses. My phone and my tablet have glass screens. One wall of my office is dominated by a great arching window with sixteen glass panes that reveal the big oak outside on the lawn, the birds on the feeder, and the roofline of the building next door. My office door has nine more panes to let me know who is in the hall. I drove here in a van with tempered glass windows. I wear a watch with a nearly indestructible glass crystal. I look through all of them without seeing them. I am the beneficiary of countless moments of discovery and invention by glassmakers and glassworkers who created some of the most colorful objects known to antiquity, but it was their effort to produce colorlessness and the resulting penetration of this material into all corners of our lives that has left the greatest legacy.

BIBLIOGRAPHY

Antonaras, A. (2006). Glass lamps of the Roman and Early Christian periods. Evidence from the Thessaloniki area. In: C.-A. Roman and R. Gudea, eds, 2008. Trade and Local Production of Lamps from the Prehistory until the Middle Age. Lychnological Acts 2. Cluj-Napoca: Editura Mega, pp. 23–30, Pl. 3–5. Barag, D. (1985). Catalogue of Western Asiatic Glass in the British Museum, volume 1. London: British Museum Publications. Beretta, M. (2004). Une révolution scientifique. In: M Beretta and G. Di Pasquale, eds, 2006. Le verre dans l’empire romain. Exposition présentée à Florence, Museo degli Argenti, “Musée “des Argents”, palais Pitti du 27 mars au 31 octobre 2004 (Arts et Sciences). Paris: Réunion des Musées Nationaux, p. 305. Bowers, B. (1998). Lengthening the Day. A History of Lighting Technology. Oxford: Oxford University Press. Brox, J. (2010). Brilliant. The Evolution of Artificial Light. New York:

COLORLESS GLASS AND BYZANTINE INTERIOR LIGHTING 181 Houghton Mifflin Harcourt. Delougaz, P. and R.C. Haines. (1960). A Byzantine Church at Khirbet al-Karak. Oriental Institute Publications 85. Chicago: University of Chicago. Dell’Acqua, F. (2004). Les Fenêtre Vitrées dans l’Antiquité romaine. In: M. Beretta and G. Di Pasquale, eds, 2006. Le verre dans l'empire romain. Exposition présentée à Florence, Museo degli Argenti, “Musée “des Argents”, palais Pitti du 27 mars au 31 octobre 2004 (Arts et Sciences). Paris: Réunion des Musées Nationaux, pp. 111–121. Di Pasquale, G. (2004). Miroirs, globes et verres ardents. In: M. Beretta and G. Di Pasquale, eds, 2006. Le verre dans l'empire romain. Exposition présentée à Florence, Museo degli Argenti, “Musée “des Argents”, palais Pitti du 27 mars au 31 octobre 2004 (Arts et Sciences). Paris: Réunion des Musées Nationaux, pp. 137–145. Egeria. Egeria. Itinerarium. Edited and translated by P. Maraval. 1997. Paris: Editions du Cerf. Grose, D.F. (1989). The Toledo Museum of Art. Early Ancient Glass: Core-Formed, Rod-Formed, and Cast Vessels and Objects from the Late Bronze Age to the Early Roman Empire, 1600 B.C. to A.D. 50. New York: Hudson Hills Press. Hadad, S. (1998). Glass Lamps from the Byzantine through Mamluk Periods at Bet Shean, Israel. Journal of Glass Studies 40, pp. 63–76. Harden D.B. (1936). Roman Glass from Karanis. Ann Arbor: University of Michigan. Ignatiadou, D. (2005). Achaemenid and Greek Colorless Glass. In: J. Curtis and S.J. Simpson, eds, The World of Achaemenid Persia. London: I.B. Tauris, pp. 419–426. Jackson, C.M and Nicholson, P.T.(2010). The Provenance of Some Glass Ingots from the Uluburun Shipwreck. Journal of Archaeological Science, 37(2), pp. 295–301. Jones, J.D. (1987). The Glass. In: S.T. Parker, ed., The Roman Frontier in Central Jordan. Interim Report on the Limes Arabicus Project, 1980–1985. Oxford: BAR-IS 340, volume 2, pp. 621–653. Jones, J.D. (2003). Recently Discovered Cage Cup Fragments from

182

JANET DUNCAN JONES

Aqaba. Journal of Glass Studies 45, 180–182. Jones, J.D. (2005). Glass Vessels from Gordion: Trade and Influence Along the Royal Road. In: L. Kealhofer, ed., The Archaeology of Midas and the Phrygians. Recent Work at Gordion. Philadelphia: University of Pennsylvania Museum Press, pp. 101–116. Jones, J.D. (2006). The Glass. In: S.T. Parker, ed., The Roman Frontier in Central Jordan. Final Report on the Limes Arabicus Project, 1980–1989, volume 2. Washington, DC: Dumbarton Oaks Research Library and Collection, pp. 393–412. Jones, J.D., 2013. Glass. In: J.P. Oleson and R. Schick, eds, Humayma, Excavation Project 2: Nabataean Campground and Necropolis, Byzantine Churches, and Early Islamic Domestic Structures. Boston: American Schools of Oriental Research, Archaeological Reports 18, pp. 515–546. Jones, J.D. (forthcoming). Glass. In: Parker, S.T., ed. The Roman Aqaba Project Final Report. Vol. 2. Keller, D.( 2006). Die Gläser aus Petra. In: D. Keller and M. Grawehr, eds, Petra-ez Zantur III. Terra Archaeological V. Mainz: Philipp von Zabern, pp. 1.1–25. Meyer, C. (1987). Glass from the North Theater Byzantine Church, and Soundings at Jerash, Jordan, 1982–1983. BASOR Supplement 25, pp. 175–222. Monteserrat, D. (1995). Early Byzantine Church Lighting: A New Text. Orientalia: Nova Series 64(4), 430–444. Moorey, P.R.S. (1994). Ancient Mesopotamian Materials and Industries: The Archaeological Evidence. Oxford: Oxford University Press. O’Hea, M. (2001). Glass from 1992–1993 Excavations. In: Z. Fiema et al., The Petra Church. Amman: American Center for Oriental Research Publications, pp. 370–376. O’Hea, M. (2007). Glass in Late Antiquity in the Near East. In: I. Lavan, E. Zanini, and A. Sarantis, eds, 2007. Technology in Transition AD 300–650. Leiden: Brill, pp. 233–248. O’Hea, M. (2012). The Glass. In: K. Politis, ed. Sanctuary of Lot at Deir ʿAin ʿAbata in Jordan Excavations 1988–2003. Amman: Jordan Distribution Agency in association with the British Museum, pp. 293–316

COLORLESS GLASS AND BYZANTINE INTERIOR LIGHTING 183 Oleson, J.P. and R. Schick, eds. (2013). Humayma, Excavation Project 2: Nabataean Campground and Necropolis, Byzantine Churches, and Early Islamic Domestic Structures. Boston: American Schools of Oriental Research, Archaeological Reports 18. Parker, S.T. and A. M. Smith II, eds. (2014). The Roman Aqaba Project Final Report, Vol. 1 – The Regional Environment and the Regional Survey. American Schools of Oriental Research, Archaeological Reports No. 19. Boston: ASOR. Parker, S.T., ed. (2006). The Roman Frontier in Central Jordan. Final Report on the Limes Arabicus Project, 1980–1989. Washington, DC: Dumbarton Oaks Research Library and Collection. Pliny the Elder. (1962). Natural History, Books 36–37, trans. D.E. Eicholz. Cambridge, MA: Harvard University Press. Prudentius. (1993). Cathemerinon V, trans. H. J. Thomson Cambridge MA: Harvard University Press. Reade, W. (2012). Glass. In: D.T. Potts, ed., A Companion to the Archaeology of the Ancient Near East. Chichester: Wiley-Blackwell, pp. 317–335. Renfrew, C., Dixon J., and J. Cann. (1966). Obsidian and Early Cultural Contact in the Near East. Proceedings of the Prehistoric Society 2, 30–72. Slane, K. W. and J. Magness. (2005). Jalame Restudied and Reinterpreted. Rei Cretariae Romanae Fautores Acta, 39, 257–261. Smith, R.H. and L.P. Day. (1989). Pella of the Decapolis. Final Report on the College of Wooster Excavations in Area IX, The Civic Complex, 1979–1985. Wooster, Ohio: The College of Wooster. Stern E.M.S. and B. Schlick-Nolte. (1994) Early Glass of the Ancient World 1600 B.C.–A.D. 500. Ernesto Wolf Collection. OstfildernRuit: Verlag Gerd Hatje. Stern, E.M.S. (2001). Roman, Byzantine, and Early Medieval Glass 10 BC–700 AD. Ernesto Wolf Collection. Ostfildern-Ruit: Hatje Cantz Publishers. Weinberg, G.D., and S.M. Goldstein. (1988). The Glass Vessels. In: G.D. Weinberg, ed., Excavations at Jalame, Site of a Glass Factory in Late Roman Palestine. Columbia, Missouri: University of Missouri Press, pp. 38–102.

184

JANET DUNCAN JONES

Young, R.S. (1981). Three Great Early Tumuli. The Gordion Excavations Final Reports, Vol. I. Philadelphia: The University Museum.

A NABATAEAN INSCRIPTION FROM BIR MADKHUR WITH AN EXCURSUS ON BASILEOPHORIC NAMES AND THE NABATAEAN DYNASTIC CULT DAVID F. GRAF & ANDREW SMITH II 1

This fragmentary Nabataean inscription was discovered in the excavations of Bir Madkhur directed by Andrew Smith in 2014. After a brief description of the project and context of the discovery of the find (Smith), the analysis of the text and its significance in an excursus will be discussed (Graf). The Bir Madhkur Project is a collaborative, multi-disciplinary field project in southern Jordan. Its primary goal is to examine from a rural, landscape perspective the history of Petra between classical and late antiquity. The intent is to shed light on the symbiotic relationship between the city of Petra and its territory. Bir Madkhur is located ca. 15 kilometers northwest of Petra. The site was occupied primarily between the Nabataean and Byzantine periods, when it served as a major way-station and military outpost in I would like to express my appreciation to Andrew Smith II for his permission to publish here this Nabataean text as a tribute to our friend and colleague S. Thomas Parker. Tom and I began our careers decades ago engaged with the Limes Arabicus, and we are both concluding them with a focus on Petra. Throughout this period, we have had many stimulating and valuable exchanges. In regard to the interpretation of this text, I wish to acknowledge important contributions by Glen J. Corbett, John F. Healey, and Laila Nehmé, and especially to Hani Hajayneh. Thomas M. Weber also contributed valuable contributions to the excursus. 1

185

186

DAVID F. GRAF & ANDREW SMITH II

the foothills of Wadi Araba, the southern section of the Great Rift Valley (Figure 1). The main features of the site include a Roman/Byzantine fort, a domestic settlement, and a caravanserai/bath complex (Figure 2). While the project has focused on the excavation of Bir Madhkur, additional excavation activity has been carried out at a regional farmhouse to the west of the site and a caravan station farther to the south. Also, the intensive, regional survey carried out north and west of Bir Madhkur, as well as south along the both banks of Wadi Musa, has generated an impressive database of greater than 1200 sites. These range from farmhouses and campsites to larger caravan stations that once supported the ancient trade passing through the region. All of these sites in the central Wadi Araba attest to a mixed, rural economy that supported the urban development of Petra. The project at Bir Madhkur began with excavations in 2008 (Smith 2010b). This was a pilot season and one area targeted was the fort itself (Area A), which remains in an exceedingly ruined state, no doubt due to multiple earthquakes in antiquity. Measuring just over 30 x 30 m, the fort is a Late Roman/Early Byzantine quadriburgium, with four corner towers and an open courtyard surrounded by rooms adjacent to the curtain wall. Bir Madhkur, in fact, was one of a chain of forts of this type constructed by Diocletian in the Araba. Others include ʿEn Hazeva and Yotvata in the western Araba, Gharandal to the south, and Qasr et-Tlah to the north (Smith 2010a). Our objectives in 2008 were to understand the occupational history of the fort as well as to clarify its construction and to determine its plan. In particular, we sought the location of the gateway. We also wanted to investigate whether the fort, as it exists today, was constructed atop an earlier, Nabataean structure. While there is a strong Nabataean and Early Roman presence at Bir Madhkur in the form of surface pottery and coins, no extant architectural features from this earlier occupation has been identified anywhere at the site. With all of that in mind, the excavation strategy in Area A was to open two adjacent trenches (A1 and A2) along the north wall of the fort in order to reveal the presumed gateway as well as to reveal the foundation level of the curtain wall. These efforts exposed a small portal gate 0.79 m wide, which opens directly onto the (presumably ancient) well immediately outside of the fort as well as the spring farther to the northeast (Figure 3). We

A NABATAEAN INSCRIPTION FROM BIR MADKHUR

187

returned to Bir Madhkur in 2009 to conserve the exposed segment along the north wall and the portal gate.

Figure 1. Location of Bir Madkhur

Figure 2. Aerial view of Bir Madkhur with major features marked (Satellite image courtesy of GeoEye)

188

DAVID F. GRAF & ANDREW SMITH II

In 2014, we returned for a three week season at Bir Madhkur and reopened the probes along the north wall of the fort (both Trenches A1 and A2). The main goal was to reassess the stratigraphy and reevaluate the field data from 2008. We also wanted to expand our conservation efforts along the wall, as some sections had been damaged since 2008. The data recovered from these probes along the north wall of the fort are consistent with what we have observed from work elsewhere at the site. In addition to a plethora of pottery, glass, bone, and other organic and inorganic remains, the most significant of the results from our work at Bir Madhkur (mainly over the course of two and half field seasons) has been the discovery of more than 800 coins and other small finds, the final analysis of which continues. Nevertheless, preliminary analysis of the coins so far suggests a date range from the first century BCE to the fifth century CE. The majority, however, are overwhelmingly fourth century types. Again, identifying the nature of the Nabataean settlement at Bir Madhkur, largely revealed through pottery and coins, remains complicated by the Late Roman/Early Byzantine military occupation at the site.

Figure 3. Image shows the gateway along the north curtain wall of the fort where the inscription was uncovered (Photo by Andrew M. Smith II)

A NABATAEAN INSCRIPTION FROM BIR MADKHUR

189

Figure 4. The Nabataean inscription found in Trench A2 (Photo by Andrew M. Smith II)

THE NABATEAN INSCRIPTION

While clearing tumble from Trench A2 north of the wall of the fort, several carefully worked blocks were removed that were originally part of the portal gate. On one of these blocks we discovered a two-line fragmented Nabataean inscription (Figure 4). Nabataean inscriptions are rather rare in the Wadi Arabah (Graf 2014), so this text offers an important contribution to the Nabataen corpus. The incised block is rectangular and slightly wedge-shaped, measuring 18 cm deep and 27 cm high, but with a length of 31 cm at its base and 33 cm at its top. The sides of the block demonstrate a slight anathyrosis. Also, the surface of the block is not smooth, as the inscribed face has shallow-angled chisel marks across half of its face. In combination with other wedge-shaped blocks from the excavation, it is likely that the portal gate was vaulted. These blocks would have served, then, as voussoirs for an arch, and the inscribed block perhaps served as the keystone. The stone has suffered damage both to the beginning and ending of the inscription, broken cleanly on the right and shattered more on the left end of the block, but may have been part of lintel for a construction of some previous structure, reused in the later fort. The inscription contains the end of the personal name of the

190

DAVID F. GRAF & ANDREW SMITH II

inscriber and part of his patronym, followed by only two letters of a noun.

ANALYSIS OF THE TEXT: […. M]LKYM br ῾BD-῾B[DT] […….]T᾽

Date: The only basis for establishing a date for the brief fragmented text is by paleographical analysis of the script, which indicates a second-third century CE date. The tāw is considered especially distinctive for this period (Healey 1990–1991: 47, s.v. “t” dating from 166/169 CE and later), but it already appears in an inscription dated to the time of Malichos II (40–70 CE) at Petra (Zayadine et al. 2003: 224, no. 116), but becomes typical later. The form is known in the Sinai inscriptions at Jebel Moneija from the second and third centuries CE (Gruendler 1993: 41: N13), the Raqāši inscription of 267 CE (JS 17 = CIS 271 = Gruendler 1993: N16; cf. Cantineau 1932: 36), and the latest dated inscription at Madāʾin Ṣāliḥ of 356 CE (Gruendler 1993: N20). See also for a parallel Al-Theeb 1993: 137, no. 64, who notes the form of the tāw in the name THYNW in a text from Northwest Arabia “is unprecedented among [t]hese inscriptions.” The final mīm is also typical of Northern Transjordan and the Ḥaurān in the second century CE (Healey 1990–1991: 47, “M”), the Sinai inscriptions in the second and third centuries CE (Yardeni 2000: 85–88, listing CIS 794, 825, 867, 914, and 1109), and is found again in the latest Nabataean inscription at Madāʾin Ṣāliḥ in 356 CE (Gruendler 1993: 96, N20). The ʿain with its curvature of the tail upward and then to the right appears at Wadi Ramm and northwest Arabia in the second century CE (Healey 1990– 1991: 47, “M” 166/169) and thereafter throughout Nabataean Arabia (Gruendler 1993: 76, N11, N17, N19–20). For the kaf see Gruendler 1993: 89, N13 [Sinai] and al-Theeb 1993: 187–189 [northwest Arabia]. These impressions suggest the text probably dates roughly to the second or third century CE, or a little later. Line 1. The partial traces of a letter preceding -LKYM are visible on the edge of the stone, so it may be suggested that the missing initial letter is a mīm, and that the name should be restored as [M]LKYM (Malkim). If could be assumed the initial name is complete and read as LKYM, perhaps derived from Arabic lakam but this provides an unattested personal name in either Nabataean or

A NABATAEAN INSCRIPTION FROM BIR MADKHUR

191

Pre-Islamic Arabian texts. 2 If could less likely be read as BKYM (cf. Arabic BKY “to cry” or ”mourn”), but it is also a name unattested in Nabataean or North Arabian texts. 3 However, the reading MLKYM is equally problematic. Although unattested in Nabataean, the name MLKYN and MLKYWN (C 431 at Petra) appears in Nabataean (Cantineau 1932: 114; al-Khraysheh 1986: 108) and MLKY is known in North Arabian Thamudic E (King 1990: 629) and Palmyrene (Stark 1971: 95). The problem is how to explain the final -M in MLKYM. The first possibility is to regard the suffix -M in MLKYM as a hypocoristicon in which the second part of the name has been suppressed or abbreviated, but it is far from clear what superlative or theophoric element is missing. The second possibility is to explain the final –M as “mimation”. In South Arabian, “mimation” is typical for personal names in Sabaic (Beeston 1984: 30, par. 14.2) and appears erratically in Beeston (1984: 61, par. 14.1). But it is extremely rare in early North Arabian Safaitic and Thamudic (see Moscati 1959: 37), appearing only occasionally: QRSM/QRŠM and QNM are proposed by Littmann (1945–49: 180), who suggests they may reflect the influence of the Minaean colony at al-ʿUlā. In addition, Hayajneh adds GLHM and ŠʿTM in Safaitic (1998: 265–66, who provides an excellent discussion of the -M suffix). There is also the rare occurrence of a final -M in North Arabian Thamudic E vocative prayers in connection with the deities LH (Allah) and LT (Allat), which occurs in the phrase “O LH-m [Grant]…” or “O LT-m [grant]…”, but in association only with these two deities (King 1990: 80, with examples in Thamudic E at 102–3). The parallel cited for this phenomenon is the -umma ending in Old Arabic allāhumma, in expressing the name of God in the vocative, where the -M is considered “a substitute for the suppressed vocative particle” (Lane 1980: 83c). But the suffix or ending of -M is far as I Lane 1980: 3013a, lakam, ‘to strike with the fist, punch, box’; cf. Kazmirski 1960: II, 1021. See Lane 1980: 2671a, lakyza “the two contended in striking each other upon the breast or beneath the chin, with the fists’. 3 But see Harding 1971: 115 for Safaitic BKY or bākiya ‘wailing woman, hired mourner’; and cf. CIK 219,1 Bādiya. 2

192

DAVID F. GRAF & ANDREW SMITH II

know completely absent in Nabataean Aramaic, although appearing in Arabic on occasion, e.g. Ğulhuma (CIK 265,1), and add Ğuʿšum, Ḥarṯama, Saʿdam, Šaʿṯamand Qarham, in “Bedouin” contexts (Walther 1966: 29). In any case, the occurrence of the final -M in this Nabataean text at Bir Madkhur remains strikingly unique. A third possibility is that MLKYM represents the ancient Ammonite national deity of Moloch or Milcōm (mlkm), sometimes used as element in theophoric personal names, such as in BDMLKM, “Badmilkom” (Aufrecht 1989: no. 1, “in the hands of Milkōm”), perhaps to be restored as [ʿ]BD-MLKM, “Servant of Milkōm” (Bordreuil and Lemaire 1972: 57). The deity MLKM also is well known also from personal names in the Bronze Age at Ebla and Mari (Heider 1992: 895–96; cf. Huffmon 1965: 230–31 and cf. Hess 1993: 112–114, no, 116; cf. Healey 1975: 235–238). If this is the case, our MLKYM may represent a rare survival of the Iron Age deity of the Ammonite kingdom. Some support for this interpretation if offered by a North Arabian text from the region of Bayir in southeast Jordan that refers to the three major Transjordanian gods of the Iron Age: “O Milkom and Kemosh and Qōs” (h mlkm wkms wqws), where the gods are named in geographical order, from north to south, Ammon, Moab and Edom (Hayajneh, Ababneh, al-Khraysheh 2015: 82). It is possible that Mlkm in this vocative text has an enclitic -m ending added to the noun “King”, but it is at best, centuries removed from our Bir Madkhur text. Nevertheless, these linguistic phenomena offer at least a possible explanation and basis for the appearance of a final -M in our text and for proposing the first name be restored as MIKYM. As for the patronym, the surviving portion of the name is ʿBD-ʿB…, and can be almost certainly be restored as ʿBD-ʿB[DT]. This basileophoric name appears throughout the Nabataean realm. 4 Line 2. The –Tʾ in the second line suggests the emphatic state feminine singular ending of a noun. The indication seems to be a dedication of some construction. Here the possibilities are numerous, including byrtʾ (“the temple”, in Nab: see CIS 163, 164; RES 2023/3; but also “fortress” elsewhere in NW Semitic: Hoftizer and Cantineau 1932: 126; Negev 1991: no. 815, lists 16 occurrences; cf. al-Khraysheh 1986: 131. 4

A NABATAEAN INSCRIPTION FROM BIR MADKHUR

193

Jongeling 1995: 155 ), mqbrtʾ (“the tomb”, CIS 196), and ʿbydtʾ (“the work” CIS 196), but excluding bnynʾ (“the building or edifice” CIS 196), and bbʾ, “the gate”: Hoftiuzer and Jongeling 1995: 142–43).

EXCURSUS: THE NABATAEAN BASILEOPHORIC PERSONAL NAMES

The name ʿBD-ʿBDT deserves additional comment and analysis. Such basileophoric names are common in Nabataea: ʿbḥrtt appears at least 10 times (Negev 1991: no. 802; add Graf 1994: 294 ), ʿbdmlkw 8 times (no. 808; add Graf 2014: 352 ), and ʿbd-rbʾl six times (Negev 1991: no. 824). Starcky suggested these “theophoric anthroponyms” are an indication of the presence of a ruler cult among the Nabataeans, and that all the Nabataean kings were divinized, including the queens, as signified by the servant names ʿbdšqlt and ʿbd-ḥldw (Starcky 1966: 1015), perhaps adopted mainly by administrative and military officials and cultic personnel (Graf 1994: 292–96; cf. Strootman 2014: 111–135). In this regard, he was following the pioneering discussion by Clermont-Ganneau who observed, if the Nabataen kings were divinized, it was only natural that their names became the theophoric component in the onomasticon. 5 Clermont-Ganneau further observed that the “pseudotheophoric” names in Nabataean were a highly unusual feature. Although the phenomenon of “servant names” is common in Phoenician and Punic inscriptions for the gods of their pantheon (see Benz 1972: 148–164. e.g., the theophoric names ʿbd-’šmn 150– 5, and ʿbd-mlqrt 155–161), they are more extensive in Nabataean than in any of the other Aramaic dialects or in other Semitic languages (Silverman 1981: 363). As an explanation for the curious phenomenon of the “pseudo-theophoric names, Silverman assumed they were a product of foreign non-Aramaic influence, namely Arabic. The adoption of the Ruler Cult of the Hellenistic period by the Nabataean kings provides a more compelling explanation. The Clermont-Ganneau 1888: 39–47, 142; for a listing of the “servant names” see Milik 1959b: 148–149. 5

194

DAVID F. GRAF & ANDREW SMITH II

only explicit literary evidence for the divinization of the Nabataen kings is a rather late tradition of the Byzantine era. According to the Arabicus of the historian Uranius, probably to be dated to the fourth century CE (West 1974: 274), the Nabataean city of ʿAvdat was “where King Obodas, whom they [i.e. the Nabataeans] deify, is buried.” This tradition is preserved by the Byzantine grammarian Stephanus of Byzantium, in probably the sixth century CE (Ethnika = FgrH 675, F 4 = ed. Meineke 482, 16–17). The tradition receives verification in a several Nabataean texts stemming from ʿAvdat that mention “Obodas the God” (Negev 1986b: 56–60, for ʿbdt ʾlhʾ) and another that asserts “as ʿObodat lives” (RES 527, ḥy ʿbdt). After the Roman annexation of Arabia, the tradition continues in Greek texts from ʿAvdat that mention Thea Oboda (Negev 1981: no. 6) and Zeus Oboda (Negev 1981: nos. 3–7; cf. 1b and 11). 6 The “cult of Obodas” was not restricted to ʿAvdat in the Negev, but was also prominent at the Nabataean capital at Petra. Less than a km south of the Theater, there was an important religious rock-cut triclinium complex at Jebel Numayr, where a century ago a statue was discovered that was dedicated to “ʿObodas the god”, offered in behalf of King Aretas IV, the queen Shaqilat, and other members of the royal family (CIS II, 354 = Milik 1959a: 559– 60; cf. Dijkstra 1995: 57–60). Recent excavations at the cultic complex discovered the fragments of a statue, ostensibly the one offered to “Obodas the God” (Nehmé 2002a: 247–50 and fig, 10–11 and 2002b: 452–55; Schwentzel 2006: 129; Kropp 2003b: 50; cf. Tholbecq et al. 2008 and Wenning: 2015: 44–46). In addition, over a hundred graffiti were found in the environs, some 40 of which seem to be associated with the open-air triclinium or chapel (Nehmé 2002a: 245; cf. 2002b). More importantly, subsequent excavations in the religious complex have revealed beneath the religious complex associated with Aretas IV an even earlier open-air triclinium dating to the mid-second century BC, based on early Nabataean pottery (Schmid Type I), Hellenistic imports and C14 tests (Tholbecq and Durand 2013; cf. Nehmé 2012). The relationship of the various stages of the cultic center still needs to be estab6

For discussion, see Erickson-Gini 2014.

A NABATAEAN INSCRIPTION FROM BIR MADKHUR

195

lished but the unpublished graffiti may provide the clues to any possible continuity in the worship of “Obodas the God” at the site. The worship of “Obodas the God” at Petra also extended beyond Jebel Numayr. At ed-Deir, on the heights above Petra, a Nabataean text indicates the presence of the members of a “symposium” (marzēḥ) who worshipped “Obodas the God” (Dalman 1912: 92–94, no. 73 = RES 1423). More recently, evidence has emerged of the worship of Obodas in the settlement of al-Gaia (modern el-Jī) outside of inner Petra. A recent inscription on a bronze oil burner found in the Falāhat region in SW Wādī Mūsā preserves a dedication to “Obodas the God” (ʿbdt ʾlhʾ) by a priest named Zwyls (perhaps Greek Zōilos, a personal name which appears several dozen times in the Greek world: cf. Wuthnow 1930: 51) and, more importantly for us, his son ʿAbd-ʿObodat for a temple in Gaia, dating to the reign of Rabbel II (Al-Salameen and Falahat 2014: 293–307). Another recent find from the same Wādī Mūsā, an inscribed bronze lamp of the first century CE, also is dedicated to “Obodas the God” (ʿbdt ʾlhʾ) by a woman ʾAmat-Allah (a priestess?) and Taym-Dushara (Al-Salameen and Shudaifat 2014: 43–45). The noteworthy personal Basileophoric and Theophoric names of these dedicants suggest the worship of “Obodas the God” was an integral part of Nabataean religion, not a small peripheral sect or a family cult. The problem is to identify which one of the Nabataean kings named ‘Obodas’ is designated by the expression “ʿObodas the God.” The most likely candidate is Obodas I (ca. 96–84 BCE), as proposed by Starcky (1966: col. 906) and generally followed by later scholars (Bowersock 1983: 62). The possible short reign of a putative Obodas II (62–61 BCE), postulated only from questionable numismatic evidence (now rejected by Huth 2010: 214–217), removes him from consideration (Graf 1992: 971; contra Meshorer 1975: 16), and the description of Obodas III (30–9 BCE, probably better designated as II) as uninterested in administrative and military affairs (Strabo 16.4.24 [781]), and “inactive and sluggish by nature” (Josephus, Antiquities of the Jews 16.200), seems to exclude him as the possible defied king. Although Negev once earlier argued he was the deified Obodas (1986a: 107), he later became more ambiguous (1986b), before finally proposing it was Obodas I (Negev 2003: 18), and this remains the preferred identification for “Obodas the god.”

196

DAVID F. GRAF & ANDREW SMITH II

But why was Obodas I divinized? The traditional explanation for Obodat I’s divinization was to connect him with the defeat of the army of Antiochus XII Dionysos Epiphanes, which resulted in the death of the Seleucid ruler. According to Milik, this event “captured the imagination of the Nabataeans” and led to the divinization of Obodas I (as cited in Starcky 1966: 906; cf. Bowersock 1983: 24–25). Several obstacles now seem to prohibit this explanation. Since Antiochus XII’s coinage begins in 87/86 BCE at Damascus and ends in 83/82 BCE (Houghton et al. 2008: 2471 and 2472A), the climactic battle and his death must be assigned to 83/82 BCE (as suggested by Bellinger 1949: 77, but virtually ignored). But the identification of the Nabataean king Obodas I with this victory is never explicitly stated. The account of the conflict only mentions the Nabataean victor is an unnamed “Arab king” (Jos. BJ 1.99–102; AJ 13, 387–91). Moreover, it now seems clear that Obodas’ reign terminated before this event, and that another Nabataean king ruled at the time of Antiochus XII’s defeat. On a statue from Qasr al-Bint at Petra, an inscription indicates it was dedicated to “Rabbel, King of the Nabataeans, [……]t, the king of the Nabataeans”, in the 18th year of Aretas the king (CIS 349). The missing phrase can be restored as “son of Aretas” ([br ḥrt]t) or “son of Obodat ([br ʿbd]t). Starcky dated it to Aretas III (Starcky 1966: 905), suggesting a dynastic sequence of Obodas [I]-Rabbel [I]Aretas [III], which means there is a king named ‘Rabbel’ who must be squeezed in between Obodas I and Aretas III. This anonymous king is better identified with Rabbel I. Uranius’ Arabika, lists under the Arab village Motho, the place “where Antigonus the Macedonian was killed by Rabbel, king of the Arabs” (FGrH 675 F 25 = Stephan of Byzantium). But it is known that the Macedonian Antigonus, one of Alexander the Great’s successors, died in the Battle of the Diadochoi at Ipsus (Phrygia) in 301 BCE (Plutarch, Demosthenes 29). This suggests that Uranius’ text is corrupt, and the proposal to emend the name ‘Antigonus’ in the text to ‘Antiochus’ and refer the incident to the Nabataean defeat of Antiochus XII in 83/82 BCE, in which the Seleucid ruler was killed. The unnamed “Arab king” (Jewish War 1.99–102; Antiquities of the Jews 13, 387–91) is best identified with Rabbel I. Since the decisive conflict is located in the environs of Kana, identified as Khirbet Gazza at the entrance of Waldi al-Qina/Qena, just east of the southern end of the Dead Sea (Schmitt 1995: 202), in proximity

A NABATAEAN INSCRIPTION FROM BIR MADKHUR

197

to modern Mauta in Moab, this provides support to the tradition. This location also agrees with Josephus’ indication that the retreating Seleucid army “succumbed to starvation” at Kana (Jewish War 1.101), where they “perished of hunger” (Antiquities of the Jews 13.391). It is possible that Rabbel I died himself from wounds suffered in the battle, for immediately afterward, the Nabataean king Aretas III is declared by the Damascenes the “King of Coele Syria” (Jewish War 1.103) and began issuing coins at Damascus (Newell 1939: 92 no. 144; Meshorer 1975: 14–16, 86–87, nos. 5–8). If Obodas I is to be identified with ‘Obodas the god’, the only event that we can associate with such a development is the single fact that we know about his reign. According to Josephus, “Obodas, king of the Arabs”, defeated the Hasmonean king Alexander Jannaeus in a battle at Garada near the Golan in c. 95–93 BCE, just after the Judean king had conquered Moab and Galaaditis (Jewish War 1.90; Antiquities of the Jews 13.375). This significant victory provides for us the only basis for his deification and the cult of “Obodas the God” (Nehmé 2012), but the circumstances remain obscure (cf. Dijkstra 1995: 319–21). The problem remains why was he was buried at ʿAvdat. Avraham Negev was forced to hypothesize he must have been killed in an unrecorded battle in trying to reconquer the territory lost in the Negev after Alexander Janneaus’ conquest of Gaza (Negev 2003: 18*). Admittedly, this is a scenario built on a series of assumptions, attributing his supposed recovery of the vital trade route between Petra and Gaza to an imaginary battle. Such conjecture and speculation exist in the absence of any concrete evidence. As a consequence, a more skeptical and pessimistic view of any apotheosis of a Nabataean king has been expressed (Wenning 2015: 55). Rather than a divinized king, it is proposed that “Obodas the God” represemts a cult for a Nabataean deity of that name, centered at the Negev city, and the name was subsequently adopted by a series of Nabataean kings, none of whom were “deified” (Dijkstra 1995: 319–321; cf. Alpass 2013: 158). As a result, these basileophoric Nabataean “servant” names are removed as evidence for the divinization of the Nabataean kings (Dijkstra 1995: 321 n. 7). Even Healey is dubious that the “servant names” may suggest the divinization of the Nabataean kings, and emphatically rejects that any of the Nabataean queens were ever divinized (Healey 2001: 150). As for the tradition preserved by the Arab historian

198

DAVID F. GRAF & ANDREW SMITH II

Uranius, Wenning considers it an aetiological legend created to explain why the cult “Obodas the god” was associated with ‘Avdat, a “political fiction” invented in late antiquity (Wenning 1997: 191; cf. Alpass 2013: 158). In the process, dubious myths and legends replace historical conjecture. Of note is none of these critics refers to Clermont-Ganneau’s argument a century ago that the basileiophoric names imply the divinization of the Nabataean kings and queens, based on far less evidence than is now available. In my opinion, the divinization of the Nabataean kings and queens still offers the best explanation for the “servant names.” The dynastic ideologies of the Hellenistic kings employed the ruler cult to legitimize their rule, and it was influential for kings throughout the Near East (Chaniotis 2003; cf. Habicht 1970). In the Ptolemaic world, a centrally organized dynastic cult, with the divinization of the ruling kings and queens, and a hierarchy of provincial priests, is well documented (Fraser 1972: I, 213–246; Winter 1978; Koenen 1993; Pfeiffer 2008; Thompson 2012: 117–128). The institution of the dynastic cult began in 272/1 BCE under Ptolemy II Philadelphus and Arsinoe I with the Theoi Adelphoi, i.e. the ruling divine brother-king and sister-queen (Quaegebeur 1998a; Hölbl 2001: 92–98; cf. Hazzard 2000). In 243/2 (Haubner 2011), Ptolemy III Euergetes and Berenike were declared Theoi Eueregetai in 243/2 BCE, followed by Ptolemy IV Philopater and Arsinoe III as Theoi Sotares in 215/4 BCE (Fraser 1972: I, 216–219), and later by Theoi Philopatores and Theoi Epiphaneis (Thompson 2012: 124). By the end of the second century BC, there were even more cult titles, and more priests and priestesses supplementing the dynastic cult (Fraser 1972: I, 16–220). Of significance is Cleopatra III (142–101 BCE), with her son Ptolemy IX Soter II (116–107 BCE), the Theoi Philometores kai Soteres, explicitly promoting their divinity as political propaganda (Quaegebeur 1978), with the garrisons and soldiers the active adherents and promoters of the dynastic cult (Chaniotis 2003: 441; cf. Quaegebeur 1989). The identification with deities culminates with Ptolemy XII Auletes and his daughter Cleopatra III identified as the Neos Dionysos and Nea Isis, respectively (OGIS 741 with Fraser 1972: 244 and n. 238, citing Plut. Ant. 54). The royal ideology now represents the personification of the King as Osiris-Dionysos (Thompson 2012: 108) and the Queen as IsisTyche (Quaegebeur 1988; Hölbl 2001: 285–293).

A NABATAEAN INSCRIPTION FROM BIR MADKHUR

199

The early connections of the Nabataean dynasty with the Ptolemaic kingdom in the third century BCE indicates the Nabataean kings were exposed to the Egyptian ruler cult at an early stage. The first attested Nabataean king is contemporaneous with Ptolemy II Philadephus, between 272 and 252 BCE, and appears closely allied with the Ptolemaic dynasty in Egypt at the time (Graf 2006). The association of the Nabataean royal god Dushara with Dionysus, particularly associated with the Ptolemaic dynasts, is recorded by Isidore of Charax, the author of the Parthian Stations, who accompanied Juba of Mauretania on Gaius Caesar’s Eastern Expedition in the time of Augustus (FGrH 781 F5; Roller 2004: 217–219). At Baidha, 7 km north of Petra, excavations of a cultic rock-cut colonnaded cultic center exposed a triclinium, which appears to be associated with a Dionysian thiasos in the reign of Malichus I, based on the associated divine and zoological sculptures, and the location in a wine-growing region (Bikai et al. 2008). A nearby inscription of a marzēah (“symposium”) at Baidha (Zayadine 1986) is now matched by similar attestations at Wadi Mūsā, the ancient al-Gaia (Al-Salameen and Falahat 2012: 37–51). Since the “god of al-Gaia” is identified in several Nabataean texts as “Dushara” (Wenning 2015: 55), it may be assumed the Nabataean supreme god was identified with Dionysos, although such marzēah are associated at Petra also with the cult of “Obodas the god” (Nehmé 2012). Where this Ptolemaic influence becomes explicit is in the iconographic representation of the Nabataean kings and queens on Nabataean coinage (Kropp 2013a: 241–43). The kings are depicted as beardless youths with long flowing hair, precisely the hair-style of Dushares, the principle god of their dynasty (Bowersock 1990:31). This is exactly the late Hellenistic iconographic representation of Dionysos. 7 What makes this clear is the Dionysiac diadem, a simple ‘white’ headband with free-hanging ends that appears on the heads of Nabataean dynasts from Aretas III to Rabbel II (Schwentzel 2005: 152–153), It was the symbol for Dionysos’ conquests in the East and became the royal symbol for the divinized kings of Asia (Smith 1988: 34–38, 44; cg. 1991: 65). The Healey 2001:99 and Zayadine 2003: 59; although considered inconclusive by Kropp 2011: 185–89 7

200

DAVID F. GRAF & ANDREW SMITH II

Nabataean dynasts are thus represented as adopting/adapting the divine imagery of the patron god of their dynasty for their own personal appearance and portraiture. 8 Dushara, as the dynastic god of the royal family is indicated in the expression “Dushara the god of our lord (the king)” (CIS 350.3–4, and RES 83, ʾlh mrʾnʾ, ‘god of our lord’; Healey 2001: 97–98). In similar fashion, the queens are represented as the Tyche (fortune deity) of the Nabataean realm, with a raised open palm, and a headdress decorated in the front with an Isis ornament. The same Isis iconographic elements are reflected in the representation of the goddess al-ʿUzza (Zayadine 1981: 117 and 1991: 283–306; cf. Schwentzel 2005: 162; and Kropp 2013a: 242–43). The Isis headdress appears first with Queen Huldu in the reign of Aretas IV in 15 CE and later with Queen Shuqilat in 27 CE (Hoover and Barclay 2010: 204), and continues with the later queens of Rabbel II, Gamilat and Hagiru (Schwentzel 2014: 156– 58). The correlations suggest the numismatic images of Nabataean kings and queens have been infused with the divine symbols of Dionysos and Isis (pace Kaizer 2010: 119–120). From almost the beginning of Ptolemaic rule, Dionysos was identified with the Egyptian rulers, as reflected in the Dionsiac festival as described in the grand procession of Ptolemy II Philadelphus, where Dionysos was recognized as the founder of the dynasty (Kahill 1996: 79–80). By the reign of Ptolemy IV Philopator (221–204 BCE), at least, the rulers were presented as Dionysos incarnate and this became characteristic of Ptolemaic royal ideology. It appears even that a Dionysiac thiasos existed at the Ptolemaic court where the ruler was presented as the reincarnation of Dionysos (Tondriau 1946: 169; cf. 1947, 1950; Van Nuffelen 1999: 183, 188; Pfeiffer 2008: 41). In sum, the later Ptolemaic kings associated themselves with Dionysos and appear convinced of their ‘divine’ character (Van Nuffelen 1999: 176). In similar fashion, the Ptolemaic queens identified themselves with Isis from Arsinoe II Philadelphos in the third century to Cleopatra VII. Arsinoe II was identified with Isis, Aphrodite and Schwentzel 2005: 154; 2008: 290; 2010: 237–38; 2014: 152–153; cf. Kropp 2011, 188–189, and 2013b: 49, who admits there are Hellenistic Alexandrian elements in the royal Nabataean portraiture 8

A NABATAEAN INSCRIPTION FROM BIR MADKHUR

201

Hathor (Müller 2009; 280–299; Caneva 2012: 12; cf. Quaegebeur 1978), but with Cleopatra III and afterwards, Isis is predominant (Van Nuffelen 1999: 179). The latter declared herself the “Sacred foal of Isis, the Great Mother of the gods,” a clear attestation the Ptolemaic queen identified herself with Isis (Fraser 1972: 221, and n. 249; cf. Colin 1994: 272–283, and Plantzos 2011: 395). By the first century BCE, the ruling king and queen were now thoroughly divinized, the Königspaar (Colin 1994: 293), incorporated into the Egyptian cult “during their lifetime (Quaegebuhr 1989: 107). This culminates with Antony and Cleopatra, identifying themselves as the Neos Dionysos and the Nea Isis (Fraser 1972: I, 245–246; Quaegebeur 1998: 53; cf. Roller 2010: 114–117). This does not mean, of course, that the Ptolemaic dynasts were incorporated into the Egyptian pantheon. They were not the object of sacrifices, and they are depicted presenting offerings themselves to the Egyptian gods (Thompson 2012: 126; cf. Koenen 1993). But they had a distinctively different status than mortals, conferred by their subjects based on their status as rulers, and the merit of their benefactions and protection of the state. The bifurcation between mortals and immortals being breached by rulers achieving this unusual ideal status was primarily communicated not only nomenclature, but also iconography (Smith 1988: 39). By contrast, in Rome, the emperor was not considered deus (‘god’) in his lifetime, but became divus after his death, when ‘heavenly honors’ were conferred to him by the Senate (Price 1984: 83; Gradel 2002). However, in the Greek East, the emperor was commonly referred to as theos, assimilated to a Zeus, Helios or Dionysos, and the recipient of isotheoi timai, “honors equivalent to the gods, on par with divinity, worthy of a god” (Montanari 2015: 986). The ruler was still clearly less than the gods, but of higher status than mere mortals, in essence “god-like” (Price 1984: 86–94). Whether the Nabataean kings and queens, represented as being assimilated to Dionysos or Isis, regarded themselves as similarly “god-like” is difficult to determine. As A. D. Nock observed, “it is not easy to draw a definite line between comparison and identification” but that a ruler could be the reincarnation of a particular deity was not foreign to ancient ideology (Nock 1928: 32–35). In the case of the Nabataean dynasts, there is epigraphic and iconographic evidence that this was the case.

202

DAVID F. GRAF & ANDREW SMITH II

The circumstances and time when the ruler cult was first adopted in the Nabataean realm are difficult to determine, as the evidence is primarily of the first century BCE and later. But it is noteworthy by this time, the monumental architecture at Petra reflecting the “second Pompeian style” (of the first century BCE) is drawn from Alexandria (McKenzie 1990: 85–100, 124–126). In addition, such architectural features at Petra as the obelisk, crowsstep and elephant can be attributed to Egyptian influence (Finlayson 2016: 83–101). These Egyptian influences correlate with the chronology of the appeance of the royal cult in Nabataea. If we are correct that the cult of “Obodas the God” is connected with Obodas I, the beginning can be traced at least back to the early first century BCE. The first appearance of the Nabataean queens in the public sphere is marked by the anonymous queen on the coins of Obodas II (30–9 BCE), with close affinity to the Ptolemaic and Seleucid issues that depicted the portraits of the kings and queens together (Meshorer 1975: 33–34 and nos. 21–31). The adoption of the cult of Isis at Petra in the late first century BCE demonstrates the influence of Ptolemaic Egypt on Nabataean culture (Merklein and Wenning 1998; Healey 2001: 137–140; Vaelske 2013: 351– 361). Idumean ostraca in Palestine of the fourth century BCE already contain “servant names” with the Egyptian deities (Porten and Yardeni 2014: 244): for Isis see ʿBD-ʾSY (A14: 3) and ʿBD-ʾS (A215.1) and for Osiris, ʿBD-ʾWSYRʾ (A215.1), where the population has a substantial population bearing Arabic-Nabataean names. But the first sign in the Nabataean realm appears later: an Aramaic text mentioning the goddess Isis at Sīʿ in the Ḥaurān is dated to 108/7 BCE (Milik 2003: 269–74, correcting his date of 104/3 BCE). At Petra, it is marked by a dedication to Isis with a relief of the goddess dated to 25 BCE at el-Mreriyye in Wadi es-Siyyagh (Milik and Starcky 1975: 120–124 = Bricault 2005: 513, no. 404/0501; cf. Janif 2004). The concentration of Isis statuettes at the Temple of Winged Lions at Petra (Roche 1987: 218) and the possible presence of Osiris in the relief at el-Mreriyye has led to the suggestion that the mysteries of Osiris were celebrated at Petra and that Isis herself had a small temple at the Nabataean capital (Bricault 1992: 39 and 45). Some twenty or more terracotta and stone figurines primarily from domestic contexts throughout the civic center (El-Khouri 2002: 11, 52–54), demonstrating the popularity of the cult at Petra,

A NABATAEAN INSCRIPTION FROM BIR MADKHUR

203

and a Greek inscription of 257 CE mentions a priest of Isis at Petra (Milik and Starcky 1975: 123 = Bricault 2005: 514, no. 404/0502). This is not an isolated phenomenon, exclusive to Petra. Elsewhere in Nabataean Arabia the evidence is admittedly more fragmentary and late. At Gerasa, statues of Sarapis and Isis were dedicated in 142/3 CE (Bricault 2005: 512, no, 404/0401), and a bust of Sarapis from Umm al-Jimāl is known from the second or third century CE (Weber 2006: 82, no. 61). But there is earlier evidence off the Ptolemaic royal cult elsewhere in Palestine. The remains of an early Hellenistic shrine of Isis and Sarapis are known at Samaria, attested by an inscription and pottery decorated with Isis crowns and headdresses (Magness 2001: 158–165; Bricault 2005: 510, n. 403/0501; cf. Bricault 1999). There is also a dedication to Serapis and Dionysus of the Augustan era at Nysa-Scythopolis (Bricault 2005: 509, no. 403/0301). This regional impact of the cult is just a reflection of the widespread popularity of Isis across the Mediterranean in the Hellenistic and Roman periods. 9 This extensive evidence makes it difficult to dismiss the evidence of Isis at Petra as merely a product of local religious ideas (pace Alpass 2010: 107), and provides a context for the adoption of the Ruler Cult in Nabataea. In addition to the basileophoric names in Nabataean inscriptions, there are other indications for the possible existence of the royal cult in Nabataea. First, at Petra dynastic statues have been discovered in the vicinity of the civic center at Petra, particularly around Qasr al-Bint at Petra. In Ptolemaic Egypt, such cultic statues in honor of the divinized rulers (agalma) were placed inside the sanctuaries of the local gods, as “temple-sharing gods,” or what the Greeks called synnaos theos (Nock 1930; Smith 1988: 15–20). This practice began with Ptolemy II and Arsinoe II, 10 and continued with the later Ptolemaic dynasts (Pfeiffer 2008: 33–45; Thomson 2012: 118), including members of the royal famly (Smith 1988: 24; Bricault 2001; Map I, for the diffusion across the Mediterranean world; for the epigraphic evidence, see Bricault 2005, and for the numismatic evidence, Bricault 2008. 10 Quaegebuhr 1998a; Collombert 2008; cf, Carney 2000: 24–30, for earlier Macedonian precedents. 9

204

DAVID F. GRAF & ANDREW SMITH II

Strootman 2014: 98–110). The conglomeration of statues of Nabataean royalty in the precinct of Qasr al-Bint suggests the Nabataeans were replicating the Ptolemaic practice. This is first attested by a statue of Rabbel I, erected in the reign of Aretas III ca. 65 BCE found in the environs of Qasr al-Bint (CIS II 349 with Schwentzel 2006: 126–127). The second is the base for a statue of Aretas IV inserted into the southern temenos wall of Qasr al-Bint (Starcky and Strugnell 1966: 244–47 = Zayadine et al 2003: 90, no. 1, and 225 no. 119; Schwentzel 2006: 127–128), but clearly in a secondary context (Graf et al. 2007 230–236). The third is that of the princess Sha‘udat, the daughter of King Malichos II, perhaps from the pronaos of Qasr al-Bint (Zayadine et al. 2003: 91, no. 3 with 224 no. 116; Schwentzel 2006: 128). In addition, there is a marble plaque found near the Temenos Gate recording an offering to Malichos II (Zayadine et al. 2003: 90–91, no. 2), which is perhaps to be associated with the gallery of statues at Qasr al-Bint. Some fragmentary statuary at Petra has been associated with the Nabataean kings and queens, but their interpretation remains speculative, as it is difficult to determine whether they represent divinities or divinized rulers (Schwentzel 2006: 131–1340). Outside of Qasr al-Bint, the only addition at Petra is the staue of the divine Obodas erected at Jebel Numayr discussed earlier (cf. Schwentzel 2006: 128–131). Elsewhere in Nabataea, there is a Greek-Nabataean inscription at Jerash (ancient Gerasa in the Decapolis) defining the sacred precinct that refers to a statue perhaps of Aretas IV dated to 80/81 CE (C. B. Welles, “the Inscriptions” no. 1 in Kraeling 1938: 371– 373), but is too fragmentary to contribute to the evidence in the more explicit inscriptions at Petra. These dynastic statues at Petra were dedications by the chiefs of clans or priests, not the monarchs themselves. Their presence in the Temple precinct suggests the Nabataean kings wished to portray themselves and the benefactors of their kingdom on the divine order (Schwentzer 2006: 127). Although the Nabataean royal statues do not survive, their depiction of the monarchs is presumably similar to the iconographic images on their coins: a royal headband with long hair of corkscrew locks or curls. This is similar to the Hellenistic royal style emerging in the late second century and first century BCE. The last Seleucid kings in the first century BCE also were portrayed as youths with long curling hair, reflecting a heroic divine-like idea (Smith 1991: 24). Other examples of a divinizing

A NABATAEAN INSCRIPTION FROM BIR MADKHUR

205

king of the second century BCE are from Skythopolis in the Decapolis (Wenning 1983: 108–111= Smith 1991: fig. 264), some basalt heads from the Hauran of the first century BCE (Weber 2009: 84; Weber 2011). A coin image of Attembelos I (47–27 BCE) of Characene (Smith 1991: 225, fig. 280), reinforces the portrayal of the Nabataean royal iconography on their coins as a divinized heroes with diadem and long corkscrew curls.. A sculptured head in the Delos Museum has also been associated with Obodas II and Syllaeus’ dedication (Schmidt 1999: 279–298), but the badly defaced head seems more likely to represent Apollo, the god of the island (Schwentzel 2006: 134–135; Kropp 2013: 71). Another statue with long-flowing corkscrew locks from Egypt now in the Louvre in Paris (Ma 3546 = Smith 1988: 166–167, no. 56) has also been identified with a Nabataean king (Schmid 2013: 757–769), presumablys Malichos I (Schwentzel 2006: 135). It has iconographic affinity with several other statues with similar masculike-like image and Isis locks, the Papyri Kleopatra (Smith 1988: no. 24) and the Vienna Kleopatra (Smith 1988: no. 74; cf. Schwentzel 2006: 136–137, suggests identity with Aretas III) renders this dubious, and it remains more compelling to identify the three as the product of “a novel female royal style” that emerged in the mid-second century associated with the Ptolemaic queens Cleopatra I, II and III, who “ruled in behalf of or through boy kings and weak kings” (Smith 1995: 208–209). Assigning them later to the Nabataean kings in the first century BCE seems adventurous stylistically and chronologically. Another indication of the Nabataean royal cult is the titles and epithets employed by the Nabataean kings on their coins and in inscriptions that were similar to those used by the Hellenistic divinized dynasts. The first title of Aretas III on his Damascene issues was philhellenos in Greek, probably to portray himself as successor to the Seleucid kings (Bowersock 1983: 25–6), so not of great import. But later kings utilize epithets in Aramaic inscriptions that are similar to those used by other Hellenistic rulers. Aretas IV is ascribed with regularity as the one who “loves or cares for his people” (rḥm ʿmh), beginning at the inception of his reign (CIS II.197), essentially equivalent to Greek philopatris (“lover of his fatherland,” as with Archelaos of Cappadocia (36 BC-AD 17), implying the “protector” of his realm (De Callatay and Lorber 2011: 452). Later, Rabbel II is proclaimed as the one “who has given life and saves his people” (dy ʾḥyy wšyzh ʿmh), beginning in 5/6 CE (Meshorer

206

DAVID F. GRAF & ANDREW SMITH II

1975: 71, 75–76). The comparable Greek term would be Sōtēr (“savior”) or Euergetēs (“benefactor”), or perhaps both (Nock 1951: 127–148, Marshak 2015: 35 n. 11 and 231–238), reflecting an “external divine rescue”, associated with Dionysos’ eastern conquests (Smith 1988: 50), and a ruler who brought peace and order out of chaos (Nock 1951: 137). What prompted or provoked the use of these epithets by the Nabataean kings, expressing god-like qualities, remains unknown. In the few known instances elsewhere, it was the people (Appian, Syr. 45 and 69) or the army (Lucian, Zeuxis 11; Bickerman 1938: 237) that produced the epithets, not the monarchs themselves. Rulers normally were promoted to divinity for specific achievements or actions in behalf of their kingdom. The circumstances and agents in the case of the epithets for the Nabataean dynasts remain obscure, but they may relate to their gaining royal status, as each emerged to the throne in their minority (Schwentzel 2013:191–192), or, in the case of Rabbel II, his quelling of a rebellion may have led to his epithet (Winnett 1973: 54–57; Bowersock 1983: 156). It is interesting that philhellene, sōtēr and euergetēs are associated regularly with the Pontus dynasts, the Ptolemies and Seleucids, the Parthians, Characene or Bactria (De Callatay and Lober 2011: 450/1 and 455). For example, Attambelos I of Characene (47–24 BCE) is proclaimed Basileus Soter and Euergetes on his coins (Smith 1988: 120). In contrast, philorōmaios is never used by the Nabataeans at all, as one might expect of a client-kingdom of Rome. All of the epithets, philopatris, sōtēr and euergetēs, could be interpreted as suggesting aspects of a “divinized” ruler. The epithets became common in the Augustan era and early imperial period to express the wellbeing of a ruler’s province or a city (Nock 1951: 138–143), so they are entirely appropriate for a Nabataean dynast governing his kingdom. In spite of these Ptolemaic influences, there has been reticence to follow these “divine” elements to their logical conclusion. Schwentzel considers the Nabataean dynasty adapted (“revue et corrigé”) the Hellenistic model, according to their “interest propres,” without any suggestion of a royal cult (2005: 164) and Kropp suggests “the cultic overtones are mere accompaniment” (2010: 242). The divine elements are merely regarded as “sound and fury, signifying nothing,” empty symbols devoid of any content. In response, it may be suggested that it is unlikely mere iconographic propaganda and nomenclature would be successful in legitimizing,

A NABATAEAN INSCRIPTION FROM BIR MADKHUR

207

unifying and consolidating the Nabataean realm if it did not incorporate the same values as the Hellenistic royal ideology. The Nabataean kings were clearly portraying themselves like other Hellenistic kings. Although the emphasis has been on the Ptolemies, the ruler cult in the Seleucid realm should be considered, although the evidence is more fragmentary (Bickerman 1938: 236–57; Van Nuffelen 2004: 300; Kosmin 2014: 177–78). By 205 BCE, Antiochus III was the grand master, promoter and organizer of the dynastic cult in the Seleucid realm (Bickerman 1938: 247–48, 255; Van Nuffelen 2004: 279–285), marked also in 193 BCE by the launching of the empire-wide cult of his queen Laodike, associated with that of the king and the progonoi of the dynastic cult of the Seleucid dynasty (Robert 1967: 281–97; Kosmin 2014: 136–37 and 328 n. 103). His successors followed suit, as is reflected in the list of divinized Seleucid kings of the second century BCE from Teos (OGIS 246), which extends from Antiochus III to Demetrius II Nicator (145–140 and 129–125 BCE), including even Ptolemy VI Philometor, the only Ptolemy ever crowned as a king in the Seleucid Empire (Piejko 1982: 129–31). But the development of a Seleucid royal cult appears even earlier. An inscription from Lydia, dating to 228/9 CE, attests the cult of ‘Zeus Seleukios’, as the patron of the Seleucid dynastic line, indicating the survival of the third century BCE cult into the Roman era (Nock 1928: 42; Bikerman 1938: 255). This text is now supported by the papyri from DuraEuropos; which mention with the priesthoods of Zeus and Apollo, the progonoi of the Seleucid dynasty, and the cult of Seleucus I Nicanor, the founder of the Seleucid realm (P. Dura 25 dated to 180 CE, with comments of Kosmin 2014: 342 n. 174). A parallel for the Nabataean realm adopting the Ptolemaic practice is offered by the Seleucid king Antiochus XII. In his coinage, his epithets introduce Dionysos for the first time in 83/82 BCE: his titulary is Antiochos Dionysos Epiphanes Philopator Kallinikos, both on tetradrachms (Houghton et al. 2008: 2482) and the bronze issues (no. 2481–2483) of his reign. The inclusion of the Ptolemaic title “Dionysus” (cf. OGIS I, 258–260) may just express that he is the son of Antiochus VIII and Cleopatra Tryphaina, but it also perhaps reflects the regaining of Damascus with Ptolemaic support (Ehling 2008: 246). The other epithets of his titular are also important: “Epiphanes” refers to the Seleucid ruler-cult, a title associated with

208

DAVID F. GRAF & ANDREW SMITH II

at least nine Seleucid kings (see Kosmin 2014: 24 and 177–78). The dynastic title “Philopator” and the political-military name “Kallinikos” follow. The epithets of his coinage provide some insight into the political and ideological climate of the day. Such reverberations of Ptolemaic propaganda in this highly politically-charged atmosphere illustrate what may have developed even earlier among the Nabataeans, the allies of the Ptolemies now for several centuries. These contacts many have influenced the Nabataean kings to divinize their dynasty. Admittedly, this may be regarded as purely a hypothetical basis for asserting the ruler cult among the Nabataeans, but it is supported by iconographic and epigraphic evidence. Furthermore, as is well known, these practices were adopted by the Roman emperors, beginning with the establishment of the imperial cult under Augustus, which spread across the Mediterranean world, including Egypt (Dörner 2014: 153–276), SyriaPalestine and Arabia (see Fritz Graf 2008: 760–63, for documentation). These divine honors were extended even to the Roman “queens,” beginning with the worship of the “empress” Livia, the Diva Augusta, after they had been vetoed by the senate after her death in 29 CE (Suet. Tib. 51), were instituted by the Emperor Claudius in 42 CE, precisely on January 17, the day of her marriage to Octavian (Suetonius, Claudius 11.2; Dio Cassius 60.5 [41] with Levick 1990: 47, citing ILS 4995 and coins). In the East, earlier recognition of Livias was given by Herod Antipas, who fortified Bethramphtha in the Peraea and renamed it Livias-Julias (Jos. AJ 18.27) and the renaming of Bethsaida in Galilee Julia, after the emperor’s daughter (Jos. AJ 18.17), joining the imperial cities of Caesarea Maritima and Sebaste created for Augustus. It has even been surmised that Livia appeared in the guise of Roma in the Augusteum at Casarea, Sebaste, and Banias (Bernett 2007; cf. Marshak 2015: 212–17). Finally, Marcus Agrippa, the administrator of the eastern empire after 18 BCE, was recognized as euergetēs kai sōtēr at Athens (ID 1593), and the Theōi Sōtēr in Larissa in Thessaly and Mytilene on Lesbos, celebrated in a festival on Kos (Sylloge3 1065), and a month named after him on Cyprus, all during his lifetime and before 12 BCE (Habicht 2005: 242–246). His association with Herod and visit to his realm in 15 CE (AJ 16.12–26; cf. 157) may have produced similar attestation in Palestine, but the evidence is lacking. Whatever the case, the Nabataeans could not have been oblivious to Herod’s promotion of the imperial cult within his realm.

A NABATAEAN INSCRIPTION FROM BIR MADKHUR

209

Finally, Angelos Chaniotis brought to my attention a similar phenomenon to baseleophoric names that appears in the Greek East: an individual, usually a member of the elite, occasionally designated a ruler as “his own god.” For example, with Mark Antony at Alexandria in 34 BCE (OGIS 195: τὸν ἑαυτοῦ θεὸν καὶ εὐεργέτην), or for Tiberius on Cyprus (OGIS 283 line 10: τῶι ἁτοῦ θεῶι), or Nero on Cyprus in AD 60/61 (ISalamis XIII no. 135: τῷ ἰδίῳ θεῷ καὶ σωτῆρι), or King Sauromates I of Hermonassa on the Black Sea c. 93–124 CE (IOSPE II 358: [τὸ]ν ἴδι[ον θεὸν? καὶ σω]τῆρα), or Marcus Aurelius (IGR IV 685: τὸν ἴδιον θεὸν [καὶ] εὐεργέτην)); or King Sauromates II; Hermonassa and Phanagoreia, ca. 174–210 CE (IOSPE II 357 and SEG LVI 931: τὸν ἴδιον θεὸν καὶ δεσπότην). The same usage is attested also for gods, but less often: In Kos, it was used by a priest, who dedicated a statue “of his own goddess” (IG XII.4.621, second century), and in Egypt for an anonymous god, possibly Pan (Bernand 1984: no. 98, τῷ ἰδίῳ θεῷ μεγίστῳ. Imperial period). The emotional weight of this phrase becomes clear when we see how it was used by the parents of a child who died at the age of four, when they affectionately call him θεὸς ἴδιος ἐπήκοος (IGUR 1702, “their own god, who listens to them”, θεῷ ἰδίῳ ἐπηκόῳ. Rome, third century CE). Such names denote – through “religious” devotion – loyalty. To name a child after a king, instead of using a theophoric name, indicates a similar display of affection and loyalty to the dynast or to the ruler cult. When all of this evidence is taken into consideration, Clermont-Ganneau’s observation gains even more credibility and remains the most plausible explanation for the phenomenon of the “servant names” in Nabataean Aramaic. The evidence of the Nabataean dynasty developing along the lines of a Hellenistic and Ptolemaic model are considerable. The baseileophoric Nabataean names are particularly common among administrative and military officials, as were the philo-basileistes names among Ptolemaic officials. The Nabataean servant-names can then be seen as not merely expressions of loyalty to the royal dynasty, but also as indicators of absorption of the ruler cult by inhabitants of the Nabataean realm (Graf 1994: 291–293). Although this Bir Madkhur text dates long after the Nabataean dynasty ended, it still constitutes an important addition to the ʿBD-ʿBDT occurrences throughout the Nabataean realm, and the first attestation between Petra and ʿAvdat at an important site along the caravan route leading to Gaza. As we know

210

DAVID F. GRAF & ANDREW SMITH II

from ʿAvdat, the worship of “Obodas the god” did not end with the Roman annexation, but survived into the early Roman imperial period.

BIBLIOGRAPHY

Al-Khraysheh, F. (1986). Die Personennamen in den nabatäischen Inschriften des Corpus Inscriptionum Semiticarum. Ph.D. Dissertation, Philipps-Universität Marburg/Lahn. Alpass, P. (2010). The Basileion of Isis and the religious art of Nabataean Petra. Syria 87: 93–113. Alpass, P. (2013). The Religious Life of Nabataea. Religions in the Greco-Roman World 175. Leiden: Brill. Al-Salameen, Z. and H. Falahat (2012). Two New Nabataean Inscriptions from Wādī Mūsā, with Discussion of Gaia and the Marzēaḥ. Journal of Semitic Studies 57: 37–51. Al-Salameen, Z. and H. Falahat (2014). An Inscribed Nabataean Bronze Object dedicated to Obodas the God from Wādī Mūsā, Southern Jordan. PEQ 140: 293–307. Al-Salameen, Z. and Y. Shudaifat (2014). A new Nabataean inscribed bronze lamp. AAE 25: 43–49. Al-Theeb, Solaiman Abdal-Rahman (1993). Aramaic and Nabataean Inscriptions from North-West Saudi Arabia. Riyadh: King Fahd National Library. Appian. Roman War, 11: Syrian Wars. Translated by H. White (1912). Harvard: Harvard University Press. Aufrecht, W. E. (1989). A Corpus of Ammonite Inscriptions. Lewiston, NY: E. Muellen. Beeston, A. F. L. (1984). Sabaic Grammar. Journal of Semitic Studies Monograph 6. Manchester. University of Manchester. Bellinger, A. R. (1949). The End of the Seleucids. Transactions of the Connecticut Academy of Arts and Sciences 38: 55–102. Benz, Frank L. (1972). Personal Names in the Phoenician and Punic Inscriptions. Studia Pohl 8. Rome: Biblical Institute Press. Bernett, M. (2007). Der Kaiserkult in Judäa unter den Herodiern und Römern: Untersuchungen zur politischen und religiösen Geschichte Judäas von 30 v. bis 66 n. Chr. Wissenschaftliche Unter-

A NABATAEAN INSCRIPTION FROM BIR MADKHUR

211

suchungen zum Neuen Testament 203. Tübingen: Mohr Siebeck. Bernand, A. (1984). Les portes du désert. Recueil des inscriptions grecques d’Antinooupolis, Tentyris, Koptos, Apollonopolis Parva et Apollonopolis Magna. Paris: Éditions du CNRS. Bickerman, E. J. (1938). Institutions des Séleucides. BAH 26. Paris: Geuthner. Bilberstein-Kazimirski, A. de (1960). Dictionnaire arabe–français contenant toites les raciness de la langue arabe. Paris: G. P. Maisonneuve. Bikai, P., Kanellopoulos, C. and S. Saudeners. (2008). Beidha in Jordan: A Dionysian Hall in a Nabataean Landscape. AJA 112: 465–507. Bordreuil, P. and A. Lemaire (1976). Nouveaux sceaux hébreux, araméens et ammonites. Semitica 26: 45–63. Bowersock, G. W. (1983). Roman Arabia. Cambridge: Harvard University Press. Bowersock, G. W. (1990). The Cult and Representation of Dushares in Roman Arabia. In: F. Zayadine, ed., Petra and the Caravan Cities. Amman: Department of Antiquities, pp. 31–36 Bricault, L. (1992). Isis Dolente. Bulletin de Institut française d’archéologique orientale 92: 37–49. Bricault, L. (1999). Sarapis et Isis: Sauveurs de Ptolémé IV à Raphia. Chronique d’Égypte 74, fasc. 148: 334–43. Bricault, L. (2001). Atlas de la diffusion des cultes isiaques (IVe siecle av. JC.-IV siecle apr. J-C.). Memoires de l’Academie des Inscriptions et Belles-Lettres 23. Paris: de Boccard. Bricault, L. (2005. Recueil des inscriptions concernant les cultes isiaques. Mémoires de l’Académie des inscriptions et belles-lettres. Nouvelle série 31. Paris: de Boccard. Bricault, L. (2008). Sylloge nummorum religionis isiacae et sarapiacae (SNRIS). Mémoires de l’Académie des Inscriptions et Belles-Lettres 38. Paris: de Boccard. Caneva, S. G. (2012). Queens and Ruler Cults in Early Hellenism: Festivals, Administration, and Ideology. Kernos 25: 75–101. Cantineau, J. (1930). Le Nabatéen, Vol. I. Paris: Librairie Ernest Leroux.

212

DAVID F. GRAF & ANDREW SMITH II

Cantineau, J. (1932). Le Nabatéen, Vol. II. Paris: Librairie Ernest Leroux. Carney, E. (2000). The Initiation of Cult for Royal Macedonian Women. Classical Philology 95: 21–43. Chaniotis, A. (2003). The Divinity of Hellenistic Rulers. In: A. Erskine, ed., A Companion to the Hellenistic World. Malden, MA: Blackwell, pp. 431–445 Clermont-Ganneau, Ch. (1888). Les noms royaux nabatéens employés comme noms divins. Recueil d’archéologie orientale I (Paris, E. Leroux): 39–47 = Revue Archéologique, Troisième Série, 5 (1885): 170–178. Colin, F. (1994). L’Isis «dynastique» et la mère des dieux phrygienne: essai d’analyse d'un processus d'interaction culturelle. ZPE 102: 271–295. Collombert, P. (2008). La ‘stèle de Saïs’et l’instauration du culte d’Arsinoé II dans la chora. Ancient Society 38: 83–99. Dalman, G. (1912). Neue Petra-forschungen und Der heilige felsen von Jerusalem. Leipzig, J. C. Hinrich. De Callatay, F. and Lorber, C. C. (2011). The pattern of royal epithets on Hellenistic coinages. In: P. Iossif et al., ed., More than men, less than gods: studies on royal cult and imperial worship. Studia Hellenistica 51. Leuven: Peeters, pp. 417–55. Dijkstra, K. (1995). Life and loyalty : a study in the socio-religious culture of Syria and Mesopotamia in the Graeco-Roman period based on epigraphical evidence. Leiden: Brill. Dörner, N. (2014). Feste und Opfer für den Gott Caesar: Kommunikationsprozesse im Rahmen des Kaiserkultes im römischen Ägypten der julisch-claudischen Zeit [30 v.Chr. – 68 n.Chr.]. Pharos 30, Rahden: VML Marie Leidorf. Ehling, K. (2008). Untersuchungen zur Geschichte der späten Seleuckiden (164–53 v. Chr.). Historia Einzelschriften 196. Stuttart: Franz Steiner. El-Khouri, L. (2002). The Nabataean Terracotta Figurines. BAR-IS 1034. Oxford: Archaeopress. Erickson-Gini, T. (2014). Oboda and the Nabateans. Srata: Bulletin of the Anglo-Israel Archaeological Society 32: 81–108.

A NABATAEAN INSCRIPTION FROM BIR MADKHUR

213

Finlayson, C. (2016). The Obelisk, the Crow-Step, and the Elephant in Nabataean Contexts. In: N. Khairy, ed., Studies in Nabataean Culture, vol. II. Amman: University of Jordan, pp. 75–104 FGrH = F. Jacoby, Fragmente der grieschischen Historike. Leiden: Brill. Fraser, P. M. (1972). Ptolemaic Alexandria. 3 vols. Oxford: Clarendon. Gradel, I. (2002). Emperor Worship and Roman Religion. Oxford: Oxford University Press. Graf, D. F. (1992). Nabataeans. Anchor Bible Dictionary IV. New York: Doubleday, pp. 970–73. Graf, D. F. (1994). The Nabataean Army and the Cohortes Ulpiae Petraeorum. In: E. Dabrowa, ed., The Roman and Byzantine Army in the East,. Kraków: Uniwersytet Jagiellonski, pp. 265–311 Graf, D. F. (2006). The Nabataeans in the Early Hellenistic Period: The Testimony of Posidippus of Pella. Topoi: Orient-Occident 14: 47–68. Graf, D. F. (2014). Nabataean Inscriptions from Wādī Musemīr in the Southeast Wādiī ʿAraba. In: S. T. Parker and A. Smith, eds, The Roman Aqaba Project Final Report. Volume 1: The Rewhional Environment and Region Survey. Boston, MA: ASOR, pp. 349–356 Graf, D. F., Schmidt, S. G., Ronza, E. and S.E. Sidebotham. (2007). The Hellenistic Petra Project: Excavations in the Qasr al-Bint Temenos Area: Preliminary Report of the Second Season, 2005. ADAJ 51: 223–238. Graf, F. (2008). Ruler Cult. Brill’s New Pauly 12: 760–763. Gruendler, B. (1993). The Development of the Arabic Scripts: from the Nabatean era to the first Islamic century according to dated texts. Harvard Semitic Studies 43. Atlanta: Scholars Press. Habicht, C. (1970). Gottmenschentum und griechische Städte. Zetemata 14. München. Beck. Habicht, C. (2005). Marcus Agrippa Theos Soter Hypeboreas 11/2: 242–246. Hayajneh, H. (1998). Die Personenname in den qatabānische Inschriften: Lexikalische und grammatische Analyse in Kontext der semitischen An-

214

DAVID F. GRAF & ANDREW SMITH II

throponomastik. Texte und Studien zur Orientalisitik 10. Hildesheim: George Olms. Hayajneh, H., Ababneh, M. A., and al-Khraysheh, D. (2015). Die Götter von Ammon, Moab und Edom in einer neuen frühnordarabischen Inscrift aus Südost-Jordanien. In: V. Golinets et al., eds, Neue Beiträge zur Semitistk, Münster: UgaritVerlag, pp. 79–105 Harding, G. L. (1971). An Index and Concordance of Pre-Islamic Arabian Names and Inscriptions. Near and Middle East Series 8. Toronto: University of Toronto. Hazzard, R. A. (2000). Imagination of a Monarchy: Studies in Ptolemaic Propaganda. Toronto: University of Toronto Press. Hauben, H. (2011). Ptolémée III et Bérénice II, divinités cosmiques. In: P. Iossif et al., ed., More than men, less than gods: studies on royal cult and imperial worship. Studia Hellenistica 51. Leuven: Peeters, pp. 357–388. Healey, J. (1975). Maliku: mlkm: Annunaki. Ugarit-Forschungen 7: 235–238. Healey, J. (1990–1991). Nabataean to Arabic: Calligraphy and script development among the pre-Islamic Arabs. In: F. Déroche, A. Gacek and J. J. Witkam, eds, Manuscripts of the Middle East, vol. 5, Leiden: Ter Lugt Press, pp. 41–51 Healey, J. (2001). The Religion of the Nabataeans: A Conspectus. Religions in the Greco-Roman World 136. Leiden: Brill. Heider, G. C. (1992). Moloch. Anchor Bible Dictionary 4: 895–98. Hess, R. (1993). Amarna Personal Names. Winnona Lake: Eisenbrauns. Hölbl, G. (2001). A History of the Ptolemaic Empire. New York: Routledge. Hoftijzer, J. and K. Jongeling. (1995). Dictionary of the North-West Semitic Inscriptions. Leiden: Brill. Hoover, O. and Barkay, R. (2010). Important Additions to the Corpus of Nabataean Coins since 1990. In: M. Huth and P. G. van Alfen, eds, Coinage of the Caravan Kingdoms: Studies in Ancient Arabian Monetization,. New York: The American Numismatic Society, pp. 197–212.

A NABATAEAN INSCRIPTION FROM BIR MADKHUR

215

Houghton, A. Lorber, C. and O. Hoover. (2008). Seleucid Coins: A Comprehensive Catalogue, part 2: Seleucus IV through Antiochus XIII, 2 vols. New York: American Numismatic Society. Huffmon, H. B. (1965). Amorite Personal Names in the Mari Texts. Baltimore: John Hopkins Press. Huth, M. (2010). Some Nabataean Questions Reconsidered. In: M. Huth and P. G. van Alfen, eds, Coinage of the Caravan Kingdoms: Studies in Ancient Arabian Monetization. New York: The American Numismatic Society, pp. 213–226 Janif, M. (2004). L’écrit et le “figuré” dans le domaine relioeux des Nabatéenns: Le sanctuaire rupestre du Sadd al-Mreriyyeh à Pétra. Syria 81: 119–130. Josephus, Antiquities of the Jews. Translated by L. Feldman and R. Marcus (1980–1981). Harvard: Harvard University Press. Josephus, Jewish War. Translated by L. Feldman and R. Marcus (1980–1981). Harvard: Harvard University Press. Kahil, Likely (1996). Cults in Hellenistic Alexandria. In: J. Walsh and T. F. Reese, eds, Alexandria and Alexandrianism. Malibu, CA: J. Paul Getty Museum, pp. 75–84. Kaizer. T. (2010). “Kings and Gods: Some thoughts on religious patterns in Oriental principalities.” In: T, Kaizer and M. Facella, eds, Kingdoms and Principalities in the Roman Near East, Orinet et Occidens 19. Stuttgart: Steiner, pp. 113–124 Kosmin, P. J. (2014). The Land of the Elephants: Space, Territory and Ideology in the Seleucid Empire. Cambridge: Harvard University Press. King, G. M. D. (1990). Early North Arabian Thamudic E. Ph.D. dissertation, University of London, School of Oriental and African Studies. Koenen, L. (1993). The Ptolemaic king as a religious figure. In: A. Bulloch et al., eds, Images and ideologies: Self-definition in the Hellenistic World, ed. Berkeley: University of California, pp. 25–115. Kraeling, C. (1938). Gerasa, City oif the Decapolis. New Haven: Yale: American Schools of Oriental Research. Kropp. A. J. M. (2011). Nabataean Dushārā (Dusares)–an Overlooked Cuirassed God. PEQ 143: 176–197.

216

DAVID F. GRAF & ANDREW SMITH II

Kropp. A. J. M. (2013a). Images and Monuments of Near Eastern Dynasts, 100 BC-AD 100. Oxford: Oxford University Press. Kropp. A. J. M. (2013b). Kings in Cuirass – Some Overlooked Full-Length Portraits of Herodian and Nabataean Dyansts. Levant 45: 45–56. Lane, W. (1980). An Arabic-English Lexicon. Beirut: Libraire du Liban. [reprint of 1863–93 edition]. Levick, B. (1990). Claudius. New Haven & London: Yale University Press. Littmann, E. (1945–49). Neues zur altnordarabischen Dialektkunde. ZDMG 99: 168–180. Lucian, Zeuxis. Magness, J. (2001). The Cults of Isis and Kore at Samaria-Sebaste in the Hellenistic and Roman Periods. Harvard Theological Review 94: 157–77. Marshak, A. K. (2015). The Many Faces of Herod the Great. Grand Rapids: Eredmans. McKenzie, J. (1990). The Architecture of Petra. Oxford: Oxford University Press. Merklein, H. and Wenning, R. (1998). Ein Verehrungplatz der Isis in Petra neu untersucht. ZDPV 114: 162–178. Meshorer, Y. (1975). Nabataean Coins. Qedem 3. Jerusalem: Hebrew University. Milik, J. T. (1959a). Notes d’épigraphie et topographie palestinniennes. RB 66: 550–75. Milik, J. T. (1959b). Notes d’épigraphie et de topographie jordaniennes. LA 10: 147–84. Milik, J. T. (2003). Une bilingue araméo-grecque de 105/104 avant J.-C. In: J. Dentzer-Feydy et al., eds, Hauran II. Bibliothèque archéologique et historique 164. Beyrouth: Institut français du Proche-Orient, pp. 269–74. Milik, J. T. and J. Starcky. (1975). Inscriptions récemment découvertes à Pétra. ADAJ 20: 111–30. Montanari, F. (2015). The Brill Dictionary of Ancient Greek. Leiden: Brill.

A NABATAEAN INSCRIPTION FROM BIR MADKHUR

217

Moscati, S. (1959). Nordarabico, sudarabico, etiopico. Rivista degli studi orientali 34: 33–39. Müller, H. (2000). Der hellenistische Archiereus. Chiron 30 : 519– 42. Müller, S. (2009). Das hellenistische Königspaar in der medialen Repräsentation: Ptolemaios II. und Arsinoe II. Beiträge zur Altertumskunde 263. Berlin/New York: de Gruyter. Negev, A. (1981). The Greek Inscriptions from the Negev. Studium Biblicum Franciscanum. Collectio Minor 25. Jerusalem: Franciscan Press. Negev, A. (1986a). Nabataean Archaeology Today. New York & London: New York University Press. Negev, A. (1986b). Obodas the God. IEJ 36: 56–60. Negev, A. (1991). Personal names in the Nabataean Realm. Qedem 32. Jerusalem: Institute of Archaeology, The Hebrew University. Negev, A. (2003). Obodas the God in a Nabataean-Arabic Inscription from the Vicinity of Oboda and a Review of Other Nabataean Inscriptions. In: R. Rosenthal-Heginbottom, ed., The Nabataeans in the Negev. Haifa: Reuben and Edith Hecht Museum, pp. 17*–20*. Nehmé, L. (2002a). La chapelle d’Obodas à Pétra. Rapport preliminaries sur la champagne 2001. ADAJ 46: 243–56. Nehmé, L. (2002b). Petra. The Obodas Chapel. AJA 106: 452–455. Nehmé, L. (2012). Le dieu Obodas chez les Nabatéens: hypothèes anciennes et découvertes récentes. In: I. Sachet and C. Robin, eds., Dieux etdéesses d’Arabie images et representations. Orient et Méditerranée 7. Paris: De Boccard, pp. 181–224 Newell, E. T. (1939). Late Seleucid Mints in Ake-Ptolemais and Damascus. Numismatic Notes and Monographs 84. New York: American Numismatic Society. Nock, A. D. (1928). Notes on Ruler-Cult, I-IV. Journal of Hellenic Studies 48: 21–43. Nock, A. D. (1930). Σύνναος Θεός. HSCP 41: 1–62 = A. D. Nock. (1972). Essays on Religion and the Ancient World I. Cambridge: Harvard University Press, pp. 202–51.

218

DAVID F. GRAF & ANDREW SMITH II

Nock, A. D. (1951). Soter and Euergetes. In: S. L. Johnson, ed, The Joy of study: Papers on new Testament and related subjects Presented to Honor Frederick Clifton Grant. New York: Macmillan, pp. 127– 48 = A. D. Nock. (1972). Essays on Religion and the Ancient World II. Cambridge: Harvard University Press, pp. 720–35. Pfeiffer, S. (2008). Herrscher- und Dynastiekulte im Ptolemäerreich : Systematik und Einordnung der Kultformen. Münchener Beiträge zur Papyrusforschung und antiken Rechtsgeschichte 98. München: Beck. Piejko, F. (1982). Ptolemies in a List of Deified Seleucids from Teos. OGIS 246; ZPE 49: 129–131. Plantzos, D. (2011). The iconography of assimilation: Isis and royal imagery on Ptolemaic seal impressions. In: P. Iossif et al., ed., More than men, less than gods: studies on royal cult and imperial worship. Studia Hellenistica 51. Leuven: Peeters, pp. 389–415. Plutarch, Antony. Translated by B. Perrin. (1912–1926). Harvard: Harvard University Press. Plutarch, Demosthenes. Translated by B. Perrin. (1912–1926). Harvard: Harvard University Press. Porten, B. and A. Yardeni. (2014). Textbook of Aramaic Ostraca from Idumea, Vol. 1. Winona Lake, IN: Eisenbrauns. Price, S. F. R. (1984). Gods and Emperors: The Greek Language of the Roman Imperial Cult. Journal of Hellenic Studies 104: 79–95. Quaegebeur J. (1978). Reines ptolémaïques et traditions égyptiennes. In: H. von Maehler and V. M. Strocka. Das ptolemäische Ägypten. ed. Mainz : von Zabern, pp. 245–62. Quaegebeur J. (1989). The Egyptian Clergy and the Cult of the Ptolemaic Dynasty. Ancient Society 20: 93–116. Quaegebeur J. (1998a). Documents égyptiens anciens et nouveaux relatifs à Arsinoé Philadelpia. In: H. Melaerts, ed., Le culte du souverain dans l'Égypte ptolémaïque au IIIe siècle avant notre ère. Leuven: Studia Hellenistica 34, pp. 73–108 Quaegebeur J. (1998b). Cleopatra VII and the cults of the Ptolemaic queens. In: R. S. Bianchi et al., eds, Cleopatra’s Egypt: Age of the Ptolemies, Brooklyn: The Brooklyn Museum, pp. 41–54 RES = Répertoire d’épigraphie sémitique (Paris).

A NABATAEAN INSCRIPTION FROM BIR MADKHUR

219

Robert, L. (1967). Encore une inscription grecque de l’Iran. CRAI 111: 281–297. Roche, M.-J. (1987). Le cultes d’Isis et l’influence égyptienne à Pétra. Syria 64: 217–222. Roller, D. W. (2004). The World of Juba II and Kleopatra Selene: royal scholarship on Rome’s African frontier. New York: Routledge. Roller, D. W. (2010). Cleopatra: A Biography. Oxford: Oxford University Press. Schwentzer, C.-G. (2005). Les thèmes du monnayage royal nabatéen et le modèle monarchique hellénisique. Syria 82: 149–60. Schwentzer, C.-G. (2006). Statues royales nabatéennes. Res Antiquae 3: 123–137. Schwentzer, C.-G. (2008). La double representation du dieu Dousares: bilinguisme figurative ou syncrétisme? Metis n.s. 6: 149– 66. Schwentzer, C.-G. (2010). Arétas IV ‘roi des Nabatéens’ d’après monnaies. Numismatica e antichità classiche 39: 233–49. Schwentzer, C.-G. (2013). Juifs et Nabatéens. Les monarchies ethniques du Proche-Orient hellénistique et romain. Rennes: University of Rennes. Schwentzer, C.-G. (2014). “La reine Huldu et coiffe isiaque. Isis et le pouvoir royal à Pétra (Ier s. av. J.-C.-Ier s. ap. J.-C.). In: L. Bricault and M. J. Versluys. Power, Politics and the Cults of Isis. Leiden: Brill, pp. 147–162 Schmid, S. G. (1999). Un roi nabatéen à Délos. ADAJ 43: 279–298. Schmid, S. G (2013) Early Nabataean Royal Portriture. SHAJ 11: 757–769. Schmitt, G. (1995). Siedlungen Palästinas in griechisch-römischer Zeit: Ostjordanland, Negeb und (in Auswahl) Westjordanland. Beihefte zum Tübinger Atlas des vorderen Orients, Reihe B, Nr. 93. Wiesbaden: Reichert. Silverman, M. (1981). Servant (ʿebed) Names in Aramaic and Other Semitic Languages. Journal of the American Oriental Society 101: 361–366. Smith, A.M., II (2010a). Wadi Araba in Classical and Late Antiquity: A Historical Geography. BAR-IS No. 2173. Oxford: Archaeopress.

220

DAVID F. GRAF & ANDREW SMITH II

Smith, A.M., II (2010b). The Bir Madhkur Project: A Preliminary Report on the 2008 Field Season. ADAJ 54: 143–52. Smith, R. R. R. (1988). Hellenistic Royal Portraits. Oxford: Oxford University Press. Smith, R. R. R. (1991). Hellenistic Sculpture: a Handbook. New York: Thames and Hudson. Smith, R. R. R. (1996). Ptolemaic Portraits: Alexandrian types, Egyptian version. In: J. Walsh and T. F. Reese, eds, Alexandria and Alexandrianism. Malibu, CA: J. Paul Getty Museum, pp. 203–213. Strabo, Geography (LCL) Starcky, J. (1966). Pétra et la Nabatène. DBS 7: cols. 886–1017. Starcky, J. and J. Strugnell. (1966). Petra: deux nouvelles inscriptions nabatéenes. RB 73: 236–243. Stark, J. K. (1971). Personal Names in Palmyrene Inscriptions. Oxford: Clarendon Press. Strootman, R. (2014). Courts and Elites in the Hellenistic Empire. Edinburgh: Edinburgh University Press. Thompson, Dorothy J. (2012). Memphis under the Ptolemies. Second edition. Princeton University Press. Tholbecq, L., C. Durand and C. Bouchaud (2008). Nabataean Rock-Cut Sanctuary in Petra: Second Preliminary Report on the ‘Obodas Chapel’ Excavation Project, Jabal Numayr (2005–2007). ADAJ 52: 235–254. Tholbecq, L. and C. Durand (2013). A late second century BC Nabataean occupation at Jabal Numayr: the earliest phase of the ‘Obodas Chapel’ sanctuary. In: M. Mouton & S. G. Schmid, eds, Men on the Rocks: The Formation of Nabataean Petra. Berlin: Logos, pp. 205–222 Tondriau, J. (1946). Les thiases dionysiaques royaux de la cour ptolémaïque. Chronique d’Égypte 21: 149–71. Tondriau, J. (1948). Rois lagides compares ou identifies à des divinités. Chronique d’Égypte 23: 127–146. Tondriau, J. (1950). La dynastie ptolémaïque et la religion dionisiaque. Chronique d’Égypte 25: 283–316.

A NABATAEAN INSCRIPTION FROM BIR MADKHUR

221

Vaelske, V. (2013). Isis in Petra: Chronological and Topographical Aspects.” In: M. Mouton and S. G. Schmid, eds, Men on the Rocks: The Formation of Nabataean Petra. Berlin: Logos, pp. 351– 361 Van Nuffelen, P. (1999). Le culte des souverains hellénistiques: le gui de la religion grecque. Ancient Society 29: 175–189. Van Nuffelen, P. (2004). Le culte royal de l’empire des Séleucides: une reinterpretation. Historia 53: 278–301. Walther, W. (1966). Untersuchungen zu vor- und frühislamischen arabischen Personennamen. Dissertation, Martin-Luther-Universität Halle-Wittenberrg. Weber, T. (2006). Sculptures from Roman Syria in the Syrian National Museum at Damascus. In Cooperation with Qasem AlMohammed. Worms: Wernersche Verlagsgellschaft. Weber, T. (2009). Die Skulpturen aus Sahr und die Statuendenkmäler der römischen Kaiserzeit in südsyrischen Heiligtümern. In: J.-M. Dentzer and T. M. Weber, eds, Hauran. 4, Sahr al-Ledja: recherches syro-européennes 1998–2008 = syrisch-europäische Forschungen 1998–2008, Vol. II. Bibliothèque archéologique et historique 184. Beyrout: Institut français du Proche-Orient, pp. 1– 263 Weber, T. (2011). Arabia in exotischem Gewand. Antike Welt 6(11): 11–16. Wenning, R. (1983). Hellenistische Skulpturen in Israel. Boreas 6: 105–118. Wenning, R. (1997). “Bemerkungen zur Gesselschaft und Religion der Nabatäer.” In: R. Albertz, ed., Religion und Gesellschaft. Studien zur ihrer Wechselbeziehung in der Kultur des Antiken Vordereen Orients. Munster: Ugarit-Verlag, pp. 177–201 Wenning, R. (2015). Obodas Theos: Der statuarische Befund eines nabatäischen Gottes. ZDPV 131: 44–58. Wenning, R. and H. Merklein. (1997). Der Götter in der Welt der Nabatäer. In: T. M. Weber and R. Wenning eds, Petra. Antike Felstadt zwischen arabischen Tradition und griechischer Norm. Mainz: Zabern, pp. 105–110. West, J. M. I. (1974). Uranius. HSCP 78: 282–84.

222

DAVID F. GRAF & ANDREW SMITH II

Winnett, F. V. (1973). The revolt of Damasī: Safaitic and Nabataean Evidence. BASOR 211: 54–57. Winter, E. (1978). Der Herrscherkult in den ägyptischen Ptolemäertempeln. In: H. von Maehler and V. M. Strocka, eds, Das ptolemäische Ägypten, Mainz: von Zabern, pp. 147–60 Wuthnow, H. (1930). Die semitischen Menschennamem in griechischen Inschriften und Papyri des vorderen Orients. Leipzig: Dieterich. Yardeni, A. (2000). Textbook of Aramaic, Hebrew and Nabataean: Documentary Texts from the Judaean Desert and Related Material. Jerusalem: Hebrew University. Zayadine, F. (1981). L’iconographie d’Al-ʿUzza-Aphrodite. In: L. Kahil and Chr. Augé, eds, Mythologie gréco-romaine. Mythologies périphériques. Études d'iconographie. Paris : CNRS, pp. 113–118 Zayadine, F (1986). A Symposiarch from Petra. In: L. Geraty and L. Herr, eds, The Archaeology of Jordan and Other Studies Presented to Siegfried H. Horn. Berrien Springs: Andrews University Press, pp. 465–474. Zayadine, F (1990). The pantheon of the Nabataean inscriptions in Egypt and the Sinai. Aram 2: 151–174. Zayadine, F (1991). L’iconographie d’Isis à Petra. Mélanges de l’École française de Rome 103: 283–306. Zayadine, F., Larche, F. and J. Dentzer-Feydy. (2003). Le Qasr alBint de Pétra: l’architecture, le décor, la chronologie et les dieux. Paris: Recherche sur les civilisations.

FOLLOWING THE ROMAN PAYMASTER JOHN W. BETLYON 1

As the Roman state grew in power and prestige, Roman military forces were deployed throughout Italy, Europe, and into North Africa and Southwest Asia. Everywhere the troops went, Roman military engineers provided for the security of the soldiers, building different types of fortresses and installations. Archaeological study of these castra and castella has yielded a history of military operations throughout lands controlled by Rome. Wherever Rome sent its soldiers, it also paid its soldiers. Archaeological work at sites such as el-Lejjūn, Umm el-Jimal, and Roman Aila, has yielded evidence for the deployment of Roman forces into Arabia. With soldiers deployed on the Limes Arabicus, the Paymaster was never far behind making certain that the troops were paid for their service in “coin of the realm.” The deployment of the Roman army throughout the empire is well documented. The army was tasked with several interlocking missions. First of all, it was forwardly deployed to the frontiers of the Empire to defend Rome and its lands against foreign invasion. Particularly enemies of the state from North Africa, Germania, and the East demanded attention. Secondly, the army was deployed to protect and enhance commerce, trade, and travel throughout the It has been my great joy to have been associated with S. Thomas Parker on several archaeological projects which have focused on Roman installations in the borderlands. We ate the same dirt at el-Lejjun and Umm el-Jimal as did the legionnaires of Rome, and I studied the coins Dr. Parker and his teams unearthed at these important frontier sites, and at Roman Aqaba, or Aila. 1

225

226

JOHN W. BETLYON

Empire. Over the years, the particulars of the identities of the enemies changed, but the threats remained real. Armies are deployed and trained to deal with these “threats.” These are not imagined. Soldiers, for example, might be sent to the eastern fringes of the Empire, into lands otherwise inhospitable for habitation, to provide a buffer against possible encroachment or invasion by a potential enemy of Rome. Thirdly, the army also provided the necessary engineering and manpower to build roads, bridges, and other public works. Roman roads linked military castra and castella with population centers. This system of roads helped to fuel a growing economy. Consequently, archaeologists find remains of fortifications on these roads and on all frontiers of the Empire. At these sites, we also find examples of the money which was used to pay those soldiers and to provide for their sustenance and maintenance in the local economies. In the East, for example, work at sites such as el-Lejjūn, Umm el-Jimal, Qatsrin, and Aila, has revealed the remains of fortifications, some of which housed military units. Similar buildings are found on the western frontiers, in Gaul, Hispania, Germania, and Britannia (Cf. Sarantis 2013: passim). Roman military doctrine of the fourth century dictated that the frontiers would be defended in depth, with large, legionnaire formations stationed intermittently, with smaller, more mobile forces, including cavalry, billeted between the larger castra (Luttwak 1976: 130–136). Frontiers were also developed to include new roads as means of communication, linking larger installations and their battle formations, with smaller, more lightly armed forces which provided “eyes” on the borders and “ears” to the ground, gathering intelligence on the whereabouts and movements of potential adversaries (Whately 2013: 236– 37). Coins were found in excavations at the several legionnaire fortresses, as well as at smaller castella and watch towers interspersed between the larger castra along the Roman-Arabian border (Cf. Betlyon 2006, for example). It was standard policy, of course, to pay the troops on a regular basis. Soldiers needed to be “kept happy” and contented in their service to the Emperor. It has been said that armies travel “on their stomachs.” That is true, but soldiers also expect to be paid in full, and on time, and they expect to be fed properly. The rigors of life in the field are part of a soldier’s life, and these things can be endured, so long as there is access to good food and their pay. Often

FOLLOWING THE ROMAN PAYMASTER

227

with little to spend their money on in the borderlands, food was probably the most important item purchased by the soldiers. We do not know if Rome allowed her troops to send money “home” to care for families far from the front. This is common practice in military units in the modern era, and probably originated in Rome. Money was withheld from a soldier’s pay for uniforms, shoes, socks, and other items (Herz 2007: 311). Third century documents show that up to a third of a soldier’s pay could be withheld by the unit to cover expenses. We also do not know many specifics about how soldiers were paid during the years of the later Empire. There is evidence to suggest that when soldiers retired, they were sometimes offered land – albeit a “homestead” – in territories where they had served as compensation, or partial compensation, for their years of service. Roman soldiers could retire after twenty to twenty-five years of service, and land could substitute for their state-sponsored pensions (See Southern 2006: 166ff.). In this way, “Romanitas” was exported with the army to the far reaches of the Empire. Soldiers knew what it meant to be a Roman, and with citizenship and land, the fringes of the Empire slowly changed and evolved into civilized Roman neighborhoods (Wesch-Kein 2007: 439ff; and Southern 2006: 77). This was true from Hadrian’s Wall in northern Britain, to the edges of the Sahara in North Africa, and to the Arabian desert in the Near East (Whately 2013: 236–237). In the early second century the Emperor Trajan built a road which paralleled the eastern frontiers of his empire. It was intended to facilitate the rapid movement of military forces when and if they would be needed anywhere, anytime. he intent of this highway, the Via nova Traiana, also marked a leap forward in the expansion of trade and commerce. It was along this road that imperial forces were stationed in garrisons of varying sizes to enhance commercial enterprises and maintain peace on the frontier. These forces could be moved rapidly along the new road as needed in defense of the realm. The fortresses – the larger castra and smaller castella – were manned by “professional” soldiers. The infantry from the castra and the cavalry units in the region provided a mobile defense when needed (Luttwak 1976: 177–79). Most important at el-Lejjūn, however, was the ʿain Lejjūn, the all important source of water without which life in that sector of the frontier would have been impossible

228

JOHN W. BETLYON

(Cf. Nicasie 1998: 177). The castrum would not have existed without the water of the spring. The deployment of these soldiers to guard such critical infrastructure provided a means by which coinage was disseminated throughout the far reaches of the Empire. Mint studies of coins from sites along the Limes Arabicus have shown that nearly 60% of the coins found singly or in hoards came from the major eastern mints of Antioch and Alexandria. However, other mints from across the Empire were also well represented. In the remains of the fourth/fifth century Golan Heights village of Qatsrin, for example, a vast hoard of over 9,000 coins was recovered with nearly all dating to ca. 310–363 CE. The exception was a zinc coin of Alexander Jannaeus dated to the “year 25”, which corresponded to 80–79 BCE (Cf. Meshorer 2001: 210; and Betlyon and Killebrew 2016: 38 n. 6). This small coin remained in circulation because it was similar in size to the Roman nummi and it undoubtedly circulated at equal value with them. Such small coins were accepted as currency by the local populace, which had high confidence in the market value of their money (Kindler 2003–2006: 64–65; Bijovsky 2000–2002: 202). Far and away the vast majority of the coins found in these eastern sites are coins attributed to the “Early Byzantine” period, that is, the years of the late Empire beginning with the reign of Constantine I. At Aila, for example, nearly 66% of the known coins are from the Early Byzantine period. At the Roman castrum of elLejjūn, that number is 70%. At other Arabian sites, including the so-called “Scroll’s Church” at Petra and the nearby domestic complex at ez-Zantur, Early Byzantine coins amount to 56% and 86% of the total finds respectively. 2 While we might be expecting that coin finds from the Egyptian Red Sea ports of Berenike and Myos Hormos (Quseir al-Qadim) would be similar, we do see some differences, with many more Roman Imperial and Provincial coins and vastly fewer Early Byzantine coins, if any. No Early Byzantine coins were found at Quseir al-Qadim, as the site was abandoned by that time. This Red Sea port reached its zenith in earlier years of Roman expansion, and fourth century trade routes were further 2

I will discuss the data in the Roman Aqaba Project Final Report.

FOLLOWING THE ROMAN PAYMASTER

229

north and east, although Berenike to the south remained an important trading site. This was a period when the imperial mints, although scattered throughout the empire, maintained strict standards. The mints were staffed by dedicated civil servants (Sutherland 1974: 260). There were some early coins of Constantine I struck as Augustus and Licinius as Caesar – SOLI INVICTO-types – dedicated to the “invincible Sun” and struck as far away as Arles and Rome in ca. 313–318 CE. Even after the so-called “Edict of Toleration,” thoroughly pagan types continued to be struck. They were dedicated to Sol and Jupiter reflecting commonly held religious standards of the Roman Elite in the early fourth century, and the basic division between Licinius (and Jupiter worship) and Constantine (and Sol worship). New series of coins soon replaced those which had celebrated Sol, Mars, and Genius (Sutherland 1974: 260). While some coins were struck honoring the divine Constantine, depicting the helmeted head of the emperor on the obverse and a quadriga on the reverse, large numbers of coins began to appear from the eastern mints in the period 330–347 CE celebrating the Army, which Diocletian had deployed beginning in the 290’s. The GLORIA EXERCITVS (“glory of the army”) coins depicted two Roman soldiers on either side of one or two legionnaire standards. The coins were struck in the “local” mints of Syria, Egypt, and throughout the Empire. Even along the eastern frontier, some of these military types which had been struck as far away as Lyons and Aquileia have been unearthed at Qatsrin. These were military types, honoring the service of devoted, loyal soldiers in the army, and intended as military pay. Aquileia, for example, was one of Constantine’s new mints, founded to increase the supply of coinage needed to grow the imperial monetary economy. These new military types dominated the Eastern marketplace, as legions marched on the roads, moving from one castrum or castellum to another. Even at town sites inside the frontier, these military types were dominant. Fully 28% of the 9,000+ coins of the Qatsrin hoard are these GLORIA EXERCITVS types. Even though the village of Qatsrin was west of the Via nova Traiana, it was no stranger to visits by Roman military formations moving on the Via nova Traiana and the Strata Diocletiana. All this was a part of Rome’s “grand strategy” in the fourth century CE (Nicasie 1998: 176–178). Military units moving from place to place, coupled with expanding

230

JOHN W. BETLYON

commerce throughout the eastern provinces, meant that coins were “on the move.” Types such as these GLORIA EXERCITVS coins were struck – in part – to pay the deployed army’s soldiers. When soldiers retired from the active service, they sometimes stayed in the area where they had served. The end result was a continuing “Romanization” of the East (Wesch-Kein 2007: 442–46). Coins of the VRBS-ROMA type also appear at this time. The Qatsrin hoard contains some of these coins, which depict the shewolf suckling Romulus and Remus – a scene from Virgil’s Aeneid – on the reverse, with the emperor’s bust on the obverse. Examples in the Qatsrin hoard of this type were struck in Rome, Constantinople, Nicomedia, Cyzicus, Antioch, and Alexandria. The reverse types always depict the aforementioned scene, with two stars above, and the mint-mark below an exergual line. A coin from the mint of Arles, however, also has the chi-rho symbol ( ) between the stars. Several writers have argued that the chi-rho was indicative of Constantine’s devotion to his newly espoused Christian religion, citing the famous story of his vision of the cross in the heavens and subsequent victory over his enemies in the Battle at the Milvian Bridge (Carroll 2001: 181; Dunning 2003: 24). Believing that his new god had blessed him, Constantine entered Rome and had the chi-rho emblazoned on his labarum, although it is doubtful that this action had much to do with the name of Jesus Christ. Constantine issued a joint imperial edict (The Edict of Toleration) with Licinius in Milan (in 313 CE), granting certain freedoms to those who espoused Christianity and Judaism. Over the next few years, Constantine ceased obvious devotion to Sol, no longer striking his coins dedicated to SOLI INVICTO in 321. The new mint in Arles, represented by the mint-mark PCONST, was staffed by artists and production personnel from the recently closed mint of Ostia, which had been part of the VRBS ROMA itself. Veterans from Constantine’s victory may have worked in the mint, celebrating the military victories of the Emperor (Hendy 1985: 381). Some scholars have argued that this symbolism comes from Constantine’s conversion to Christianity (Dunning 2003: 7). But the symbol chi-rho ( ) itself represents a strong heavenly sign, which to some recalled worship of SOL, to others the cult of Mithras, and to others the mystical Egyptian ankh (Cf. Bruun 1966: 61). The symbol, adapted from the Roman cavalry standard, came to be associated under Constantine the Great with

FOLLOWING THE ROMAN PAYMASTER

231

the Emperor’s victory in battle – a victory which Christianity achieved in the Edict of Milan. In Siscia, Licinius used the symbol on coins struck in 320 (Bruun 1966: 440–41). The chi-rho appears on a labarum on reverse types with the inscription SPES-PVBLIC struck in Constantinople (327 CE; Bruun 1966: 572–73, nos. 19, 26). These coins also celebrated a Constantinian military victory in the battle of Chrysopolis and the decision to restore Byzantium as his eastern capital (Bruun 1966: 562). Bruun argued that “Christian symbolism has no place” on the coins of Constantine (1966: 64). This is correct, for here the symbols were not religious, but were rather a secular rallying point for Constantine’s military forces, symbolizing military formations and adorning shields, helmets, and the labarum. These symbols are analogous to military symbols used by other armies in other times, such as the ✚ cross: this “cross” was the emblem of the VI Corps of the Army of the Potomac during the mid-19th century American Civil War. The symbol had no Christian religious significance in this context: It was a designation for a large military formation. These symbols were very important to the soldiers serving in Constantine’s ranks, including the many Celts and Germans who formed the bulk of his legionnaire personnel (Finney 1997, 659–60). The name of Jesus Christ meant nothing to these “barbarians” at this time. Only later would the symbol truly be used in a Christian context on the gold and aes issues of Valentinian I, Valens, Gratian, and Procopius in 364–367 CE (Pearce 1951: 215–217). Magnentius used the chi-rho in the early 350’s at Arles (Kent 1981: 199–200, 216–217). Kent suggests that these coins may have been struck by Magnentius to recognize a “transient success against Constantius.” The use of the symbol was also, therefore, in recognition of military success on the battlefield. Christian symbolism would not be fully incorporated into the coinage until the fifth century, when the last vestiges of opposition to Christianity were disappearing. By far the largest group of coins associated with the midfourth century CE were the types with variations of the reverse ethnic FEL TEMP REPARATIO – “the return of felicitous times,” or “happy days are here again.” These coins were struck as militaryspecie on the occasion of Rome’s 1,100th anniversary – an auspicious occasion made possible by the Roman army (Betlyon 1990: 118; see also Mattingly 1933: 182–201). Some of these coins bear

232

JOHN W. BETLYON

the galley type, while others depict the scene of the Emperor leading captives out of a small hut. But most are the common coins depicting the “fallen horseman,” on which a Roman soldier is spearing his fallen enemy, presumably a Persian soldier from the East. While the eastern limitanei were threatened by marauding Beduoin or armed incursions from Persia, this coin type reassured Roman citizens that the Emperor and his Army were firmly in control – even if this was not completely true. Constantius II, Constans, and Gallus (as Caesar) struck these coins in ca. 348–355 CE using several variants of the reverse type. Indeed these were the only bronze coins struck in the Empire in the early 350’s. All of the usual mints are represented, including Arles in faraway Gaul. These bronze minimi circulated throughout the East with the formations of soldiers on the move, merchants plying the trade routes, and a population becoming more and more mobile. The economy, according to the Classical sources, was more robust than it had been a century earlier. These were good times, although the end of Diocletian’s freeze on prices of goods and services meant that inflation once again became a problem for policy makers. The use of price controls was intended to stop the inflationary spirals of the 290’s, which it did. But artificial controls on monetary transactions without confidence and hope in the ancient consumer’s mind did little to instill long-term faith in the coinage (Sutherland 1974: 253–4). Three small hoards of coins from this period were found at Petra at the villa of Ez-Zantur. These hoards contained approximately 270 coins, nearly all of which were from the fourth century CE, and over half of which were FEL TEMP-types (Peter 1996). The military presence at Udhruh (ancient Augustopolis, ca. 10 km east of Petra, where the legionary fortress has now been definitively dated to the Diocletianic era by a recently discovered inscription) certainly impacted the local economy of Petra (Kennedy and Falahat 2008). These coins were probably struck as specie for military paymasters! Coins provide important data when studying inflation, trade routes, and circulation patterns. Throughout the Roman Empire, the army played a large role in bolstering local monetary economies. Soldiers, by their very presence, helped to increase trade, and “provided employment for people who clothed and fed legionaries,” as well as bankers and shippers, and a high level of urbanization (Evans 2006: 29; and Hopkins 1980: 102).

FOLLOWING THE ROMAN PAYMASTER

233

Most of the coins found at sites such as Umm el-Jimal, elLejjūn, and Roman Aila relate to the fourth century’s military buildup under the Tetrarchy, Constantine, and his sons. The legionary castra built at Udhruh and el-Lejjūn and the transfer of an additional legion from Jerusalem to Aila were part of this strategic military expansion. But the expansion was expensive. Inflationary pressures caused the bronze coinage to lose value through the fourth and fifth centuries, as evidenced in shrinking coin flans. Gold and silver issues were used instead of vast quantities of nearly worthless bronze minimi, but few of these more valuable coins are found by archaeologists. Soldiers did not generally lose them. One small aureus of Valentinian I was discovered at el-Lejjūn, a coin struck in 364–367 CE. We can only speculate how or why a soldier lost this substantial part of his pay. Had he hidden it in his barracks’ room, only to be forgotten? Or had the soldier gone out on patrol, never to return? This particular coin is now in the Amman Museum. Over time, the bronze issues became virtually worthless monetarily because of inflation. The monetary and metric reforms of Anastasius and Justinian I in the early sixth century changed things. The hoards of bronze minimi of the fourth century suggest that bags or purses of small change still circulated, when in sufficient quantity to complete a viable commercial transaction, or when needed to pay the troops. That was the business of the military paymaster!

BIBLIOGRAPHY Betlyon, J. W. (1990). The Coins from the 1987 Season of the Limes Arabicus Project. In: S. T. Parker. Preliminary Report of the Limes Arabicus Project. BASOR Supplement, 26, pp. 89–136. Betlyon, J.W. (2006). The Coins. In: S. T. Parker, ed., The Roman Frontier in Central Jordan: Final Report on the Limes Arabicus Project, 1980–1989. Vol. 2. Washington, D.C.: Dumbarton Oaks Research Library and Collection, pp. 413–43. Betlyon, J. W. and A. Killebrew (2016). A Fourth Century CE Coin Hoard from the Qazrin Village. In: A. Killebrew, ed., Viewing Ancient Jewish Art and Archaeology: Essays in Honor of Rachel Hachlili. Supplement to the Journal for the Study of Judaism. Leiden: E. J. Brill, pp. 33–47.

234

JOHN W. BETLYON

Bijozsky, G. (2000–2002). The Currency of the Fifth Century CE in Palestine – Some Reflections in Light of the Numismatic Evidence. Israel Numismatic Journal 14, 196–210. Bruun, P. M.(1966). The Roman Imperial Coinage. Vol. 7. Constantine and Licinius, A.D. 313–337. London: Spink and Son Ltd. Carroll, J.(2001). Constantine’s Sword: The Church and the Jews. New York: Houghton Mifflin. Dunning, M. (2003). First Christian Symbols on Roman Imperial Coins. The Celator, 17/12, pp. 6–27. Evans, J. D. (2006).The Coins and the Hellenistic, Roman and Byzantine Economy of Palestine. Vol. VI, The Joint Expedition to Caesarea Maritima Excavation Reports. Boston, MA: The American Schools of Oriental Research. Finney, P. C.(1997). Labarum. In: E. Ferguson, ed., Encyclopedia of Early Christianity, vol. 2. New York and London: Garland, pp. 659–60. Hendy, M. F. (1985). Studies in the Byzantine Monetary Economy, ca. 300–1450. Cambridge: Cambridge University Press. Herz, P.(2007). Finances and the Costs of the Roman Army. In: P. Erdkamp, ed., A Companion to the Roman Army. Malden, MA: Blackwell. pp. 306–22. Hopkins, K.(1980). Taxes and Trade in the Roman Empire (200 B.C. – A.D. 400). JRS 70, 101–25. Kennedy, D. L., and H. Falahat. (2008). Castra Legionis VI Ferratae: a building inscription for the legionary fortress at Udhruh near Petra. JRA 21, 151–69. Kent, J. P. C. (1981). The Roman Imperial Coinage. Vol. 8. The Family of Constantine I, A.D. 337–364. London: Spink and Son Ltd. Kindler, A. (2003–2006). A Hoard of Quadrates (Prutot) of the Jewish War (66–70 CE) from Khirbet Zeita. Israel Numismatic Journal 15, 64–68. Luttwak, E. N. (1976). The Grand Strategy of the Roman Empire from the First Century A.D. to the Third. Baltimore: The Johns Hopkins University Press. Mattingly, H. (1933). Roman Coins. New York: MacVeagh.

FOLLOWING THE ROMAN PAYMASTER

235

Meshorer, Y. (2001). A Treasury of Jewish Coins from the Persian Period to Bar Kokhba. Jerusalem: Yael Ben-Zvi Press and Amphora Books. Nicasie, M. J. (1998). Twilight of Empire: The Roman Army from the Reign of Diocletian until the Battle of Adrianople. Amsterdam: J C. Gieben. Peter, M. (1996). Die Fundmünzen. In: A. Bignasca, et.al., Petra: Ez Zantur I. Mainz: Verlag Philipp von Zabern, pp. 91–127. Pearce, J. W. E. (1951). The Roman Imperial Coinage. Vol. 9. Valentinian I – Theodosius I. London: Spink and Son Ltd. Sarantis, A.(2013). Military Encounters and Diplomatic Affairs in the North Balkans during the Reigns of Anastasius and Justinian. In: A. Sarantis and N. Christie, eds., Late Antique Archaeology, Vol. 8, War and Warfare in Late Antiquity. Leiden: E. J. Brill, pp. 759–808. Southern, P. (2006). The Roman Army: A Social and Institutional History. Santa Barbara, CA: ABC-CLIO, Inc. Sutherland, C. H. V. (1974). Roman Coins. London: Barrie and Jenkins. Wesch-Kein, G. (2007). Recruits and Veterans. In: P. Erdkamp, ed., A Companion to the Roman Army. Malden, MA: Blackwell Publishing, pp. 435–50. Whately, C. (2013). Organization and Life in the Late Roman Military: A Bibliographic Essay. In: A. Sarantis and N. Christie, ed., Late Antique Archaeology, Vol. 8, War and Warfare in Late Antiquity. Leiden: E. J. Brill, pp. 209–38.

THE MODULAR PLANNING OF ROMAN FORTIFICATIONS IN THE NEAR EAST: PRINCIPLES AND PROCESS JOHN PETER OLESON 1

How did Roman military engineers pick or create a design for a legionary fort, auxiliary fort, castellum, or watchtower, and how did they lay out these architectural plans on the ground? Several ancient literary sources provide information and ideas, but can we see evidence for these procedures in the surviving structures? 2 Although many British scholars have applied modular analysis to Roman forts in Britain and northern Europe (Walthew 1981, 2005; Millett 1982; Connolly, Davison, Van Driel-Murray 1989; Henderson 1991; Crummy 1993; Evans 1994; Richardson 2000; Taylor 2000), these issues have seldom been addressed in the context of the Roman Near East (Butler 1909: 73; Kennedy 2004: 81). My excavation of the early second-century Roman auxiliary fort at Humayma, Roman Hauarra, and the subsequent research are almost unique in applying thorough modular analysis to the archaeological remains of a Roman military structure in this region. The It is a pleasure to submit this study to a volume in honor of Tom Parker, who took an early interest in my work on the fort at Humayma and who provided very useful advice over the years. The many references to his work in this study will make clear the importance of Tom’s research and excavation to our knowledge of Roman military architecture in the Near East. 2 On the planning and surveying of Roman forts and camps, see Johnson 1983: 36–44; Lenoir 2011: 11–29; Grillone 2012: 69–96. 1

237

238

JOHN PETER OLESON

results clearly show the application of rational multiples of the Roman foot of 0.296 m (pes monetalis, plural pedes monetales, henceforth abbreviated pm) in both the fortification walls and the interior structures. 3 The question immediately arises as to whether the same thoroughgoing approach was applied at other military construction sites in the Roman provinces of Arabia, Syria, Palestine, and Egypt. Only a sample of the evidence can be given in this context. 4

THE TRAJANIC AUXILIARY FORT AT HAUARRA

The remains of a large Trajanic Roman auxiliary fort are clearly visible at the very north edge of the habitation centre at Humayma (Fig. 1), and nearly all the early visitors to the site commented on the structure, although they did not always interpret its function correctly (Oleson 2010: 9–20, 57–60). The fort is not mentioned by the ancient literary sources that concern Hauarra, although the Notitia Dignitatum indicates the presence of Equites sagittarii indigenae (Or. 34.25; Oleson 2010: 53–55). The Medieval illustration in the manuscript of the Notitia Dignitatum shows a stylized fort (Fig. 2), while the Peutinger Table records only the place name and distance to Zadagatta (“xx” Roman miles) without any architectural embellishment (Oleson 2010: 54–55). The inscription on a late-second or early third-century altar found in the vicus was dedicated by a vexillation of the Legio III Cyrenaica, a detachment from which presumably was stationed in the fort at that time (Oleson et al. 2002: 103–5, 112–16.) It is likely that a detachment from the Legio VI Ferrata was also stationed in the fort at some point, perhaps at the time of its foundation soon after Trajan’s conquest of the Nabataean kingdom in 106 (Kennedy 1980; Freeman 1996; Oleson et al. 2002: 104, 114; Oleson 2010: 59). The rectangular plan of the structure (Fig. 3), with a gate in each of the four walls, the outside dimensions Hultsch 1882: 88–98; Campbell 2000: 90–91, 482. Some scholars, including Campbell, use the measurement 0.2957 m for the pes monetealis, but it seems to me illusory to expect that the value of the unit can be determined to the tenth of a millimetre. In any case, the rounding up does not affect the calculations in the present study. 4 This study builds upon the information supplied in Oleson 2010. 3

MODULAR PLANNING OF ROMAN FORTIFICATIONS

239

(206.32 m x 148.32 m, corresponding to 500 x 700 pm, with 0.2–0.4 % deviation), and the interior dimensions of the reservoir in its northwest corner (29.40 m x 14.20 m x 3.05 m, corresponding to 100 x 50 x 10 pm with 0.7 to 3.0 % deviation) persuaded me of its Roman origin even before excavation. Prior to excavation, the location of Hauarra on the Via Nova Traiana and the probable inclusion of the fort in the limes system for the Provincia Arabia also suggested to most twentieth-century visitors that the structure should date sometime between the second and the fourth centuries (Graf 1983: 657; 1995: 257). Excavation quickly revealed that the fort was built soon after the conquest of Nabataea in 106.

Figure 1. al-Humayma/Hauarra, aerial view of fort from south (Photo: D. Kennedy, APAAME E116049).

The rectilinear plan, uniform throughout the entire complex, is oriented 6 degrees east of true north (ca. 9 degrees east of magnetic north in 2013. 5 During the Roman imperial period the celestial pole was approximately equidistant from α Ursae Minoris (Cynosura) and β Ursae Minoris (Kochab), and the former star, the brighter of the NOAA Geomagnetism Home: http://www.ngdc.noaa.gov/geomag-web/#declination. 5

240

JOHN PETER OLESON

two, was probably used for navigation. In the fifth century Stobaeus describes α Ursae Minoris as “always visible,” although the star was at that time removed from the celestial pole by about 8 degrees (Stobaeus, Anthologium 1.40.1, line 23). It is clear that the Roman military surveyors intended to lay out the camp on a grid oriented to celestial north, and that they came very close.

Figure 2. al-Humayma/Hauarra, illustration of forts of Arabia and Palaestina in Notitia Dignitatum, including Hauarra (after Seek 1876: 72).

MODULAR PLANNING OF ROMAN FORTIFICATIONS

Figure 3. al-Humayma/Hauarra, plan of fort and excavated structures.

241

The location selected for the fort was a gentle slope 500 m northeast of and approximately 10 m above the Nabataean reservoirs marking the centre of Nabataean Hawara. The north end of the fort is at approximately the 965 m contour, the south end at approximately the 960 m contour, providing sufficient fall for both water distribution and drainage (slope ± 2.4 %). No conclusive evidence has yet been found that there were pre-existing, Nabatae-

242

JOHN PETER OLESON

an structures at the site selected for the fort, which presumably was chosen because it was slightly apart from and above the settlement centre. The location, design, and even some of the details of the Hauarra fort correspond well with the specifications provided by Pseudo-Hyginus, De munitionibus castrorum. This work, whose author and original title are unknown, is now felt to have been written by a military engineer active in the east during Trajan’s reign. 6 The author recommends both a slightly sloping site and that the main gate, the porta praetoria, face the enemy, in this case the Nabataean settlement to the south (De munitionibus castrorum 56). nam quod attinet ad soli electionem in statuenda metatione, primum locum habent quae ex campo in eminentiam leniter attolluntur, in qua positione porta decimana eminentissimo loco constituitur, ut regiones castris subiaceant; porta praetoria semper hostem spectare debet. 7

As far as the choice of ground is concerned at the beginning of a site survey, the best is that on a gentle rise above the plain. At such a location the porta decimana [rear gate] is laid out at the highest point, so that the various sections of the camp lie below it. The porta praetoria [front gate] must always face the enemy.

A Nabataean aqueduct pool and garden lay 90 m west of the southwest corner of the fort, and there is some evidence for substantial Nabataean structures south and southeast of the pool, in the area of the later civilian Roman settlement (vicus). An additional factor was the course of the Nabataean aqueduct, which passed 65 m west of the northwest corner of the fort (Oleson 2010: 99–100). A branch line was built immediately from the aqueduct to a large reservoir in this corner of the fort, from which terracotta pipes carLenoir 1979: 113–33. Grillone (2012: 14–19) feels the date remains ambiguous but probably sometime between the early second and midthird centuries. 7 Lenoir 1979: 22, 91–92. The first sentence has textual problems. For attinet ad soli electionem in statuenda metatione, Grillone (2012: 170) restores ad sollicitudinem instituendae metationis, rendering the clause “as far as concerns the careful attention to laying out the survey…” 6

MODULAR PLANNING OF ROMAN FORTIFICATIONS

243

ried water throughout the fortified area (Oleson et al. 2008: 330–31; Oleson 2010: 175–81). The road rebuilt as the Via Nova Traiana about the same time the fort was constructed paralleled the aqueduct for the last 5 km of its course, for the last two kilometres on the east side of the aqueduct. As a result, the road, too, passed close by the east wall of the fort, east of the aqueduct, providing a direct link with the main regional communication route (Graf 1995: 257; Reeves et al. 2009: 246–51; Oleson 2010: 29–30, fig. 2.7). Pseudo-Hyginus recommends this arrangement as well: praecipue observari debebit via quae lateribus castrorum supersit (“It is particularly important to ensure that a road passes by the side of the camp.”De munitionibus castrorum 57; Lenoir 1979: 22; Grillone 2012: 170). It looks as if the fort was built to supervise both passage along the Via Nova and the small Nabataean caravan stop of Hawara, although by the fourth century its function may have been more defensive in character (Wheeler 2011: 237, 253–54). In addition to its immediate strategic and practical advantages, the location allows line-of-sight vision to the small Roman fort at Quweira, on the routes to Ramm and to Aila (Parker 1986: 105–8). A group of low hills 200 m north and northeast of the fort cuts off the view of some parts of the al-Shara escarpment to the north, but placement of a sentry on the closest hill would have solved this deficiency. There are no remains of watchtowers on these hills, suggesting that this obstruction was not an important issue. The fort at Legio in Israel, built for either the Legio II Traiana or the Legio VI Ferrata early in the second century, was also located at a position better suited to water supply and road access than for its strategic value (Tepper 2007: 62). Ultimately, survey, excavation, and clearing for consolidation of the curtain wall determined that the fort had four square corner towers projecting beyond the adjacent walls, rectangular projecting towers flanking each of the four gates, and 12 rectangular, projecting intermediate towers along the curtain wall, making a total of 24 towers. Unlike the arrangement common in northern Europe during the early empire, the towers did not project inward from the wall face, and the corners of the fort were not rounded (Gregory 1996). All the towers are 5.97 m (20 pm) wide and project 1.80 m (6 pm) from the wall face. The walls of the fort are approximately 3.00–3.06 m thick (10 pm), of which 0.69 m (2.5 pm?) is a separate wall built up against the inside face of the main wall to accommo-

244

JOHN PETER OLESON

date frequent access stairs. Since no traces of the top surface of the main wall survive, there is no evidence for the thickness or configuration of the parapet wall. If the parapet, however, was the same width as the access stair, the space remaining as a walkway (W ca. 1.62 m, ca. 5.5 pm) would have been more than sufficient for men to pass each other and deploy their equipment. A platform (W 0.09 m, L 5.22 m; 3 x 18? pm) built up against the wall 18 m north of the southeast corner (Area P) provided an extra 0.20 m of working space, possibly to allow installation of a ballista. 8 The presence of a similar platform north of the east gate suggests that this was a regular feature of the fortification, although excavation is required to reveal further examples. The exterior surface of the wall was sealed with white plaster. The height of the wall and parapet can only be estimated, but 15 pm (4.44 m) for the wall and 5 pm (1.48 m) for the parapet seem reasonable. Parker estimates a slightly greater height for the walls at Lejjun (Parker et al. 1987: 347, 2006: 189). Deviation from proposed design measurements for all these structural units within the fort is small, from 0.9 to 1.4 %. The east and west gates (porta principalis sinistra and porta principalis dextra, respectively) are not located in the centre of their respective walls, but were shifted to the south: the east wall extends 119 m north of the centre point of the east gate, and 87 m south of it (402.0 and 293.9 pm, 400 and 300 pm design measurement, with 0.5 to 2.0 % deviation). The offset of the via principalis (the main east/west road) to the south was not uncommon in Roman forts and allowed for a courtyard or parade ground in front of the principia, or headquarters building (Lenoir 1979: figs. 1, 13; Johnson 1983: 27–35; Campbell 2006: 33; Grillone 2012: figs. 2, 5). The courtyard in front of the principia measured 120 x 100 pm (28.9 x 35.1 m; 1.1 to 2.3 % deviation). The well-preserved east gate was 4.20 m (14 pm?) wide between the doorjamb blocks, although the design measurement was likely 15 pm between the centre points of those blocks. It is likely that all four gates had the same design and dimensions, although the south gate now appears to be substantially wider than the others as a result of erosion of the soil and architecOleson et al. 2008: 328–29. Josephus, Jewish War 3.80 mentions the presence of artillery on the wall between the towers. 8

MODULAR PLANNING OF ROMAN FORTIFICATIONS

245

tural remains by runoff water. A GPR survey confirmed the presence of a curved earth mound outside the north gate intended to block a direct attack on the gate from the flat plain to the north, a feature probably termed titulum (Pseudo-Hyginus, De munitionibus castrorum 49–50; Oleson et al. 2003: 52–53). The titulum is visible now as only a barely perceptible rise in the ground level; no evidence survives for similar barriers outside the other gates. The widths of the via principalis and via praetoria were documented as 8 m (27 pm), curb to curb, or 8.9 m (30 pm) including sidewalks (0.1 to 0.2 % deviation) (Oleson et al. 2003: 43). Although the walls and towers were probably only 4 or 5 m high, the whitewashed fortification must have seemed to loom over the smaller, lower structures of post-conquest Hauarra. To judge from the frequent spolia of cornice blocks, column drums, bases, and capitals, and diagonally trimmed ashlar blocks found built into the fort structures, the main public structures of Nabataean Hawara did not survive the events of the Roman occupation, whether the conquest was peaceful or contested. Although the cause is still disputed, there is evidence throughout the Nabataean kingdom for widespread destruction around the time of the Roman occupation, from earthquake or military activity or both (Schmid 1997; Parker 2009: 144–45). The fort was well placed to oversee the settlement, the largest in this part of the Hisma, its water resources and agricultural products, and to scrutinize traffic along both the Via Nova Traiana and the informal, traditional paths across the desert. Just as the Nabataean pool and garden fed by the aqueduct had provided striking visual proof of the Nabataean mastery of the desert region, so also the fort advertised the new Roman masters. In modern aerial photographs of the site, the remains of the military complex still stand out as the most impressive feature (Fig. 1).

PLANNING AND EXECUTION OF THE INTERIOR BUILDINGS OF THE HAUARRA FORT

Although this was an auxiliary fort, most likely designed to accommodate a quingenary ala, a unit of approximately 500 men de-

246

JOHN PETER OLESON

tached from some legion in the province or wider region, 9 the standard structures found in a large legionary fort were provided here as well (Fig. 4). By the end of excavation in 2005, the following buildings had been identified and plans produced: principia (headquarters building, Area G) with parade ground in front (Area K), praetorium (commander’s residence, Area I), and horrea (granary, Area J). 10 A partial plan was produced as well of a barracks building that later housed a forge (Area H), a possible craft area with latrine (Area N), and a basin that may have formed part of a stable (Area Q). The principia, praetorium, and horrea all reveal clear evidence for modular planning. The remains of the barracks and latrine area were in poor condition and only partially excavated, and it has not yet been possible to document modular design from their remains.

Figure 4. al-Humayma/Hauarra, plan of central structures in fort, with indication of pedes monetales.

Parker 1986: 105. This calculation is based largely on the area within the walls, 3.1 ha. 10 The term horreum is sometimes used for “granary,” but Latin authors more often use the plural form horrea for such a building. 9

MODULAR PLANNING OF ROMAN FORTIFICATIONS

Figure 5. al-Humayma/Hauarra, plan of praetorium, with indication of pedes monetales.

247

The horrea (granary) is located immediately east of the principia (Headquarters Building) in the northeast quadrant of the fort. This structure is identified as the granary because of the following plan and construction features, which are shared with granaries at other Roman forts: heavy, buttressed walls, long narrow interior rooms roofed with stone slabs carried on cross arches; floors carefully paved with bricks or thick stone slabs; careful arrangement for

248

JOHN PETER OLESON

drainage; location at the highest and thus driest point within the walls, and the predominance in this location of fragments of storage wares (Petrikovits 1975: 82–85; Johnson 1983: 142–57; Richardson 2004; Parker 2006: 235–40). The location of the structure near the central administrative area is paralleled in most other Roman forts. Given the isolated and environmentally marginal location of Hauarra, the provision of food and other supplies for a unit of 500 men, baggage animals, and mounts would have been logistically very challenging (Richardson 2004). The items making up the standard Roman military diet have been estimated to weigh 1.4 kg per person per day, so the unit of approximately 500 men suggested by the size of the Hauarra fort would have required approximately 700 kg of foodstuffs daily (Davies 1971; 1989: 193; Johnson 1983: 195–202). Any mounts housed in the fort would have required further supplies: 3 kg of grain per day for a pack animal, 5.5 kg/day for a horse or twice this amount of high quality pasturage (Shirley 2001: 109). Camels, which may have provided the mounts for the local equites, required less fodder, and local rough grazing could have filled some of their needs. Storage of at least three months supply of food, probably totalling more than 100 metric tons, was essential to the function of the outpost. A few details at the south end of the structure require further excavation, but it looks as if the engineers laid out the main portion of the horrea as a rectangle 75 pm wide from east to west, and 50 pm long (22.2 x 14.8 m; Fig. 5). This space was then subdivided into three large rooms 25 pm wide and 50 pm long (7.5 x 14.88 m), each with a door 5 pm wide in the centre of the south wall. The walls were built straddling the survey lines, and they vary in width. The storage area was roofed with heavy slabs carried on cross arches (Oleson 2009: 541). While the horrea was crucial for subsistence, the principia was the central administrative structure in the fort, where the unit’s gods and standards were kept and the commander’s decisions passed on to the officers and men assembled in the courtyard in front, to the south (Fig. 4). As in most Roman forts, the engineers laid this structure out on the centre line of the fort, on the same orientation as the fortification walls, facing south on to the east/west via principalis and the north/south via praetoria. The benchmark (gromae locus; see below) for the original survey probably was set up on what became the centre line of the principia, on the

MODULAR PLANNING OF ROMAN FORTIFICATIONS

249

line of its south wall, with a clear view to the sites of the four gates, two-thirds of the distance from the north wall to the south entrance. The groma is the standard Roman instrument for surveying right angles. Pseudo-Hyginus specifies just this location (De munitionibus castrorum 12; Lenoir 1979: 6, 52–53; Grillone 2012: 132, 208–9): in introitu praetorii partis mediae ad viam principalem gromae locus appellatur, quod turba ibi congruat sive in dictatione metationis posito in eodem loco ferramento groma superponatur, ut portae castrorum in conspectu rigoris stellam efficiant. et professores eius artis causa supra scripta gromatici sunt cognominati.

The spot on the via principalis along the centre line of the praetorium is called the place of the groma, because the crowd gathers there. In calling out the measurement, when the groma has been set up on an iron mark placed on that spot, it should happen that the fort gates are all in sight at right angles. The practitioners of this skill are known as gromatici because of what is written above.

For the principia the engineers laid out a rectangle 100 pm wide and 150 pm long (29.4 x 44.1 m; deviation 0.5–0.7%). A crosswall placed 30 pm from the north wall defined space for four offices and the central shrine, subdivided by four party walls 20 pm apart. The parade ground area in front consequently was 100 x 120 pm in size. It may have been surrounded by a colonnade. Several surviving statue bases show that statues drew attention to the façade of the office area and commemorated important officials (Oleson 2009: 541). The praetorium, or Commander’s Residence, is the third of these central structures. The planning procedure can be reconstructed in more detail for this building, given its better preservation, greater complexity, and more complete excavation (Fig. 5). A square was laid out at the appropriate orientation, 90 pm on a side (26.6 m); an east/west line was then laid out across the square, 60 pm north of the south side (17.8 m). These long dimensions may be based on the quarter-actus (30 pm). Two further lines were then laid out north/south, 20 pm in from the east and west sides (5.9 m). These lines defined a central courtyard 60 pm long north to south, and 50 pm wide. The long rectangles framing the east and west

250

JOHN PETER OLESON

sides of the courtyard were then each divided into rooms theoretically 15 pm wide and 20 pm deep (4.4 x 5.9 m). Five long, rectangular rooms were laid out across the northern third of the structure, all 30 pm long, oriented north/south: two outside rooms 20 pm wide (possibly subdivided in length); two at the northeast and northwest corners of the courtyard, 12.5 pm wide; a grand central room 25 pm wide. The present measurements of the structure vary slightly from these ideals, depending on whether the walls were constructed with the outer or inner face on the surveyed line, or the medial line of the wall itself. The walls in the praetorium – built for the most part of rubble set in mud, with occasional use of blocks at corners and doorways – range in thickness from 0.64–0.70 m (2.16–2.36 pm), but the design width was probably 2 pm. The addition of plaster after construction and during renovation, and the gradual dissolution of the fabric after abandonment have thickened them. As built, the outside north/south dimensions of the praetorium are just over 93 pm (27.16 m), suggesting that one east/west wall was built outside the surveyed line, and the other straddling it. Similarly, the courtyard is just over 50 pm wide (14.88 m), but only 58 pm long (17.16 m). In this case, the north/south walls and one east/west wall were built outside the survey line, but the other east/west wall was built inside it. The doors range around 5 pm in width. As in the horrea, most rooms in the praetorium had flat slab roofs supported by cross arches, while some of the smaller rooms may have been roofed with poles and palm thatch covered with an impermeable roof plaster. The military surveyors – mensores or gromatici – who laid out the fortifications and interior buildings of the fort at Hawara were following a long tradition that specified the designs and procedures to be used. Sketches or even verbal descriptions may have been sufficient, given the standardized plans, and with the groma to survey right angles and the measuring chain or pole to determine distances, the team could have quickly laid out the wall lines on the ground with coloured flags and pegs – as described by Polybius (Histories, 6.26–31, 41; Sherk 1974: 546–53; Lewis 2001: 20–21, 59– 60, 120–33). Although Polybius describes an early form of the Roman marching camp used in the mid-second century BC, and the details of the design remain subject to controversy, the characteristics of

MODULAR PLANNING OF ROMAN FORTIFICATIONS

251

symmetry, proportion, logical organization, and standardization are already clear. A site from which the camp can best be overseen is selected for the commander’s tent and marked with a flag. A square 200 pm on a side is marked out with this flag at the centre. The rest of the camp is laid out around this square, with consideration of local conditions, such as the best direction from which to take water and forage. Ὅταν δὲ κατὰ τὰς πορείας ἐγγίζωσι στρατοπεδεύειν, προπορεύονται χιλίαρχος καὶ τῶν ταξιάρχων οἱ πρὸς τοῦτο τὸ μέρος ἀεὶ προχειρισθέντες, οἵτινες ἐπειδὰν συνθεάσωνται τὸν ὅλον τόπον, οὗ δεῖ στρατοπεδεύειν, ἐν αὐτῷ τούτῳ πρῶτον μὲν διέλαβον τὴν τοῦ στρατηγοῦ σκηνὴν οὗ δεήσει θεῖναι κατὰ τὸν ἄρτι λόγον… τούτων δὲ προκριθέντων διαμετροῦνται τὴν περίστασιν τῆς σκηνῆς, μετὰ δὲ ταῦτα τὴν εὐθεῖαν, ἐφ’ ἧς αἱ σκηναὶ τίθενται τῶν χιλιάρχων, ἑξῆς δὲ τὴν ταύτης παράλληλον, ἀφ’ ἧς ἄρχεται τὰ στρατόπεδα παρεμβάλλειν… γενομένων δὲ τούτων ἐν βραχεῖ χρόνῳ διὰ τὸ ῥᾳδίαν εἶναι τὴν καταμέτρησιν, ὡς ἁπάντων ὡρισμένων καὶ συνήθων ὄντων διαστημάτων, μετὰ δὲ ταῦτα σημαίαν ἔπηξαν μίαν μὲν καὶ πρώτην, ἐν ᾧ δεῖ τόπῳ τίθεσθαι τὴν τοῦ στρατηγοῦ σκηνήν, δευτέραν δὲ τὴν ἐπὶ τῆς προκριθείσης πλευρᾶς, τρίτην ἐπὶ μέσης τῆς γραμμῆς ἐφ’ ἧς οἱ χιλίαρχοι σκηνοῦσι, τετάρτην, παρ’ ἣν τίθενται τὰ στρατόπεδα. καὶ ταύτας μὲν ποιοῦσι φοινικιᾶς, τὴν δὲ τοῦ στρατηγοῦ λευκήν. τὰ δ’ ἐπὶ θάτερα ποτὲ μὲν ψιλὰ δόρατα πηγνύουσι, ποτὲ δὲ σημαίας ἐκ τῶν ἄλλων χρωμάτων. γενομένων δὲ τούτων ἑξῆς τὰς ῥύμας διεμέτρησαν καὶ δόρατα κατέπηξαν ἐφ’ ἑκάστης ῥύμης.

When the army on the march is near the place of encampment, one of the tribunes and those centurions who are specially charged with this duty go on in advance, and after surveying the whole ground on which the camp is to be laid out (στρατοπεδεύειν) first of all determine from the considerations I mentioned above where the consul’s tent (στρατηγοῦ σκηνὴν) should be placed… When they have decided on this, they measure out first the area of the tent, next the straight line along which the tents of the tribunes are erected, and next the line parallel to this, starting from which the troops form their encampment… All this is done in a very short time, as the marking out is quite an easy matter, all the distances being

252

JOHN PETER OLESON fixed and familiar. Then they plant flags, one in the spot intended for the consul’s tent, another on the side of it they have chosen for the camp, a third in the middle of the line on which the tribune’s tents will stand, and a fourth on the other parallel line along which the legionaries will encamp. These latter flags are crimson, but the consul’s is white. On the ground on the other side of the consul’s tent they plant either simple stakes or flags of other colours. After this they go on to lay out the streets and plant stakes in each street. (Polybius, Histories 6.41.1–8; translation J.P. Oleson, after Paton 1923: 362–66).

There are striking parallels between this straightforward procedure and the modern use of variously coloured flags in laying out building sites, or lawn irrigation systems. Polybius, of course, wrote 250 years before the fort was built at Hauarra, but the procedures are simple and logical, and he notes that “all the distances (are) fixed and familiar.” Josephus was equally impressed by this Roman innovation. ὅπῃ δ’ ἂν ἐμβάλωσιν εἰς ἐχθρῶν γῆν, οὐ πρὶν ἅπτονται μάχης ἢ τειχίσαι στρατόπεδον. τὸ δὲ οὐκ εἰκαῖον οὐδὲ ἀνώμαλον ἐγείρουσιν οὐδὲ πάντες ἢ ἀτάκτως διαλαβόντες, ἀλλ’ εἰ μὲν ἀνώμαλος ὢν τύχοι χῶρος, ἐξομαλίζεται· διαμετρεῖται δὲ παρεμβολὴ τετράγωνος αὐτοῖς. καὶ τεκτόνων πλῆθος ἕπεται τῶν τε πρὸς τὴν δόμησιν ἐργαλείων. Καὶ τὸ μὲν ἔνδον εἰς σκηνὰς διαλαμβάνουσιν, ἔξωθεν δ’ ὁ κύκλος τείχους ὄψιν ἐπίχει πύργοις ἐξ ἴσου διαστήματος κεκοσμημένος. ἐπὶ δὲ τῶν μεταπυργίων τούς τε ὀξυβελεῖς καὶ καταπέλτας καὶ λιθοβόλα καὶ πᾶν ἀφετήριον ὄργανον τιθέασιν, πάντα πρὸς τὰς βολὰς ἕτοιμα. πύλαι δὲ ἐνοικοδομοῦνται τέσσαρες καθ’ ἕκαστον τοῦ περιβόλου κλίμα, πρός τε εἰσόδους τῶν ὑποζυγίων εὐμαρεῖς καὶ πρὸς τὰς ἐκδρομὰς αὐτῶν, εἰ κατεπείγοι, πλατεῖαι. ῥυμοτομοῦσι δ’ εὐδιαθέτως εἴσω τὸ στρατόπεδον, καὶ μέσας μὲν τὰς τῶν ἡγεμόνων σκηνὰς τίθενται, μεσαίτατον δὲ τούτων τὸ στρατήγιον ναῷ παραπλήσιον.

Whenever the Romans invade hostile territory, they do not engage in battle before they have built a camp (στρατόπεδον). This camp is not erected at random or unevenly; they do not all work at once or in disorderly parties. If the ground is uneven, it is first levelled; a site for the camp is then measured out

MODULAR PLANNING OF ROMAN FORTIFICATIONS

253

in the form of a rectangle. For this purpose the army is accompanied by a multitude of workmen and of tools for building. The interior of the camp is divided into rows of tents. From outside, the circuit gives the appearance of a city wall and is furnished with towers at regular intervals; and in the spaces between the towers are placed rapid-fire devices, catapults, and stone-throwers, every variety of artillery engine all ready for use. There are four gates in this surrounding wall, one on each side, wide enough for pack animals to enter without difficulty and for sallies of troops in emergencies. The camp is intersected by streets symmetrically laid out. In the middle are the tents of the officers, and precisely in the centre is the headquarters of the commander (στρατήγιον), resembling a small temple. (Josephus, Jewish War 3.76–82; translation J.P. Oleson after Thackeray 1927: 26–28).

Execution seems often to have been a little more haphazard, as walls were sometimes built on top of the theoretical wall line, or to one side or the other, as seen at Hauarra. Such minor variations from the ideal may have been intentional, as the builders tweaked a standard plan to deal with practical problems involving the site or available materials. The construction teams may also have varied in competence and attention to detail. In any case, consideration of the process of design provides a better understanding of the structures in the permanent forts considered here.

THE DOCUMENTATION OF PLANNING METHODS FOR ROMAN MILITARY STRUCTURES

Do any other Roman military structures in Jordan or elsewhere in the Near East show the same thoroughgoing planning seen at Hauarra? Are there chronological or geographical patterns in certain types of planning, and are these patterns different in the Near East than elsewhere in the Roman Empire? The analysis of comparative material, unfortunately, is not simple. At Hauarra, I had the advantage of personal familiarity with the site, thorough, careful measurement of the structures, and an interest from the very start of my excavation in the possible use of the pes monetalis. The publications concerned with other military sites not infrequently include approximate or incomplete measurements, or they do not specify from which points the measurements were taken. Furthermore, it is

254

JOHN PETER OLESON

not always easy to determine from which points on the shaky line of a fallen ancient wall measurements should be taken. Combined with our ignorance of where the ancient builders calculated the lines of their walls – along the inside or outside edges or down the middle – the modern measurements can distort the evidence for ancient planning. In addition, a project begun with a regular, modular plan sometimes had to adapt to the irregular topography or restricted size of the building site. We have no information on what margin of error the ancient surveyors tolerated, but the skewed plan of the fort at Lejjun, for example, can be attributed to an initial error in orientation transmitted to the rest of the plan by continued survey from the central groma (Parker 2006: 188). Finally, we have to decide the tolerable margin of deviation when transforming metric measurements into pedes monetales. Roman engineers, of course, also make use of other modules, such as the north European pes Drusianus of 0.332 m (Hultsch 1882: 693–95; Walthew 1981; Millett 1982; Crummy 1993: 118–19), or – much less likely – some local form of cubit, as in the cubit of 51.5 cm at Bu Njem in Libya (Rebuffat 1989). At Hauarra, the “Egypto-Palestinian” or “Long Egyptian” cubit used at Petra (52.5 cm) would be a likely candidate (Dentzer-Feydy 1990: 229; Zayadine 2003: 77–79). None of these alternate modules, however, work well with the metric measurements compiled below. The pertica or decempeda, a module consisting of 10 pm, was often used in surveying(Campbell 2000: 206–7, 240– 41, 270–71, 482). Nevertheless, we should not always expect lengths based on 50 or 100 pm, for example, since the Romans often calculated long measurements in terms of the actus of 120 pm, resulting in measurements of 12 (0.10 actus), 30 (0.25 actus), 120, 180 (1.5 actus), or 240 pm (2 actus). 11 The stadium of 625 pm might also be employed; it consisted of 125 passus of 5 pm each (Campbell 2000: 206–07, 240–41, 270–71, 482). Finally, calculation of measurements in Roman feet does not always work because of the occasional reliance on proportional measurement for laying out plans: i.e. the engineers took one dimension of the structure as the definConnolly et al. 1989: 209; Henderson 1991; Walthew 2005: 271– 77. The agrimensonres often calculate important dimensions with the actus; see Campbell 2000: 206–7, 240–41, 270–71. 11

MODULAR PLANNING OF ROMAN FORTIFICATIONS

255

ing length, and set up a grid or laid out the rooms in relation to that dimension. e.g. on a grid of 1/24th or 1/30th the length of the façade (Evans 1994; Blagg 2000: 141; Walthew 2005). PseudoHyginus indicates that the exterior (?) measurements of a temporary marching camp for three legions and assorted auxiliaries should be 2400 x 1600 pm, an intentional ratio of 3 : 2 (De munitionibus castrorum 21). He also presents the dimensions of internal features in pedes monetales: e.g. via prinicpalis 60 pm wide, the praetorium 720 x 180 pm, via quintana (behind the praetorium) 40 pm wide, intervallum (open space inside the fortification wall) 60 pm wide, roads between tents 20 pm wide. Although the description of a camp for two legions and associated auxiliaries in Polybius (6.27–42) is too early and too problematic to be of much use in this context, a foot (ποῦς), most likely the pes monetalis, was the module used: exterior 2150 x 2150 pm, via principalis 100 pm wide, via praetoria 50 pm wide, intervallum 200 pm wide. 12 The late fourth century author Vegetius does not describe in detail the process for laying out a camp or fort, but he does mention the role of the surveyors, modules, and proportions (De re militari 2.7): Mensores qui in castris ad podismum demetiuntur loca, in quibus tentoria milites figant… (“Surveyors, who in the camps measure out in feet the places where the soldiers might erect their tents…”). ...pro necessitate loci uel quadrata uel rotunda uel trigona uel oblonga castra constitues, nec utilitati praeiudicat forma, tamen pulchriora creduntur quibus ultra latitudinis spatium tertia pars longitudinis additur. ita autem ab agrimensoribus podismum mensurae colligi oportet, ut ad quantitatem concludatur exercitus... agger erigitur; supra quem ad similitudinem muri et pinnae et propugnacula conponuntur. opus uero centuriones decempedis metiuntur, ne minus foderit aut errauerit alicuius ignauia... (Vegetius, De re militari 3.8; Lang 1885: 82–84.) Walbank 1957: 709–16; Miller and DeVoto 1994: 1–43; Johnson 1983: 27–29. Walbank points out that Polybius does not give the precise dimensions of the walls, but provides enough information to provide an approximate length and width of 2150 pm (1957: 715). Johnson (1983: 27) for some reason suggests 2017 pm. 12

256

JOHN PETER OLESON As the location demands, you will set up either a square, circular, triangular, or oblong camp. The form does not damage a camp’s utility, but camps are considered more attractive if their length is a third part more then their width. The surveyors (agrimensores), however, should calculate a module of measurement in feet, so that the army might be enclosed according to its size… A mound is built up, and both a stockade and towers are placed on it, so it looks like a wall. The centurions measure the work with 10–foot poles (decempedis) lest some dig less or cause mistakes through laziness.

It is interesting that Vegetius acknowledges that variations in plan are caused by the character of the site, that modules are calculated as particular numbers of feet, that the centurions kept track of the design with 10–foot measuring rods, and that errors could creep in through carelessness. Unless measurements are taken from well preserved ashlar wall faces, we can place more confidence in the longer measurements for the structures analyzed below – perhaps 30 m or more – since in this situation both the ancient and modern percentage of error is reduced. The data from the 45 better documented structures recorded in Appendix 1 and summarized in Table 1 show that all but seven involve deviations of less than 3.2 percent, – usually much less – between the proposed measurement in pedes monetales and the actual metric measurement. Since the range of deviation depends not only on errors in ancient measurement, but also on adaptation to the building site, loss of ancient wall lines through structural collapse, and errors or approximations in modern measurements, it is not possible to set a figure for maximum allowable deviation. For the equivalences proposed below, I have usually regarded an apparent range of error greater than three percent as an indication that something is amiss with the raw data or the interpretation. In any case, the overall weight of the evidence is in favour of modular design. Table 1. Proposed design dimensions and proportions for the better-documented military structures

Site name (Polybius)

Structure type marching fort

Date Mid-2C BC

Design dimensions 2150 x 2150 pm

Proportions 1:1

Deviation 0%

MODULAR PLANNING OF ROMAN FORTIFICATIONS Site name (PseudoHyginus) Qreiyé Nedwiyât al Qdeyr SouriyahSura Soukneh

Structure type marching fort marching camp marching camp? fort

Date

257

Early 2C?

Design dimensions 2400 x 1600 pm

Proportions 3:2

Deviation 0%

mid-2C

750 x 750 pm

1:1

0.2 %

Early 2C

400 x 267 pm

3:2

1.2 to 1.3 %

Late 3C

1:1

2.1 %

37:27

0.1 to 0.4 %

auxiliary fort? castellum castellum fort castellum

Early 2C

480 x 480 pm, 4 x 4 actus 370 x 270 pm

2nd half 2C 293–305 late 1C Tetrarchic

300 x 300 pm 520 x 340 pm? 500 x 300 pm 150 x 150 pm

1:1 26:17 5:3 1:1

1.0 to 1.3 % 1.3 to 1.6 % 0.9 to 1.4 % 3.1 %

castellum

Tetrarchic

200 x 200 pm

1:1

2.1 %

castellum

3–4C?

170 x 170 pm?

1:1

0.6 %

fort fort

mid-3C Early 2C

1:1 5:4

2.0 % 0.9–1.4 %

ad-Diyatheh

fort

Ca. 300

4:3

0.9–3 %

Deir el-Kahf Deir el-Qinn

fort castellum

Ca. 300 2–4C

1:1 4:3

1.4–4.8 % 2.8 -5.1 %

Khirbet esSamra al-Hadid

castellum

Tetrarchic

300 x 300 pm 1500 x 1200 pm, 12.5 x 10 actus 240 x 180 pm, 2 x 1.5 actus 200 x 200 pm 240 x 180 pm, 2 x 1.5 actus 200 x 200 pm

1:1

2.1 to 6.4 %

fort?

0.4 to 1.1 %

marching camp castellum

540 x 360 pm; 4.5 x 3 actus 625 x 416.6 pm; 1 x 2/3 stadium 250 x 250 pm

3:2

Qasr elAzraq Qasr elAzraq Qasr elAzraq (alternate) Umm elQuttein Qasr alHallabat Azamia

Late Roman? 2C?

3:2 1:1

0.05 to 1.8 % 2.7 %

Qdeym Bkhara Kharbaqa Khan atTrâb Khan atTrâb Khan as Sâma Sa‘neh Bostra

Azamia (alternate)

4C

castellum

4C

240 x 240 pm; 2 x 2 actus

1:1

1.3 to 7 %

auxiliary fort castellum

2–4 C

500 x 400 pm

5:4

1.4 to 5.4 %

Tetrarchic

150 x 150 (?) pm

1:1

2.3 %

marching camp? marching camp?

Ea. 2C

360 x 240 pm; 3 x 2 actus 375 x 250 pm

3:2

3.1 to 5.6 %

3:2

0.9 to 1.4 %

Ea. 2C

258 Site name Umm erResas Umm elWalid Khirbet ezZona Qasr Bshir Qasr ethThuraiya Qasr ethThuraiya (alternate) Muhattet elHaj Lejjun Khirbet elMureigha Rujm elFaridiyyeh Qasr el-Bint at-Tuwana Jurf edDarawish Udruh Daʿjaniya Khirbet elAil es-Sadaqa Humayma Gharandal Yotveta Mampsis Be‘er Sheva Luxor Nag alHagar

JOHN PETER OLESON Structure type fort?

Design dimensions 540 x 480 pm; 4.5 x 4 actus 135 x 135 pm

Proportions 9:8

1.1 to 2.2 %

1:1

0.6 %

150 x 135 pm

10:9

0.1 to 0.9 %

castellum castellum

Late 3–ea. 4C 293–305 3C

190 x 190 pm 125 x 115 pm

1:1 25:23

0.1 to 1.4 % 1.4 %

castellum

3C

140 x 135 pm

28:27

0.6 to 0.9 %

castellum

Late Roman

170 x 160 pm

17:16

0.8 to 1.6 %

legionary fort fort?

ca. 300

800 x 650 pm

16:13

0.2 to 0.5 %

300?

6:5

1.4 to 1.5 %

castellum

300–500

7:6

1.3 to 1.4 %

watch tower Roman fort? castellum

Tetrarchic?

360 x 300 pm; 3 x 2.5 actus 140 x 120 pm; 1 1/6 x 1 actus 40 x 30 pm

4:3

1.3 %

2–4C?

400 x 275 pm

16:11

1.4 to 1.7 %

200–300

1:1

1.3%

legionary fort castellum fort?

late 3C

120 x 120 pm; 1 x 1 actus 800 x 600 pm

4:3

0.3 to 4.7 %

4C Roman?

340 x 340 pm 225 x 200 pm

1:1 9:8

0.4 to 0.9 % 0.9 to 1.3 %

auxiliary fort auxiliary fort castellum small fort fort auxiliary (?) fort legionary fort fort

2C

400 x 275 pm

16:11

1.4 %

Ea. 2C

700 x 500 pm

7:5

0.2 to 0.4 %

293–305 293–306 Late Roman? 3C

125 x 125 pm 130 x 130 pm 200 x 135 pm 600 x 400 pm

1:1 1:1 40:27 3:2

0% 2.4 to 3.2 % 0.1 to 1.4 % 0.8 to 1.4 %

Tetrarchic

900 x 700 pm

9:7

0.1 to 0.6 %

Tetrarchic

480 x 480 pm; 4 x 4 actus

1:1

0.01 %

castellum? castellum?

Date Late 3–ea. 4C 1–4C?

Deviation

MODULAR PLANNING OF ROMAN FORTIFICATIONS

CONCLUSIONS

259

The data from the better documented structures in the gazetteer are summarized in Table 1, for the sake of easy reference. Structures in the Gazetteer that are poorly dated or apparently only measured approximately have been omitted. Forty-five separate structures are listed, along with two literary sources and an additional alternate interpretation of the original modules for each of three structures. Of the 50 entries, all but seven involve deviations of less than 3.2 %, usually much less. The statistics strongly suggest that the Roman mensores or gromatici did in fact use various multiples of the pes monetalis in laying out at least the fortification walls of their camps and forts, in the following modules or their subdivisions: 625 pm (stadium), 120 pm (actus), 100 pm, 50 pm, 25 pm, 10 pm (pertica or decempeda), and 5 pm. The plans often were laid our according to specific proportions of length to width. The proportion of 1:1 is most common, with 20 examples, while that of 3:2 (as specified by Pseudo-Hyginus) is second most common, with seven examples. Proportions also appearing in multiple examples are 4:3 (4 examples), 5:4, 9:8, and 16:11 (2 examples each). With more careful measurement of the military structures, it is likely that the statistics would become even more convincing. It is also likely that careful measurement of structures within the fortification walls will in many cases show some of the same type of extensive modular planning as seen in the fort at Hauarra. One result of this study is confirmation that Polybius and Pseudo-Hyginus have provided us with statistics about camp design that are generally verifiable, and that as a result their comments on the practical methods of surveying and laying out these structures are likely to be generally accurate as well. As Lenoir notes, it is unlikely that the mensores closely reproduced standard plans published in handbooks (commentarii), but rather used standard plans or general principles or practices as the starting point, to be adapted to local materials, conditions, and topography (2011: 384–85). Although the results of this study seem to be self-evident– indicating that the notoriously well-organized Roman military engineers did in fact, as contemporary literary sources state, work with modules and proportions based on the pes monetalis–the documentation had not been assembled previously for the Roman Near East. As a result, archaeologists have generally ignored the potential of this type of analysis in understanding the structures they excavate. I

260

JOHN PETER OLESON

hope that this study will lead scholars and archaeologists working on Roman military structures to pay more attention to the possible relevance of Roman modules in the planning of the structures with which they are concerned.

APPENDIX 1: GAZETTEER OF ROMAN MILITARY STRUCTURES IN THE NEAR EAST FOR WHICH USE OF THE P ES M ONETALIS IS LIKELY

The approach to analysis of modules taken at Hauarra has seldom been applied to the other Roman forts excavated in the region, so it is useful to collect statistics to see if this approach to analysis of dimensions has promise. This selection is for the most part based on metric dimensions of fortifications published in regional or topical surveys (Parker 1986, 2000; Kennedy 2004; Kennedy and Bewley 2004; Lenoir 2011; Bishop 2012). The two surviving literary sources that discuss the layout of military camps are placed first. The list of archaeological sites is arranged more or less geographically, by region, and the source of the data is specified. The modern name is used, with the ancient name (where known) is provided in parentheses. Unless otherwise indicated, the sites are in the territory of modern Jordan. I have not done a systematic survey of Roman military sites in Egypt, so only two are included here. I have used the published metric measurements to calculate actual measurements in pedes monetales (taken as 0.296 m), and used these in turn to propose the theoretical design dimensions. The deviation between the theoretical and actual measurement is then indicated in parentheses as a percentage of the theoretical unit. 13 Structures whose chronology is doubtful, and those for which the modern measurements appear to be approximate, have in general been Reid 2005: 173–77 makes a diligent and successful attempt to verify the use of the pes monetalis in the Roman temple at Petra (included in the catalogue below), but she simply uses the difference between the theoretical measurement and the actual measurement as an indication of accuracy. This approach distorts analysis, since a larger difference in a longer measurement can represent a lower percentage deviation than a smaller difference on a shorter measurement. 13

MODULAR PLANNING OF ROMAN FORTIFICATIONS

261

omitted. Deviation from the proposed design lengths may be the result of inaccuracy in the ancient or modern survey, or discrepancy between the intended ancient layout line (inside or outside the wall, or along the middle) and the points at which the modern measurements were taken, or both. For the most part, the modern measurements seem to have been taken from the outside faces of the curtain walls, a choice that often appears to reflect the intentions of the original Roman surveyors. Occasionally, measurements taken from the outside of projecting towers appear to provide a better solution. Modern measurements taken from the edge of the spill of a fallen wall, or from a point approximately over the original wall face, usually will be inaccurate. Table 1 summarizes the proposed design dimensions of the better documented structures listed below, the proportions of length to width, and the percentages of deviation from the proposed measurement in pedes monetales.

Polybius, Histories 6.27–42: legionary marching fort (?), mid-2C BC. 2150 x 2150 pm.

Pseudo-Hyginus, De munitionibus castrorum 21: legionary marching fort, early 2C? 2400 x 1600 pm, an intentional ratio of 3 : 2.

Qreiyé,Syria (Lenoir 2011: 64): marching camp, mid-2C. 220 x 220 m = 743.2 x 743.2 pm = 750 x 750 pm (0.2 % deviation).

Nedwiyât al Qdeyr,Syria (Lenoir 2011: 66): marching camp? Early 2C. 120 x 80 m = 405.4 x 270.3 pm = 400 x 267 pm (ratio of 3 : 2; 1.2 to 1.3 % deviation).

Souriyah-Sura,Syria (Bishop 2012: 125): legionary fort, Tetrarchic. 145 x 145 m = 489.9 x 489.9 pm = 480 x 480 pm (4 x 4 actus; 2.1 % deviation). Soukneh,Syria (Lenoir 2011: 68): auxiliary fort? Early 2C. 110 x 80 m = 371.6 x 270.3 pm = 370 x 270 pm (0.1 to 0.4 % deviation).

Qdeym (Acadama), Syria (Lenoir 2011: 72): castellum, second half 2C. 87.9 x 87.6 m = 297.0 x 296.0 pm = 300 x 300 pm (1.0 to 1.3 % deviation).

Bkhara (Avatha), Syria (Lenoir 2011: 81–82): castellum, 293–305. 152 x 99 m = 513.5 x 334.5 pm = 520 x 340 pm? (1.3 to 1.6 % deviation).

262

JOHN PETER OLESON

Kharbaqa,Syria (Lenoir 2011: 88–89): fort, late 1C. 150 x 85 m = 506.8 x 287.2 pm = 500 x 300 pm (0.9 to 1.4 % deviation). Khan at-Trâb (Vallis Diocletiana), Syria (Lenoir 2011: 95): castellum, Tetrarchic? 43 x 43 m without corner towers = 145.3 x 145.3 pm = (3.1 % deviation). 58 x 58 m with towers = 195.9 x 195.9 pm = 200 x 200 pm (2.1 % deviation). 150 x 150 pm?

Khan as Sâma/Khan abu-Shamat (Thama?) Syria (Lenoir 2011: 96): castellum, 3–4C? 50 x 50 m = 168.9 x 168.9 pm = 170 x 170 pm? (0.6 % deviation). Saʿneh, Syria (Lenoir 2011: 101–2): fort, mid-3C. 87 x 87 m = 293.9 x 293.9 pm = 300 x 300 pm (2.0 % deviation).

Bostra, Syria (Kennedy 2004: 218): legionary fort, early 2C. 440 x 360 m = 1486.5 x 1216.2 pm = 1500 x 1200 pm (0.9 to 1.4 % deviation). 463 x 363 m = 1564.2 x 1226.4 pm = 1500 x 1200 pm (2.2 to 4.3 % deviation). Alternatively: 12.5 x 10 actus (2.2 to 4.3 % deviation).

ad-Diyatheh, Syria (Parker 2000: 124): fort, ca. 300. 71.7 x 51.7 m with the corner towers = 242.2 x 174.7 pm = 250 x 175 pm (0.03 to 3.1 % deviation). Alternatively: 240 x 180 (2 x 1.5 actus; 0.9 to 2.9 % deviation). Without the corner towers (Lenoir 2011, 99), 69.5 x 49.5 m = 234.8 x 167.2 pm = 240 x 170 pm (1.6 to 2.2 % deviation). The 2 x 1.5 actus estimate seems the best solution.

Deir el-Kahf (Speluncae) (Parker 1986: 22): fort, after 306. 62 x 61 x 60 m x 61 = 209.5 x 206.1 x 202.7 x 206.1 pm = 200 x 200 pm (1.4 to 4.7 % deviation). Deir el-Qinn (Kennedy 2004: 79): castellum, 1–4 C. 73 x 56 m = 246.6 x 189.2 m = 250 x 200 pm (1.4 to 5.4 %). Alternatively 240 x 180 pm (2 x 1.5 actus; 2.8 to 5.1 % deviation).

Umm el-Jimal (Parker 1986: 26–27): fort, early 4C. One side is either 100 m = 337.8 pm, or 110 m = 371.6 pm. Kennedy (Kennedy 2004: 86) gives the dimensions of ca. 95–112 m = 320.9 to 378.4 pm = 320 to 380 pm (0.3 to 0.4 % deviation). Lenoir (2011: 104) reports 116 m, 97 m, and 105 m, or 391.9, 327.7, and 354.7 pm. There is too little certainty in the measurements to allow confident application of the pes monetalis.

MODULAR PLANNING OF ROMAN FORTIFICATIONS

263

Khirbet es-Samra (Hatita) (Lenoir 2011: 107): castellum, Tetrarchic. 63 x 58 m = 212.8 x 195.9 pm = 200 x 200 pm (2.1 to 6.4 % deviation). The square plan was skewed in execution, explaining the large deviation. al-Hadid (Aditha) (Parker 1986: 32–33): fort? Late Roman? 158 x 107 m = 533.8 x 361.5 pm = 540 x 360 pm (4.5 x 3 actus; 0.4 to 1.1 % deviation).

Qasr el-Azraq (Gregory and Kennedy 1985: 263, pl. 58b): marching camp, 2C? 185.1 x 125.5 m = 625.3 x 424.0 pm = 625 x 416.6 pm (1 stadium x 2/3 stadium; 0.05 to 1.8 % deviation). Although these measurements are approximate, having been taken from a drawing based on an aerial photograph, the dimensions are logical and convincing.

Qasr el-Azraq (Amata) (Lenoir 2011: 119): castellum, 4C, with significant later rebuilding. 76 x 72 m = 256.8 x 243.2 pm = 250 x 250 pm (2.7 % deviation), or 240 x 240 pm (2 x 2 actus; 1.3 to 7 % deviation).

Umm el-Quttein (Kennedy 2004: 81): auxiliary fort, 2–4 C. 156 x 120 m = 527.0 x 405.4 pm = 500 x 400 pm (1.4 to 5.4 % deviation).

Qasr al-Hallabat (Veriaraca) (Butler 1909: 71–73): castellum, Tetrarchic. West side 31.0 m + 5.9 +6.45 m for towers = 43.35 m = 146.5 pm = 150 pm (2.3 % deviation). The wall length plus half towers (to get ideal building block) is 31.0 +6.17 m = 37.17 m = 125.6 pm = 120 pm (1 actus; 4.7 % deviation). Khirbet ʿAin (Kennedy 2004: 109): small fort, Roman? 48 x 39 m = 162.2 x 131.8 pm = 160 x 130 pm (1.4 % deviation). There are many uncertainties about the date and dimensions of this structure.

Azamia (Kennedy and Bewley 2004: 174–75): Roman marching camp? Early 2C? Ca. 110 x 75 m = ca. 371.6 x 253.4 m = 360 x 240 pm (3 x 2 actus; 3.1 to 5.6 % deviation). The deviation is large, but it appears that the metric measurements are only approximate. Another resolution is 375 x 250 pm (proportions of 3 : 2; 0.9 to 1.4 % deviation).

Umm er-Resas (Mefaat) (Kennedy 2004: 137): fort? Late 3 to early 4C. 158 x 139 m = 533.8 x 469.6 pm = 540 x 480 pm (4.5 x 4 actus; 1.1 to 2.2 % deviation).

264

JOHN PETER OLESON

Umm el-Walid (Parker 1986: 42–43): castellum? 1–4C? Exterior 40.2 x 40.2 m = 135.8 x 135.8 pm = 135 x 135 pm (0.6 % deviation). Interior: 30.45 m = 102.9 pm = 100 pm (2.9 % deviation).

Khirbet ez-Zona (Parker 1986: 45): castellum? late 3 to early 4C. 44 x 40 m = 148.6 x 135.1 pm = 150 x 135 pm (0.1 to 0.9 % deviation). Qasr Bshir (Praetorium Mobeni) (Parker et al. 1987: 457–77, esp. 47): castellum, 293–305. 56.3 x 57.05 x 56.75 x 55.45 m = 190.2 x 192.7 x 191.7 x 187.3 pm = 190 pm square (0.1 to 1.4 % deviation). Towers project 3.05 m = 10.3 pm = 10 pm (3.0 % deviation), giving outside dimensions of approximately 210 pm. Qasr eth-Thuraiya (Parker 1986: 50): castellum, 3C. 37.5 x 34.5 m = 126.7 x 116.6 pm = 125 x 115 pm (1.4 % deviation). Re-interpreting Domaszewsik’s plan, Lenoir (2011: 131) calculates dimensions of 41.2 x 39.6 m = 139.2 x 133.8 pm = 140 x 135 pm (0.6 to 0.9 % deviation).

Muhattet el-Haj (Lenoir 2011: 126): castellum, Late Roman. Mean wall lengths 49.5 x 47 m = 167.2 x 158.8 pm = 170 x 160 pm (0.8 to 1.6 % deviation). All four walls of this quadrilateral shape differ in length, so the original plan is not entirely clear.

Lejjun (Betthora) (Parker 2006: 115, fig. 3.4): legionary fort, ca. 300. Ca. 238 x 192 m = 804.1 x 648.6 pm = 800 x 650 pm (0.2 to 0.5 % deviation). Principia is 63 m x 52.5 m = 212.8 x 177.4 pm = 200 x 175 pm (1.4 to 6.4 % deviation). The deviation in the principia pm measurements seems too high to be convincing.

Khirbet el-Fityan (Parker et al. 1987: 431): castellum above Lejjun, 300. N wall 73 m = 246.6 pm = 250 pm (1.4 % deviation). With towers, N wall 77.6 m = 262.1 pm. The other walls partly lost. E wall at least 71 m long, S wall probably 85 m long. Wall 1.8 m thick = 6.1 pm = 6 pm (1.7 % deviation).

Khirbet el-Mureigha (Kennedy 2004: 163): Nabataean or Roman fort? 300? 105 x 90 m = 354.7 x 304.1 pm = 360 x 300 pm (2.5 x 3 actus; 1.4 to 1.5 % deviation).

Rujm el-Faridiyyeh (Kennedy 2004: 165): road station/castellum, 300–500. 42 x 36 m = 141.9 x 121.6 pm = 140 x 120 pm (1 1/6 x 1 actus) (1.3 to 1.4 % deviation). The proportions come out as 7 x 6 units of 20 pm.

MODULAR PLANNING OF ROMAN FORTIFICATIONS

265

Qasr el-Bint (Parker 1986: 91): watch tower, Tetrarchic? 12 x 9 m = 40.5 x 30.4 pm = 40 x 30 pm (1.3 % deviation).

at-Tuwana (Kennedy 2004: 167): Roman fort? 2–4C? Enclosure 120 x 80 m = 405.4 x 270.3 pm = 400 x 260 (1.4 to 4.0 % deviation), or 400 x 275 pm (1.4 to 1.7 % deviation).

Jurf ed-Darawish (Kennedy 2004: 168): castellum; 200–300. Ca. 36 x 36 m = 121.6 x 121.6 pm = 120 x 120 pm = 1 actus square (1.3% deviation).

Udruh (Adrou) (Parker 2000: 132; Kennedy 2004, 178): legionary fort, late 3C. Very close in size to Lejjun, but the plan is distorted. N wall 246 x E wall 207 x S wall 248 x W wall 177 m = 831.1 x 699.3 x 837.8 x 598.0 pm = 800 x 600 pm as an original plan (0.3 to 4.7 % deviation, excluding the E wall). Petra (Reid 2005: 173–77): Roman temple, 2C. 14.6 x 14.6 m = 49.3 x 49.3 pm = 50 x 50 pm (1.4 % deviation). Although this is not a Roman military structure, it is included here because of Reid’s careful attempt at analysis of the Roman modules.

Daʿjaniya (Parker 1986: fig. 41): castellum, 4C. 101.5 x 100.2 m = 342.9 x 338.5 pm = 340 x 340 pm (0.4 to 0.9 % deviation). Khirbet el-Ail (Kennedy 2004: 180): fort? Roman? Ca. 60 x 66 m = 223.0 x 202.7 pm = 225 x 200 pm (0.9 to 1.3 % deviation).

es-Sadaqa (Zadagatta) (Kennedy 2004: 187): auxiliary fort, 2C. Ca. 120 x 80 m = 405.4 x 270.3 pm = 400 x 275 pm (1.4 to 1.7 % deviation). Tel Abara (Kennedy 2004: 180): marching camp? Trajanic? Ca. 150 x 120 m = 506.8 x 405.4 pm = 500 x 400 pm (1.4 % deviation).

Khirbat Abu Safat (Kennedy and Bewley 2004: 175–76): Roman marching camp? 2–3C? 200 x 130 m = 675.7 x 439.2 pm. The metric measurements appear to be approximate, resulting in ambiguous results in pedes monetales.

al-Humayma (Hauarra) (Oleson 1995: 321): auxiliary fort, Trajanic. 206.3 x 148.3 m = 697.0 x 501.0 pm = 700 x 500 pm (0.2 to 0.4 % deviation).

el-Quweira (Kennedy 2004: 198): fort, early 2C, rebuilt 3C to 4C? 32.5 x 31.5 m = 109.8 x 106.4 pm. = 100 x 100 pm? (3.2 to 4.9 %

266

JOHN PETER OLESON

deviation) or 120 x 120 pm (8.5 to 11.3 % deviation). These dimensions do not seem to conform to reasonable numbers of pedes monetales, but all measurements of the fort appear to have been based on estimating the position of partly concealed walls.

Khirbet el-Khalde (Praesidio) (Parker 1986: 108–9): castellum, 3–4C. 54 x 33 m = 182.4 x 111.5 pm = 180 x 120 pm (1.5 x 1 actus; 1.3 to 7.1 % deviation). The margin of error is too large for the 1 actus measurement, but a width of 110 pm does not seem to correspond well to a length of 180 pm. Kennedy and Bewley (2004, 191) provide the dimensions 49.5 x 32 m = 167.2 x 108.1 pm, but these also leave the original design measurements ambiguous.

Qasr el-Kithara (Parker 1986: 110–11): early 2C. 49 x 48 x 33.5 x 31.6 m = 165.5 x 162.2 x 113.2 x 106.8 pm. Although the surveyors may have had the intention of measuring out walls at 160 pm (1 and 1/3rd actus), the very irregular outline of the site probably dictated the final dimensions. Gharandal (Ariedela) (Kennedy 2004: 210): castellum, 293–305. 37 x 37 m = 125 x 125 pm (0 % deviation), or 1 actus x 1 actus (4.2 % deviation).

Yotveta, Israel (Kennedy 2004: 221): small fort, 293–306. 39.7 x 39.4 m = 134.1 x 133.1 pm = 130 x 130 pm (2.4 to 3.2 % deviation), or 140 x 140 pm (4.2 to 4.9 % deviation). Neither equivalent is very striking, but 130 pm works well enough. Giʾvat Shaul, Israel (Gregory 1995: 2.414): small fort, early 4C. 17.5 x 15.5 m = 59.1 x 52.4 pm = 60 x 50 pm (1.5 to 4.8 % deviation). The ratio of 6 to 5 is reasonable, although the deviation of 4.8 % for the shorter wall seems excessive.

Mampsis, Israel (Kennedy 2004: 224): fort, Late Roman? 60 x 40 m = 202.7 x 135.1 pm = 200 x 135 pm (0.1 to 1.4 % deviation). A length of 140 pm would be more characteristically Roman, but would result in a deviation of 3.5 %. The ratio of 6 to 4, however, is reasonable (e.g. Beer Sheva).

Beʿer Sheva, Israel (Fabian 1995: 237): auxiliary (?) fort, 3C. 180 x 117.5 m = 608.1 x 397.0 pm =600 x 400 pm (0.8 to 1.4 % deviation).

MODULAR PLANNING OF ROMAN FORTIFICATIONS

267

Khirbet Khiraf, Israel (Hirschfeld 1991: 174): fort, 2C. 93 x 87.7 m = 314.2 x 296.3 pm = 300 x 300 pm? (1.2 to 4.7 % deviation). 300 pm is 2.5 actus, or 10 x ¼ actus. Luxor, Egypt (Bishop 2012: 123): legionary fort, Tetrarchic. 268 x 207 m = 905.4 x 699 pm = 900 x 700 pm (0.1 to 0.6 % deviation).

Nag al-Hagar, Egypt (Mackensen 2009): fort, Tetrarchic. 142 x 142 m = 479.7 x 479.7 pm = 480 x 480 pm (4 actus x 4 actus; 0.01 % deviation), or – less likely – 500 x 500 pm (4.1 % deviation).

REFERENCES

Bishop, M. C. (2012). Handbook to Roman Legionary Fortresses. Barnsley: Pen and Sword. Blagg, T.F.C. 2000. “The Architecture of the Legionary Principia.” In: R.J. Brewer, ed., Roman Fortresses and their Legions. London: Society of Antiquaries, pp. 139–47 Butler, H.C. (1909). Ancient Architecture in Syria, Section A: Southern Syria, Part 2, The Southern Hauran. Leiden: Brill. Campbell, B. (2000). The Writings of the Roman Land Surveyors: Introduction, Text, Translation, and Commentary. Journal of Roman Studies Monograph, 9. London: Society for the Promotion of Roman Studies. Campbell, D.B. (2006). Roman Legionary Fortresses 27 BC AD 378. Abingdon: Osprey. Connolly, P., and Davison, D., and C. Van Driel-Murray. (1989). The barracks of the Roman army from the 1st to 3rd centuries A.D: a comparative study of the barracks from fortresses, forts and fortlets with an analysis of building types and construction, stabling and garrisons. 3 vols. BAR-IS S472. Oxford: BAR Crummy, P. (1993). Metrological analysis of Roman fortresses and towns in Britain. In: S.J. Greep, ed., Roman Towns: the Wheeler inheritance. A Review of 50 years’ research. CBA Research Report 93. London: Council for British Archaeology, pp. 111–19. Davies, R. W. (1971). The Roman Military Diet. Britannia, 2, 122– 42. Davies, R. W. (1989). Service in the Roman Army. Edinburgh: Edinburgh University Press.

268

JOHN PETER OLESON

Dentzer-Feydy, J. (1990). Khirbet edh-Dharih: Architectural Decoration of the Temple. ARAM, 2, 229–34. Evans, E. (1994). Military Architects and Building Design in Roman Britain. Britannia, 25, 143–64. Fabian, P. (1995). The Late Roman Military Camp at Beer Sheba: A New Discovery. In: J.H. Humphrey, ed., The Roman and Byzantine Near East: Some Recent Archaeological Research. JRA Suppl. 14. Ann Arbor: JRA, pp. 235–240. Freeman P. (1996). The Annexation of Arabia and imperial Grand Strategy. In: D. L. Kennedy, ed., The Roman Army in the East. JRA Supp. 18. Ann Arbor: JRA, pp 91–118. Graf, D. (1983). The Nabataeans and the Hisma: In the Steps of Glueck and Beyond. In: Carol L. Meyers and M. O’Connor, eds., The Word of the Lord Shall Go Forth: Essays in Honor of David Noel Freedman. Winona Lake: Eisenbrauns, pp. 647–64 Graf, D. (1995). The Via Nova Traiana in Arabia Petraea. In: J. H. Humphrey, ed., The Roman and Byzantine Near East: Some Recent Archaeological Research. JRA Suppl. 14. Ann Arbor: JRA, pp. 241–67 Gregory, S. (1995). Roman Military Architecture on the Eastern Frontier, A.D. 200–600, 3 vols. Amsterdam: Hakkert. Gregory, S. (1996). Was There an Eastern Origin for the Design of Late Roman fortifications? Some Problems for Research on Forts of Rome’s Eastern Frontier. In: D. L. Kennedy, ed., The Roman Army in the East. Journal of Roman Archaeology, Suppl. 18. Ann Arbor: JRA, pp. 169–210 Henderson, C. G. (1991). Aspects of the planning of the Neronian fortress of legio II Augusta at Exeter. In V. A. Maxfield, and M. J. Dobson, eds., Roman Frontier Studies, 1989. Proceedings of the XVth International Congress of Roman Frontier Studies. Exeter: University of Exeter Press, pp. 73–83. Hirschfeld, Y. (1991). Khirbet Khiraf: A 2nd-Century Fort in the Jordan Valley. JRA 4, 170–83. Hultsch, F. (1882). Griechische und römische Metrologie. Berlin: Weidmann.

MODULAR PLANNING OF ROMAN FORTIFICATIONS

269

Johnson, A. (1983). Roman Forts of the 1st and 2nd Centuries AD in Britain and the German Provinces. London: Adam & Charles Black. Josephus, Jewish War. Translated by H. Thackeray. (1927). The Jewish War, Books III–IV. Cambridge MA: Harvard University Press. Kennedy, D. L. (1980). Legio VI Ferrata: The Annexation and Early Garrison of Arabia. Harvard Studies in Classical Philology 84, 282–309. Kennedy, D. L. (2004). The Roman Army in Jordan. 2nd ed. London: Council for British Research in the Levant. Kennedy, D. L., and R. Bewley. (2004). Ancient Jordan from the Air. London: Oxbow. Lenoir, M. (1979). Pesudo-Hygin, Des fortifications du camp. Texte établi, trduit et commenté. Paris: Les Belles Lettres. Lenoir, M. (2011). Le Camp Romain: Proche-Orient et Afrique du Nord. Rome: École Française de Rome. Lewis, M. J. T. (2001). Surveying Instruments of Greece and Rome. Cambridge: Cambridge University Press. Mackensen, M. (2009). The Tetrarchic fort at Nag al-Hagar in the province of Thebaïs: preliminary report (2005–8). JRA 22, 287–311. Miller, M. C. J., and J. G. Devoto (1994). Polybius and Pseudo-Hyginus: The Fortification of the Roman Camp. Chicago: Ares. Millett, M. (1982). Distinguishing between the Pes Monetalis and the Pes Drusianus: Some problems. Britannia, 13, 315–20. Oleson, J.P. (2009). Trajan’s Engineers and the Roman Fort at Humayma (ancient Hauarra, Jordan). SHAJ 10, 535–48. Oleson, J.P. (2010). The Humayma Excavation Project, vol. 1: Resources, History, and the Water-Supply System. Boston: American Schools of Oriental Research. Oleson, J.P., ʿAmr, K., Schick, R., and R. Foote. (1995). Preliminary Report of the Humeima Excavation Project, 1993. ADAJ 39, 317–54. Oleson, J.P., Baker, G., de Bruijn, E., Foote, R., Logan, J., Reeves, M.B., and A.N. Sherwood. (2003). Preliminary Report of the

270

JOHN PETER OLESON

Al-Humayma Excavation Project, 2000, 2002. ADAJ 47, 37– 64. Oleson, J.P., Reeves, M.B., and B. Fisher. (2002). New Dedicatory Inscriptions from Humayma (Ancient Hawara), Jordan. ZPE 140, 103–21. Oleson, J.P., Reeves, M. B., Baker, G., de Bruijn, E., Gerber, Y., Nikolic, M., and A.N. Sherwood. (2008). Preliminary Report on Excavations at al-Humayma, Ancient Hawara, 2004 and 2005. ADAJ 52, 309–42. Parker, S.T. (1986.) Romans and Saracens: A History of the Arabian Frontier. Winona Lake, IN: Eisenbrauns. Parker, S.T., et al. (1987). The Roman Frontier in Central Jordan. Interim Report on the Limes Arabicus Project, 1980–1985. BAR-IS 340. Oxford: BAR. Parker, S. T. (2000). Roman Legionary Fortresses in the East. In: R. J. Brewer, ed., Roman Fortresses and their Legions: Papers in honour of George C. Boon, London: Society of Antiquaries, pp. 121–138. Parker, S. T. (2006). The Roman Frontier in Central Jordan: Final Report on the Limes Arabicus Project, 1980–89. 2 vols. Washington D.C.: Dumbarton Oaks. Parker, S.T. (2009). The Roman frontier in southern Arabia: a synthesis of recent research. In: W.S. Hanson, ed., The Army and Frontiers of Rome. JRA Suppl. 74. Portsmouth R.I.: JRA, pp.143–152. Petrikovits, H. von. (1975). Die Innenbauten römischer Legionslager wahrend der Prinzipatzeit. Opladen: Westdeutscher. Polybius, Histories. Translated by W. Paton. (1923). The Histories, volume III. Cambridge MA: Harvard University Press. Pseudo-Hyginus, De munitionibus castrorum. See Lenoir 1979. Rebuffat, R. (1989). Notes sur le Camp Romain de Gholaia (Bu Njem). Libyan Studies 20, 155–67. Reeves, B. et al. (2009). Preliminary Report on Excavations in the Nabataean Town and Roman Vicus at Humayma (Ancient Hawara). ADAJ 53, 229–63. Reid, S.K. (2005). The small temple: an imperial cult building in Petra, Jordan. Piscataway NJ: Gorgias Press.

MODULAR PLANNING OF ROMAN FORTIFICATIONS

271

Richardson, A. (2000). The Numerical Basis of Roman Camps. Oxford Journal of Archaeology 19,425–37. Richardson, A. (2004). Granaries and Garrisons in Roman Forts. Oxford Journal of Archaeology 23, 429–42. Schmid, S. (1997). Nabataean Fine Ware Pottery and the Destruction of Petra in the Late First and Early Second Century AD. SHAJ 6, 413–20. Seeck, O. (ed.). (1876), Notitia dignitatum. Accedunt notitia urbis Constantinopolitanae et latercula provinciarum. Berlin: Weidmann. Sherk, R. K. (1974). Roman Geographical Exploration and Military Maps. ANRW 2.1, 534–62. Shirley, E.A.M. (2001). Building a Roman Legionary Fortress. Brimscombe Port: Tempus. Stobaeus, Anthologium. Edited by A. Grillone. (2012). Gromatica militare: lo ps. Igino. Prefazione, testo, traduzione e commento. Collection Latomus, 339. Brussels: Éditions Latomus. Taylor, D. J. A. (2000). The Forts on Hadrian’s Wall. A Comparative analysis of the form and construction of some buildings. BAR British Series 305. Oxford: BAR. Tepper, Y. (2007). The Roman Legionary Camp at Legio, Israel. In: A.S. Lewin andP Pellegrini, eds., The Late Roman Army in the Near East from Diocletian to the Arab Conquest. BAR-IS 1717. Oxford: Archaeopress, pp. 57–71. Vegetius, De re military. Edited by C. Lang. (1885). Flavi Vegeti Renati, Epitoma Rei Militaris. Leipzig: Teubner. Walbank, F.W. (1957). A Historical Commentary on Polybius, vol. 1. Oxford: Clarendon Press. Walthew, C.V. (1981). Possible Standard Units of measurement in Roman Military Planning. Britannia 12, 15–35. Walthew, C.V. (2005). Modular Planning in First-Century A.D. Romano-British Auxiliary Forts. Britannia 36, 271–310. Wheeler, E.L. (2011). The Army and the Limes in the East. In: P. Erdkamp, ed., A Companion to the Roman Army. Chichester: Wiley-Blackwell., pp. 235–66.

272

JOHN PETER OLESON

Zayadine, F., F. Larché, and J. Dentzer-Feydy. (2003). Qasr al-Bint de Pétra. L’Architecture, le Décor, la Chronologie et les Dieux. Paris: Éditions Recherches sur les Civilisations.

THE MILITARY PRESENCE IN PETRA DURING THE ROMAN-BYZANTINE PERIOD ZBIGNIEW T. FIEMA

Romans and Saracens (1986) by S. Thomas Parker, remains one of the most influential works on the military aspects of the RomanByzantine East, including the archaeology and history of the military infrastructure in Transjordan. Perceiving a military threat from nomadic populations within and without the eastern provinces, and concentrating on the defensive measures as best manifested by the limes Arabicus, Parker’s work has significantly contributed to the ongoing debate on the nature of the Roman/Byzantine military presence in the East. It is, however, necessary to note that an alternative interpretation shifts emphasis from a pure defensive role to one underlining the needs of internal security and the functioning of local administration and economy within the provinces. Such studies emphasize the role of the Roman army as an occupation force and an instrument of imperial policy in controlling local economy and the sociopolitical affairs of indigenous populations living under the Roman rule (e.g., Graf 1989; Isaac 1990). The role of the Roman army as an enforcer of peace in large population centers of the East features prominently in this perspective. Petra in southern Jordan, was the royal city of the Nabataeans (Flavius Josephus BJ 1, 125 [1, 6, 2],) and a major commercial center in the Near East. The city flourished with the incense trade in which the Nabataeans actively participated. Admittedly, some scholars (e.g., Sartre 1985: 54–56, Bowersock 1983: 21, 64) have suggested that Petra experienced political and economic decline in the 1st century CE, triggered by the redirection of the incense trade to the Egyptian harbors and the subsequent collapse of the Naba273

274

ZBIGNIEW T. FIEMA

taean trading network. In this interpretation, a direct consequence of the eclipse of Petra was then elevation of Bostra, ascribed to the last Nabataean king, Rabbel II (70/71–106 CE), who is thought to have transferred the capital from Petra to Bostra. However, neither the transfer nor other assumptions listed above can be verified and these are, in fact, belied by historical and archaeological data related to Petra during the late Nabataean and the Early Roman period (Fiema 1988, 2003). The city remained the most important center of the Nabataean Kingdom, and the wealth of material culture at the site indicates the outstanding prosperity of the city in the 1st century CE and for at least a century after the annexation of the Nabataean Kingdom by Trajan in 106 CE. The main corpus of archaeological and epigraphic data, and historical information concerning the Roman military presence in Transjordan is contained in three indispensable works: Speidel (1977), Parker (1986), and Kennedy (2004). Additionally, the military infrastructure (forts, fortlets, watchtowers) and the epigraphy, specifically located in southern Transjordan, have also been presented and discussed (Fiema 1995). Furthermore, attention has been brought to the military during the Byzantine period (4th–early 7th centuries CE), specifically in Petra's countryside (Fiema 2002a, 2002b) with an emphasis on the information provided by the 6th century Byzantine Petra Papyri (Fiema 2007). The following text concentrates on the military presence in Petra itself during the Roman-Byzantine period.

THE ROMAN PERIOD

Due to the paucity of ancient sources related to the Roman annexation of Nabataea, neither the sociopolitical situation of the Nabataean Kingdom at the end of its existence nor the nature of the annexation are fully understood and the proposed reasons for the takeover cannot be verified (Fiema 1987, Freeman 1996). Equally enigmatic is the course of the event and its immediate aftermath, for example, whether or not the actual takeover involved a military action. Destructions were reported on some sites within the former Nabataean realm, with the assumption that those they were due to the military confrontation (Schmid 1997). In Petra, the ez-Zantur I residential complex, located on a ridge in the southern part of the city, displays extensive destruction by fire, dated to the early 2nd century CE (Stucky 1996: 14, 21). Some evidence from the so-called

THE MILITARY PRESENCE IN PETRA

275

“Great Temple” in Petra was also interpreted as related to the Roman assault of the structure (Joukowsky 2007: 229). Neither a natural nor an intentional cause of the destructions mentioned above can be fully ascertained. Equally puzzling is that the Romans advertised the annexation only five years later, and the early provincial coins feature the term adquisita rather than capta (Bowersock 1983: 81; Spijkerman 1978: 32–35). Furthermore, the treatment of the “acquired” territory, and its ruling elite was unusually harsh: the name of Nabataea was suppressed, the royal family vanished into obscurity, and there is no indication of the incorporation of Nabataean aristocracy into the Roman Senate. Nabataean military forces were apparently incorporated into the Roman army and organized as cohortes Ulpiae Petraeorum (Graf 1994). Regardless, whether there was or was not any active Nabataean resistance to the Roman takeover, it is apparent that the kingdom experienced some internal, inter-tribal, socio-political conflicts before the annexation (al-Otaibi 2011: 86–94, 100) and, as proposed here, such conflicts may have been persistent also after 106 CE. Specifically, there are two variables in operation here which need to be considered. The first one, which has so far been largely overlooked, is the possibility of significant internal unrest caused by the foreign takeover. The collapse of the Nabataean royal authority and the apparent fall of until then influential persons, families and clans might have activated considerable political unheaval and social disorder, at least among the inhabitants of the royal city of Petra, resulting in riots, arson, destruction of property and other acts of dissatisfaction or revenge. In a large population center, inhabited by people of different origins, social and economic interests and political orientation (e.g., Strabo XVI.4.21 on multiethnic inhabitants of Petra), as Petra undoubtedly was in the 1st–2nd centuries CE, internal socio-political conflicts must have been common, further stirred and amplified by a major political event, such as the end of the Nabataean monarchy. Secondly, political opposition and dissatisfaction, social tension and religious controversies all manifested in often violent demonstrations, clashes, riots and other urban-type unrest in large, multiethnic and multi-religious cities in the East in the Early Roman Period so Petra would likely not be an exception. Notably, the distribution of the major military units (legions) during the 1st–2nd centuries, with their bases in major population centers clearly indicates that the maintaining of internal

276

ZBIGNIEW T. FIEMA

peace and security was equally important as the defense against the Parthian neighbors (Butcher 2003: 405–411). Consequently, the epigraphic record indicates that during the early post-annexation period, major cities of Roman Arabia, such as Bostra, Gerasa, Gadara and Philadelphia were garrisoned by the army units (Freeman 1996: 101, 105–07). Again, one would assume that Petra, the old royal city and the most important urban center of the new province, would also have had a post-annexation garrison. A short excursus on the significance of the two major population centers in Arabia – Petra and Bostra – is instructive. The army of the Roman province of Arabia consisted, between 106 and the end of the 3rd century, of a legion (two/three attested, with the III Cyrenaica becoming a permanent garrison), and several alae and cohortes, with the actual number of units present in the province fluctuating in time (for all attested units, see Speidel 1977, Kennedy 2004: 47–49). The headquarters of the Arabian legion – legio III Cyrenaica (Ptolemy Geogr. V.17) – were located in Bostra. The provincial governor, being also the commander of the Arabian legion, would undoubtedly reside occasionally in the city but this fact hardly constitutes proof of the Bostra's administrative dominance, as some have suggested (e.g., Tracy 1999, 56). After all, being the main center of the occupation forces hardly adds to the political prestige (vide Jerusalem after the 1st Jewish Revolt), and the presence of the legion at Bostra rather reflects this city's strategic position relative to the Parthian frontier. Neither the extant evidence for a governor’s seat nor the fact that periodic assizes held by a governor occurred in different places, would support the preeminence of either Petra or Bostra (Haensch 1997: 238–44, for discussion). The location of its two most important urban centers – one in the north and the other in the south – would have necessitated the governor’s frequent journeys between Petra and Bostra and his periodic residence in both places (Gagos 2009: 389). As for the evidence of the civic status, titles and honorifics bestowed by the Romans on Petra and Bostra, these of Petra – Augusta colonia Antoniniana nobilis ingenua mater coloniarum (or metrocolonia) Hadriana Petra Metropolis Arabiae (P. Petra III 23) – are particularly impressive and were still in use in 6th century. Under Trajan, Petra was recognized as a metropolis of Arabia (Sartre 1993: 67–68, no. 37), it received the honorific, Hadriana Petra, probably during the emperor’s visit to the Near East in 130–31, and was granted the

THE MILITARY PRESENCE IN PETRA

277

status of colonia, possibly as early as 209–212, by Caracalla (Gitler 2005: 189–90), or later, by Elagabalus. The metrocolonia honorific of Petra is epigraphically attested in the 3rd century (Starcky and Bennett 1968: 60–62, No. XIII; Sartre 1993, 76–78, no 48), possibly granted by Severus Alexander. This rather unprecedented title is attested in the East only for Palmyra, Antioch and perhaps Emesa (Millar 1990, 41, 44–45, 51–52). On the other hand, titles and honorifics of Bostra – metropolis and colonia – appear to post-date those of Petra and the rare, honorific metrocolonia as well as the qualifying adjectives – nobilis and ingenua – are not attested for Bostra at all. The evidence of honorifics, status designations and the dates of bestowal would undoubtedly reflect an escalation of civic rivalry between Petra and Bostra in which Petra apparently held the upper hand. The meteoric advancement of Petra in the constellation of Roman provincial cities is, nevertheless, astonishing and various explanations for this phenomenon were offered. For example, it was suggested that the title of metropolis was granted as a “compensation for Bostra’s elevation to the status of the provincial capital” (Sartre 1985: 74). This argument, however, appears unconvincing, and not only because the idea of a provincial “capital” in the Early Roman period cannot be sustained (Millar 1993: 94–95). It has been convincingly demonstrated that until the reign of Hadrian, provinces possessed only one, sole metropolis – a status officially sanctioned by Rome and bestowed upon the most important urban center in a province (Bowersock 1985: 79–80). Furthermore, if a “compensation” was meant it might perhaps better reflect the recognition of the imperial authorities of the Petraean effort in overcoming the difficulties associated with the transition from the Nabataean to the Roman rule. Further, the impressive honorifics and status designations of Petra must indicate the pre-eminence of Petra in terms of administrative and status-related advancement within the Roman world as well as the loyalty of its inhabitants. It is then apparent neither the presence of the III Cyrenaica in Bostra nor its honorifics imply any superior status of that city over Petra. Nevertheless, a Roman military presence in Petra appears necessary considering the internal security as well as the strategic requirements of the early province (Kennedy 1980: 293), but no unequivocal archaeological remains of military camp or other installations have so far been identified in Petra. At Hegra (modern Madain Salih), the southernmost Nabataean town and an important

278

ZBIGNIEW T. FIEMA

commercial center, the remains of what appears to have been a fortified military camp or cantonment, most probably from the Roman period, were recently discovered as well as several Roman military inscriptions (see Villeneuve and Fiema in Abu-Azizeh et al., forthcoming). Also at Bostra, the remains of the legionary fortress have long been known (Lenoir 2002; Bishop 2012: 58–59). Despite this, the earliest indication of the Roman military presence in Petra is provided by Papyrus Michigan 466 (26th March, 107 CE) which mentions soldiers of an unnamed legion (probably the III Cyrenaica) in Petra. 1 In fact, the III Cyrenaica is only legionary unit attested by several 2nd-3rd century inscriptions in Petra, and since the headquarters were located in Bostra, these must have been smaller detachments. It is also possible that some of the inscriptions listed below may have been left by the military personnel on visit to Petra but not stationed there. The list includes a short carving stating the III Cyr in Greek and located in the Siq of Petra (Zayadine and Fiema 1986, 199–200; Sartre 1993, 52, no. 19). The Latin epitaph of C. Antonius Valens, a trooper in the legionary cavalry of the III Cyrenaica, and roughly dated to the 2nd-3rd century, was found in the environs of Petra (Bennett and Kennedy 1978; Sartre 1993: 87, no 52). A small altar, currently in the museum in Petra, preserves a Latin inscription of soldiers of cohors Aurelia or Aureliana belonging to the III Cyrenaica (Sartre 1993: 71–72, no 44). The museum also holds an incomplete Greek epitaph mentioning a son or a daughter of a soldier of the III Cyrenaica (Sartre 1993: 97, no 61). Additionally, one should mention the Greek epitaph of Arrianos, a soldier, probably of the same legion (Sartre 1993: 91–94, no. 55) and a Greek dedication by Victorinus beneficiarius, found in the Siq of Petra (Zayadine 1981, 352; Sartre 1993, 53–54, no 20). Several military inscriptions were found on the temenos of the Qasr el Bint temple in the ceremonial center of Petra. One is the Latin honorific inscription mentioning opt(iones) l(egio)nis, probably referring to the III Cyrenaica (Starcky and Bennet 1968: 59–60, no. XII; Sartre 1993: 75–76, no 47). The Greek inscription honoring See Speidel 1977: 691–94. and Kennedy 1980: 289–92; 2004, 46, 175–76, for the analysis of this document and the discussion on the early garrison of Arabia 1

THE MILITARY PRESENCE IN PETRA

279

M. Ulpius Andromachus, praefectus of Ala II Ulpia Auriana and set up by the decuriones of this unit (Starcky and Bennett 1968: 57–59, no. XI; Sartre 1993: 78–79, no 49) was also found on the temenos. A Latin inscription found reused as a step in the Ridge church in Petra mentions the same honorand and his unit, Lucius signifer being in charge of the setting-up the monument (Salomies and Fiema 2009). However, due to the fact that Ala II Auriana was never attested in Petra or Arabia but in Cappadocia, and Syria (Spaul 1994: 61), it is generally assumed that both inscriptions were set up in Petra only because it was Andromachus' hometown (Speidel 1974: 936; 1977: 707).

THE BYZANTINE PERIOD

The military arrangements in the area of Petra during the Byzantine period (early 4th-early 7th century A.D.) are much better known not only through the extant archaeological remains but also as substantially expanded through the information in the NotitiaDignitatum (Or. XXXIV, ed. Seeck). Under the command of the dux Palaestinae, military garrisons were stationed to the east of Petra, in Admatha (modern al-Hammam), to the southwest in Arieldela (ʿAin Gharandal), to the north in Robatha (Khirbat Ruwath?), and to the south in Zodocatha/Kastron Zadacathon (Sadaqa) and Hauana/Hauarra (Humayma). Other military units were also stationed in Biʿr Madhkur and Wadi Sabra, while Roman/Byzantine fortlets or watchtowers around Petra, include Qasr Umm al-Rattam, Qasr Wadi at-Tayyiba, al-Mutrab, Ail, and Fardhakh, among others. 2 At Udhruh (ancient Augustopolis), ca 10 km east of Petra, a major legionary fortress was built around A.D. 300 for the legio VI Ferrata (Kennedy and Falahat 2008; Bishop 2012: 127), and the legion was stationed there at least in the early 4th century. The 6th century Greek Byzantine papyri from Petra provide further information on the military presence in the Petra’s countryside, prominently featuring military personnel in Kastron Zadacathon but also in other locations (Fiema 2007). For all sites mentioned, see Parker 1987, 2000; Graf 1992, 1995 and 1997; Fiema 1995. 2

280

ZBIGNIEW T. FIEMA

Undoubtedly, in addition to the usual social problems in the urban context which needed to be monitored and neutralized by military forces, the 4th–5th century period in the East witnessed another, powerful conflict dimension – the struggle between paganism and Christianity. Being often violent and associated with arson, destruction and armed clashes, these events often required the intervention of the army. 3 Petra, being the Byzantine provincial capital and the metropolitan bishopric (since 451 CE) must have had its fair share of such conflicts. The city is not mentioned as garrisoned in the Notitia Dignitatum. However, except for the reference to the 363 earthquake, basically all other extant references to Petra in the 4th century emphasize the conflicts between pagans and Christians (e.g., Eusebius, Comm. in Isaiam II. 23.38–41, Sozomen, HE VII.15,11). Undoubtedly, Petra, as many other cities of the Byzantine East, was a potentially volatile place where the supporters of polytheism, Jews, and the proponents of different theological orientations of Christianity clashed with each other and where the ritualized violence was a common occurence. The apparently continuing pagan ceremonies and observances could have been either intentionally, or through ignorance, misinterpreted by the Christians as mocking their religion and would have provoked violent disputes, and clashes. This was not helped by the fact that Christians at Petra were themselves deeply divided over theological matters (see Fiema 2002b, 2015 for sources and discussion). A single tantalizing piece of existence comes from the mid-5th century and it might, possibly, relate to the situation in Petra following the apocryphical story of the Christian monk Bar Sauma, which supposedly took place several years earlier (Nau 1927; see Sivan 2008: 175–86, 215–17 for other exploits of that monk). Bar Sauma had encountered the pagan temples in Petra around 419–23. Upon the monk’s call to abandon pagan practices and temples, the inhabitants of Petra closed the city gates against him, despite his threats that he would burn the city down. Only divine intervention – a miraculous rain in the city suffering a long drought – changed See Sivan 2008: 117–23, 145–56, 167–75, 333–37 for examples of such in the Palestinian cities. 3

THE MILITARY PRESENCE IN PETRA

281

their attitude and led to mass conversions, including the pagan priests. Only two decades later, the so-called Urn Tomb, an old Nabataean funerary monument, was converted into a cathedral and dedicated on July 24, 446, under Bishop Jason (Sartre 1993: 81–84, no. 50); the first known and dated Christian edifice in Petra. Admittedly, the tomb is of monumental proportions, with a spectacular view over the Valley of Petra and, furthermore, may reflect pragmatic economizing by local Christians in recycling the already extant structure. One may equally propose that local Christian community lacked either means or influence to construct a proper cathedral building in the center of the city. But what is particularly puzzling here is the specific mention in the dedicatory inscription of a numerus, a Late Roman army unit, in whose presence the consecration of the cathedral took place. It may be that only 20 years after the Bar Sauma episode, the act of the appropriation by Christians of a significant monument in the center of the city required not only an official imperial sanction but also an active support, including the security and protection provided by the military forces rather than the city authorities. The name of the military unit in question is reconstructed as numerus of the Most Brave Tertiodalmatians (Sartre 1993: 82). Under the command of dux Palaestinae, the Notitia dignitatum lists Equites Dalmatae Illyriciani (Or. XXXIV. stationed in Berosaba (Beersheva). Also, Equites Nonodalmatae are attested in Umm al-Jimal (Byzantine Arabia) by the inscription from 371 CE (Speidel 1977: 718–719). Theoretically, a detachment of the Dalmatians might have been transferred to Petra but whether it formed a permanent garrison or was brought in only to secure peace during the conversion ceremony, is virtually unknown. A poorly preserved epigram which prevents exact interpretation, appears to commemorate the heroic exploits of a certain Orion (Horion) who gave money and saved the city and all the inhabitants of Palaestina Salutaris from a barbarian-speaking enemy. The city wall is also mentioned, which was presumably restored/rebuilt (Bikai and Tracy 2002, Moralee 2011). The date is uncertain but the mention of Palaestina Salutaris would imply the date after ca. 392 but before 409 (Cod. Th. 7.4.30) when Salutaris at least officially became Tertia (for reorganizations, see Sipilä 2009: 166–74). However, the name of Salutaris seems to have lingered for a while, e.g., still mentioned in the acts of the Council of Ephesus (431). Thus it is

282

ZBIGNIEW T. FIEMA

safe to assign the end of the 4th-5th century date to this inscription. It is impossible to assess the claims of Orion or associate his exploits with any known historical, large-scale event. However, two important facets of late antique reality in the East are well-reflected by this inscription. The first is the increased threat posed to the settled populations by the so-called Saracens, the nomadic tribes (for discussion, see Parker 1987, 2000). This threat, beginning from the 4th century, was addressed by the imperial authorities by increasing dependency on the local phylarchs and their tribes for the defense of the countryside. The process culminated in 530 or 531 with Justinian’s recognition of the Ghassanid/Jafnid al-Harith bin Jabala as the King of all Arabs and his brother, Abu Karib ibn Jabala, as the phylarch of Palaestina Tertia (Fiema and Nehme 2015: 385–88 for discussion; see also Fiema 2007). Although the Saracen threat in southern Transjordan seems to have been of low level and intensity, a possibility of a raid repulsed by the inhabitants of Petra is not unlikely. Secondly, the inscription confirms the late antique trend of the cities’ increasing reliance on local defensive means, such as urban militias and city-walls, the upkeep and maintenance of which being largely the responsibility of the local populations. In this context, it is worth to mention a 4th or 5th century inscription found in the Wadi Sleissel, on the northwestern approach to Petra, in which Abdobodas, son of Abdobodas is commemorated by soldiers of ex-magister. Despite the grand-sounding title that magister was either a relatively junior commander of the city’s garrison or, more likely, a leader of a local militia which patrolled the access roads (Zayadine 1992: 218–22; Sartre 1993: 65–66, no. 36.) Returning to the internal security concerns, it is also worth mentioning that during the reigns of Anastasius (491–518) and Justin I (518–527), Petra acquired a dubious distinction of being a place of banishment for both ecclesiastics and common criminals (for details, see Fiema 2002b: 193; 2015). It would seem that in case of exile resulting from the religious opposition of an ecclesiastic to the creed professed by the ruling emperor, the said ecclesiatic should be sent to a place where his own religious views would not find the majority of adherents. However, in this case, both the Chalcedonian and Monophysite emperors banished the opposing ecclesiastics to Petra. Thus the argument that the city was considered a safe and loyal city to house individuals “dangerous” to the central government does not seem to apply. Rather, by that time,

THE MILITARY PRESENCE IN PETRA

283

Petra’s status was reduced to a distant and insignificant place where the presence of the exiles would be harmless to the interests of Constantinople and where, most probably, the military garrison would be capable of dealing with potential sedition or religiouslyincited rebellion. Neither the Persian occupation of the Byzantine East (614– 629) nor the Muslim conquest of Syro-Palestine (633–636) are associated with any historical information referring to Petra. The fact that Petra did not warrant even a short note in the account of the Islamic conquest, as opposed to smaller towns in the region such as Aila, Udhruh, and al-Jarba, cannot be the result of a simple oversight; the political significance of Petra then must already have dwindled to a minimum. 4 It is thus most probable that by the early 7th century, Byzantine military forces were no longer stationed in Petra.

BIBLIOGRAPHY

Abu Azizeh, W., N. Delhopital, C. Durand, Z. T. Fiema, Y. Gerber, and F. Villeneuve. (forth). Report on the 2014 and 2015 Excavation Seasons in Some Areas at Madâʾin Sâlih. ATLAL – Journal of Saudi Arabian Archaeology. Arjava, A., M. Buchholz, and T. Gagos, eds. (P2007). The Petra Papyri III. American Center of Oriental Research Publications 5. Amman: American Center of Oriental Research. Bennett, C. and D. Kennedy. (1978). A New Roman Military Inscription from Petra. Levant 10, 163–65. Bikai, P. M. and S. V. Tracy. (2002). Petra. In: R. Merkelbach and J. Stauber, eds., Steinepigramme aus dem griechischen Osten. Band 4. Die Südküste Kleinasiens, Syrien und Palaestina. Munich: K. G. Saur Verlag, p. 445. Bishop, M. C. (2012). Handbook to Roman Legionary Fortresses. Barnsley: Pen and Sword. Bowersock, G. (1983). Roman Arabia. Cambridge, Mass: Harvard University Press. 4

See Kaegi 1992: 82–83 for sources and discussion.

284

ZBIGNIEW T. FIEMA

Bowersock, G. (1985). Hadrian and Metropolis. In: J. Béranger and J. Straub. Bonner Historia-Augusta-Colloquium, 1982–1983, Bonn: Habelt, pp. 75–88. Butcher, K. (2003). Roman Syria and the Near East. Los Angeles: Getty Publications. Eusebius. (1975). Der Jesajakommentar. Edited by J. Ziegler. GChS. Volume 9, Eusebius Werke. Berlin: Akademie-Verlag. Fiema, Z. (1987).The Roman Annexation of Arabia: A General Perspective. The Ancient World 15, (1–2), 25–35. Fiema, Z. (1988). The Era of Bostra – A Reconsideration. In: R. Campanati and G. Fiaccadori, eds., The Proceedings of the XXXV Corso di cultura sull’arte Ravennate e Bizantina. Ravenna: Edizioni del Girasole, pp. 109–121. Fiema, Z. (1995). Military Architecture and “Defense” System in Roman-Byzantine Southern Jordan. A Critical Appraisal of Recent Interpretations. SHAJ 5, 261–270. Fiema, Z. (2002a). The Military Presence in the Countryside of Petra in the 6th Century. In: P. Freeman, J. Bennett, Z. Fiema and B. Hoffmann, eds, Limes XVIII. Proceedings of the XVIIIth International Congress of Roman Frontier Studies held in Amman, Jordan, September 2000. BAR_IS 1084. Oxford: Archaeopress, pp. 131–36. Fiema, Z. (2002b). Petra and Its Hinterland during the Byzantine Period: New Research. JRA Supplementary Series 49, Portsmouth, RI: JRA, pp. 191–252. Fiema, Z. (2003). Roman Petra (A.D. 106–363). A Neglected Subject. ZDPV 119(1), 38–58. Fiema, Z. (2007).The Byzantine Military in the Petra Papyri – A Summary. In: Ariel S. Lewin and Pietrina Pellegrini, eds, The Late Roman Army in the Near East from Diocletian to the Arab Conquest. Proceedings of a Colloquium held in Potenza, Acerenza and Matera, Italy, 10–13 May 2005. BAR-IS 1717. Oxford: Archaeopress, pp. 313–319. Fiema, Z (2015). Petra. In: G. Schöllgen et al., eds, Reallexikon für Antike und Christentum. Stuttgart: A. Hierseman, Band XXVII, Columns 349–88. Fiema, Z. and L. Nehmé. (2015). Petra and Hegra between the

THE MILITARY PRESENCE IN PETRA

285

Roman Annexation and the Coming of Islam. In: G. Fisher, ed., Arabs and Empires before Islam. Oxford: Oxford University Press, pp. 373–95. Freeman, P. (1996). The Annexation of Arabia and Imperial Grand Strategy. In: D. Kennedy, ed.,The Roman Army in the East. JRA Supplementary Series 18, Ann Arbor: JRA, pp. 91–118. Gagos, T. (2009). New Inscriptional Evidence and Petra's Role in the Province of Arabia. JRA 22, 381–389. Gitler, H. (2005). A Group of 120 Clay Bullae from Petra with Titles of the City. The Numismatic Chronicle 165, 185–92. Graf, D.(1989). Rome and the Saracens: Reassessing the Nomadic Menace. In: T. Fahd, ed., L’Arabie préislamique et son environnement historique et culturel. Actes du Colloque de Strasbourg 1987. Leiden: Brill, pp. 341–400. Graf, D. (1992). Nabataean Settlements and Roman Occupation in Arabia Petraea. SHAJ 4, 253–60. Graf, D. (1994). The Nabataean Army and the Cohortes Ulpiae Petraeorum. In: E. Dabrowa, ed., The Roman and Byzantine Army in the East. Kraków: Drukarnia Uniwersytetu Jagiellońskiego pp. 265–309. Graf, D. (1995). The Via Nova Traiana in Arabia Petraea. In: J. Humphrey, ed., The Roman and Byzantine Near East: Some Recent Archaeological Research. JRA Supplementary Series, 14. Portsmouth, RI: JRA, pp. 241–65. Graf, D. (1997). The Via Militaris and the Limes Arabicus. In: W. Groenman-van Waateringe, B. L. van Beek, W. J. H. Willems, and S. L. Wynia, eds., Roman Frontier Studies 1995. Oxford: Archaepress, pp. 123–33. Haensch, R. (1997). Capita Provinciarum. Statthaltersitze und Provinzialverwaltung in der Römischen Kaizerzeit. Mainz: P. von Zabern Isaac, B. (1990). The Limits of Empire: The Roman Army in the East. Oxford: Oxford University Press. Josephus. (1962). De Bello Judaico. Der Jüdische Krieg. Edited by O. Michel and O. Bauernfeind. Darmstadt: Kösel-Verlag Joukowsky, M. (2007). Petra Great Temple. Vol. II. Archaeological Contexts of the Remains and Excavations. Providence, R.I: Brown University.

286

ZBIGNIEW T. FIEMA

Kaegi, W. (1992). Byzantium and the Early Islamic Conquests. Cambridge: Cambridge University Press. Kennedy, D. L. (1980). Legio VI Ferrata: The Annexation and Early Garrison of Arabia. Harvard Studies in Classical Philology 84, 283–309. Kennedy, D. L. (2004). The Roman Army in Jordan. London: Council for British Research in the Levant. Kennedy, D. L. and H. Falahat. (2008). Castra Legionis VI Ferratae. A Building Inscription for the Legionary Fortress at Udhruh near Petra. JRA 21, 151–169. Lenoir, M. (2002). Le camp de la légion IIIa Cyrenaica a Bostra. Recherches récentes. In: P. Freeman, J. Bennett, Z. Fiema and B. Hoffmann, eds, Limes XVIII. Proceedings of the XVIIIth International Congress of Roman Frontier Studies held in Amman, Jordan, September 2000. BAR-IS 1084. Oxford: Archaeopress, pp. 175–84. Millar, F. (1993). The Roman Near East 31 BC – AD 337. Cambridge: Harvard University Press. Moralee, Jason. (2011). Petra and the Saracens: New Evidence from a Recently Discovered Epigram. In: R. W. Mathisen and D. Shanzer, eds., Romans, Barbarians, and the Transformation of the Roman World. Farnham: Ashgate, pp. 233–240. Nau, F. (1927). Deux épisodes de l’histoire juive sous Théodose II (423 et 438) d’après la vie de Barsauma le Syrien. Revue des etudes Juives 83, 184–206. Notitia Dignitatum. (1876). Edited by O. Seeck. Berlin: Weidmann. al-Otaibi, F. (2011). From Nabataea to Roman Arabia: Acquisition or Conquest? BAR-IS 2212. Oxford: Archaeopress. Parker, S. T. (1986). Romans and Saracens: A History of the Arabian Frontier. Philadelphia, Pa.: American Schools of Oriental Research. Parker, S. T. (1987). Peasants, Pastoralists, and Pax Romana: A Different View. BASOR 265, 35–51. Parker, S. T. (2000). The Defense of Palestine and Transjordan from Diocletian to Heraclius. In: L. E. Stager, J. A. Greene, and M. Coogan, eds, The Archaeology of Jordan and Beyond: Essays

THE MILITARY PRESENCE IN PETRA

287

in Honor of James A. Sauer. Winona Lake, IN: Eisenbrauns, pp. 367–88. Ptolemy. (1932). The Geography by Claudius Ptolemy. Translated by E. L. Stevenson. New York: New York Public library. Salomies, O. and Z. Fiema. (2009). Latin Inscription in the Ridge Church at Petra. ARCTOS 43, 117–140. Sartre, M. (1985) Bostra. Des origines à l'Islam. Paris: Institut Francais d’Archéologie du Proche-Orient. Sartre, M. (1993). Pétra et la Nabatène méridionale du Wadi al-Hasa au golfe de ‘Aqaba. Inscriptions de la Jordanie. Tome IV. (Inscriptions Grecques et Latines de la Syrie. Tome XXI.) Paris: P. Geuthner. Sipilä, J. (2009). The Reorganisation of Provincial Territories in Light of the Imperial Decision-Making Process. Helsinki: Societas Scientiarum Fennica. Schmid, S. (1997). Nabataean Fine Ware Pottery and the Destructions of Petra in the Late First and Early Second Century AD. SHAJ 6, 413–420. Sozomenus. (1960). Kirchengeschichte. J. Bidez and G. Ch. Hansen, eds, Berlin: Akademie-Verlag. Spaul, J. (1994). Ala2. The Auxiliary Cavalry Units of the PreDiocletianic Imperial Roman Army. Andover: Nectoreca Press. Sivan, H. (2008). Palestine in Late Antiquity. Oxford: Oxford University Press. Speidel, M. (1974). Exercitus Arabicus. Latomus, 33, 934–39. Speidel, M. (1977). The Roman Army in Arabia. ANRW II.8 pp. 687–730 Spijkerman, A. (1978). The Coins of the Decapolis and Provincia Arabia. Studium Biblicum Franciscanum Collection Maior, 25. Jerusalem: Franciscan Printing Press. Starcky, J. and C. Bennett. (1968). Les incriptions du temenos (Petra). Syria 45, 41–65 Strabo. (1961). The Geography of Strabo. Edited by H. L. Jones. Loeb Classsical Library 241. Cambridge: Harvard University Press. Stucky, R. (1996). Die Nabatäische Bauten. In: A. Bignasca et al., eds, Petra ez Zantur I. Ergebnisse der Schweizerisch-Liechtensteinischen

288

ZBIGNIEW T. FIEMA

Ausgrabungen 1988–1992. Terra archaeologica II. Mainz: Verlag Philipp von Zabern, pp.13–50. Tracy, S. (1999). The Dedicatory Inscription to Trajan at the “Metropolis” of Petra. In: J. H. Humphrey, ed., The Roman and Byzantine Near East II. JRA Supplementary Series 31. Portsmouth, RI: JRA, pp. 51–58. Zayadine, F. (1981). Recent Excavations and Restoration of the Department of Antiquities (1979–1980). ADAJ 25, 341–355. Zayadine, F. (1992). L’espace urbain du Grand Pétra. Les Routes et les stations caravanieres. ADAJ 36, 217–39. Zayadine, F. and Z. Fiema. (1986). Roman Inscriptions of the Siq of Petra. Remarks on the Initial Garrison of Arabia. ADAJ 30, 199–206.

THE ECONOMY OF CAESAREA PALAESTINAE IN LATE ANTIQUITY: STRUCTURE AND SCALE KENNETH G. HOLUM From their harvests of grain, wine, and oil were they enriched (Psalms 4:8). 1

The psalmist’s words, set into a mosaic pavement decorating a mansion in Caesarea’s northeastern suburb, focus our attention on the economies of ancient Mediterranean cities. Commenting at the 2004 ASOR meeting in Philadelphia, I took issue with Tom Parker’s interpretation of the economy of Aila, modern Aqaba in Jordan, the Roman city that he excavated from 1996–2002. In this 2004 paper as in the admirable first volume of the final reports, Parker expresses as the overall objective of his project evaluating modern theories about the ancient Roman economy (Parker and Smith 2014: 1–5). Against Parker and an emerging consensus – at least among archaeologists in the theorizing community – I argued in 2004 that despite the Aila evidence we should accept with the late M. I. Finley and his school the view that the economies of the Mediterranean-style cities were “primitive,” based on agriculture, as opposed to “modern” cities drawing wealth primarily from trade and industry (Finley 1999). Corollaries of the “primitivist” school, as Finley represented it, are that the Mediterranean cities were CIIP 2: 1173 (my translation), presenting the Septuagint text. I am grateful to Marsha Rozenblit and Mark Holum for reading this paper and suggesting improvements, to Yoram Tsafrir and Anna Iamim for assistance with fig. 3, and to Joseph Patrich for permission to use figs. 4–6. 1

291

292

KENNETH G. HOLUM

“consumer” cities that extracted their wealth from the surrounding countryside, and furthermore that their economies were “embedded” in a moral system that impelled the local elite to concentrate wealth not for purely economic reasons, as in capitalism, but for the purpose of displaying and reinforcing status. 2 In 2004, however, I did not have the opportunity to draw upon my own research on Caesarea Palaestinae in order to explain my adherence to the Finley school. Having left myself in a weak position, I propose in the following pages to offer Caesarea as a counterpoint to Aila, and thus to commend Parker warmly not only for extracting new evidence on this topic but for stimulating discussion.

Figure 1. Caesarea’s fertile territory, looking southwest toward Caesarea, across the valley of the Nahal Tanninim and the Nahal Ada. At right are Khirbet Shuni and the southern “nose” of Mt. Carmel. Aaron Levin photo (1986)

A larger city than Aila, Caesarea extended over 113.5 ha inside its fortification walls, and accommodated perhaps 35,000 inhabitants Finley and his school owed much to the theorizing of the Hungarian-American economist Karl Polyani who taught Finley at Columbia and of the great German sociologist Max Weber. See especially Morris’ foreword to Finley, Finley 1999: ix–xxxvi). 2

THE ECONOMY OF CAESAREA PALAESTINAE

293

within them (fig. 1). 3 Located in a very different ecological zone, Caesarea nevertheless resembled Aila as a harbor city and nexus of trade routes. I focus on Caesarea during the last three centuries of ancient Mediterranean civilization, termed “Byzantine” commonly in the region but more generally “Late Antique,” because the evidence from this period is most plentiful, as is the case for Aila. As will become clear, within Late Antiquity I have in mind especially the sixth century, on account of several critical passages in the Secret History of Procopius, himself a native of Caesarea. Along the lines of recent research on the ancient economy, I propose to discuss the evidence, both written and material, in terms of structure and scale. As I employ these terms, scale means statistically-based study of wealth-producing activity and of growth or recession over time, while structure is the study of social and institutional foundations of such activity according to sectors, such as agriculture and stock raising, commerce and markets, transportation, manufacturing, the building trades, social and religious institutions, and the government sector. 4 In the ensuing discussion, where actual statistics are lacking, I have tried to put forward educated guesses that are on the modest side. Fundamental to my approach is the view – uncontested for the most part but often underestimated in the scholarship – that most ancient Mediterranean cities corresponded more or less closely to an ideal type known in Greek as the polis in the eastern Mediterranean and in the Latin West as the civitas. This type of human settlement emerged in Greece and western Asia Minor between the eighth and sixth centuries BCE and passed through a more-or-less continuous history of more than a millennium until after the barbarian settlements in the western Mediterranean and the Muslim Holum 2011: 22, n. 44 accepting the coefficient (400 persons per ha) of Broshi 1980: 1–10, but adopting 113.5 ha as the total of Caesarea’s intramural space; cf. Decker 2009: 20. 4 Scheidel 2012: 7–8; Erdkamp 2012: 247–53; and, employing the terms in a different sense, Duncan-Jone 1990. For an alternative approach to the Caesarea economy see Patrich 2011 assembling a large body of evidence for production and commerce, especially archaeological data. 3

294

KENNETH G. HOLUM

conquest in the East. 5 A. H. M. Jones accepted the number of about 900 poleis in the East during Late Antiquity and at least an equal number of civitates in the West (Jones 1966: 2.712–18). Despite variations, several common features normally characterized the polis or civitas. Peculiar to this settlement type was the integration of city and the countryside with its villages and isolated farms into a single economic, political, and religious organism. As Horden and Purcell declared, we should conceive of the cities themselves, i.e. the conglomerated urban centers, less as separate and clearly definable entities and more as loci of contact or overlap between different ecologies… [as] settings in which ecological process may be intense, and in which the anthropogenic effect is at its most pronounced…

but not … more than that. And they should not be presented as conceptually detachable from the remainder of the spectrum of settlement types (Horden and Purcell 2000: 100–101).

To put it bluntly, from the economic perspective city and countryside constituted a single terrain and ought not be conceived as in opposition to one another. Furthermore, the factor that united city and countryside into a single terrain was a distinct social and economic group, the local elite composed of landlords inhabiting the built-up urban center who extracted wealth to support an extravagant urban lifestyle from their agricultural estates in the countryside. These were essentially rentiers whose economic objective was to live from the labor of others, yet they not only leased parcels of land to peasant farmers in order to receive rents from them in cash or in kind, but they also carried out or presided over various public duties in their cities (munera, leitourgiai) including, at least until sometime in the course of the sixth century, administering the often brutal process of collecting taxes from the same peasants and forwarding the amounts extracted to agents of the Roman state. At the beginning of our period these were the curiales in the West, bouleutai in the East, or “councilors,” but by the sixth century the wealthiest among the local elite had come to be known generally as possessores 5

Holum, 2005, 2011, 2014; Whittow 1990; Liebeschuetz 1972.

THE ECONOMY OF CAESAREA PALAESTINAE

295

or ktētores, “landlords,” because it was from their estates in the territory that they extracted their incomes. In the Novels of Justinian, they also appear as nobiles viri (30 praef.), and modern scholars generally call these elite of the elite simply “notables.” 6 Finally, it is universally acknowledged that what created the monumentality and civic solidarity that characterized cities in the High Empire, and that still impresses tourists today, was the quest of these elites for status through acts of euergetism or munificence, as they financed, out of the profits of their rural estates, the construction of temples, public baths, and porticoes, and the production of spectacles in the theater, amphitheater, and hippodrome. 7 This quest continued in Late Antiquity, but increasingly the elite poured their wealth into building and decorating their domiciles, ever more resplendent markers of social esteem, and into churches, of course, and the care of the poor, thus “storing up treasures in heaven,” while they financed other public building increasingly by redirecting tax money, with the emperor’s permission, to projects in their own cities. 8 Of course, were this fundamental typology of the ancient Mediterranean city accepted, it would already be clear that the Finley approach to the ancient economy has much to recommend it.

In general Laniado 2002, 2014; Liebeschuetz 1972: 104–36; for Caesarea, Holum 1995. 7 The classic study is Veyne 1990, especially chapter one. 8 Brown 2012: 53–90, especially 64: “Everything that had made a Roman city what it was sprang from … the love of a leading citizen for his hometown; also Di Segni 1995, 1999; cf. Duncan-Jones 1990: 178–82, on building in Roman Thugga, Tunisia, where the preponderance of temple construction indicates that pagans, too, in their own fashion, might “store up treasures in heaven.” 6

296

KENNETH G. HOLUM

Figure 2. Caesarea in Late Antiquity. Map by Anna Iamim

THE ECONOMY OF CAESAREA PALAESTINAE

Figure 3. Caesarea in its territory, map from Tsafrir, Di Segni, and Green, Iudaea-Palaestina, fold-out, edited by Anna Iamim, by permission.

297

298

KENNETH G. HOLUM

Let us nevertheless examine briefly aspects of Caesarea’s economy that might support the Finley approach, or might alternatively encourage us to modify it. First, I would like to address the agricultural sector. In contrast with Aila, where the hinterland was largely desert, Caesarea was mistress of an expansive territory, chōra in Greek, that extended ca. 45 km. north-south and as much as 30 km. east-west, bordering on the territories of Dora to the north, Apollonia on the south, Legio-Maximianopolis on the east, and Samaria-Sebastē on the southeast (figs. 2–3). 9 Caesarea itself was “situated among the sands,” declared Rabbi Abbahu early in the fourth century (Babylonian Talmud Megillah 6a), referring to dunes that formed along the coast during prehistory in the city’s immediate environs, where the inhabitants were nevertheless able to establish irrigated garden plots, fertilized with city garbage, that contributed to feeding the population. 10 Beyond the dunes, the territory included, on the north, the southernmost “nose” of Mt. Carmel and probably also the hilly “Inner Carmel” at least as far northward as the excavated site known today as Summaqa, but most of it was relatively flat land in the northern part of the coastal Plain of Sharon, extending from the sea on the west into the hill country of Menasseh on the northeast and of Samaria to the east. Intersecting the Sharon here were east-west wadi systems that drained the slopes westward across the flats to the sea – Nahal Tanninim (“Crocodile River”), with its tributary the Nahal Ada, then the Nahal Hadera and the Nahal Alexander with their respective tributaries. Passing through the zone of sand dunes as they approached the shore, the streams tended to form marshes, but not far inland the fertile soils of the inner Sharon and the hills (terra rossa, rendzina) yielded abundantly to the farmers’ labors. In antiquity forests of Tabor oak, called the Drymos, now all but vanished, still occupied See my forthcoming article “Caesarea Palaestinae: City and Countryside.” The exact boundaries of Caesarea’s territory are now uncertain but, as I assert, were well-defined in antiquity. 10 See (again) my article “City and Countryside” (forthcoming) and Porath 1993: 3; 2008: 1663. 9

THE ECONOMY OF CAESAREA PALAESTINAE

299

tracts of the Sharon. 11 Nevertheless, of somewhere around 1,200 km2 total land area in Caesarea’s territory, 12 a third to a half will have been productive ploughland, olive groves, and vineland that was relatively well-watered by winter rains averaging 500–600 mm. per annum in modern times. 13 In his Secret History the historian Procopius, a native son, calls Caesarea’s territory “the best land [chōra] in the world” (11.30). 14 When we turn to settlement patterns, considerations of scale come in. The Archaeological Survey of Israel has published surveys of four 10 x 10 km grid squares that certainly lay entirely within Caesarea’s territory, of which about 300 km2 was on dry land – the rest of the grid squares extending into the sea. The surveyors discerned a wide variety of sites, including villages and hamlets, isolated farmsteads, wine- and oil-presses, columbaria, quarries, and cemeteries. Sites of all types in the 300 km2 totaled 171 in the Roman period, first through third centuries, and rose to 238 in the Byzantine period. 15 In my view, the total number of sites in the For references see Tsafrir, Di Segni, and Green 1994: 114 “Drymus” and 223 “Sarona II,” and cf. Orni and Efrat 1973: 170. 12 This is the estimate of G. Beyer 1936: 15, perhaps a bit high for the Crusader period but about right for Late Antiquity. For comparison: in the fourth century the Egyptian nomes, analogous to city territories elsewhere, ranged between 72 km2, the Ombite nome, and 1140 km2, the Hermopolite nome; see Bagnall 1990: 333–35. The city territory of Cyrrhus, a city in northern Syria, measured 40 x 40 (Roman) miles in the fifth century, thus 3,500 km2, but much of this was mountainous; see Duncan-Jones 1990: 202–3, and Decker 2009: 20. 13 For geology and climate see Orni and Efrat 1973: 35–44, 47–49, 68–73, 142–48, 154–55. 14 Similarly the Expositio, a fourth-century text, hails Caesarea and other cities of Palestine as “abundans omnibus” and “fructiferae in frumento, vino et oleo” (26, 31). 15 Holum 2011: 18–19, summarizing Ne’eman, Sender, and Oren 2000; Ne’eman 1990; Olami, Sender, and Oren, Map of Binyamina. Also entirely within Caesarea’s territory, but not in my calculation, lay the survey areas of Gadot and Tepper, Map of Regavim, available at http://www.antiquities.org.il/survey/new/default_en.aspx, and of Po11

300

KENNETH G. HOLUM

territory is a proxy for the density of settlement and the level of agricultural exploitation, so in this case the statistics demonstrate an upswing during Late Antiquity in an already high level of rural prosperity. Estimating still unsurveyed tracts, one might posit within ca. 1,200 km2 total land area about 500 settlements of all types, including perhaps 80–120 villages and hamlets, 16 several hundred isolated farmsteads, and a rather dense rural population of 100,000 or a bit more. 17 The contrast with modern settlement patterns, in which far more people inhabit the city than rural areas, could hardly be more striking. Within Caesarea’s territory, stratigraphic evidence from known sites of Late Antique settlement is in short supply, so we may apply a broader study published twenty years ago in which Yizhar Hirschfeld established a general typology of rural settlement in ancient Palestine. 18 He discerned three distinct settlement types. Isolated farmsteads consisted of a single dwelling unit with facilities for livestock, hamlets generally of three to ten such units located 40 m. or so apart, accommodating a few dozen occupants at most. Villages were larger and usually more compactly settled, ranging in size upward from 1–2 ha in space and from a few hundred to a few thousand inhabitants, featuring churches or synagogues and public baths but lacking the fortification circuits and urban planning that were characteristic of actual cities. Villages also lacked centralized public markets, so exchanges occurred at the site of production, in rath, Dar, and Apelba’um 1985, Emek Hefer. In my opinion the grid covered in Olami 1981 likewise lay entirely within Caesarea’s territory, but my view may be controversial. 16 According to Bagnall 1990: 110, n. 1, the Hermopolite nome, extending over ca. 1,000 km2, had about 100 villages, hamlets not being typical there. Occupation in the Nile Valley would have been somewhat denser than in Caesarea’s hinterland. 17 Decker 2009: 20, suggests a reasonable “urbanization rate” of 25%, so if 35,000 dwelled within the walls (above), the rural population was about 105,000. I fail to see how Decker calculates 175,000–225,000 in Caesarea’s hinterland, which seems far too high. 18 Hirschfeld 1997. On settlement patterns see also Safrai 1994: 17– 103 (with caution), and Decker 2009: 33–65.

THE ECONOMY OF CAESAREA PALAESTINAE

301

periodic fairs, or through the hands of itinerant peddlers. 19 In all three types of settlement, the inhabitants engaged in craft and small-scale industrial production (e.g. pottery and glass manufacture) and in processing agricultural products (employing e.g. wineand oil-presses, grain mills), but in all three types the primary modes of production were animal husbandry and cultivating the soil. Many in these settlements owned their land and left their dwellings daily to work it themselves, while others tended the fields of others, of rentiers, who might be wealthier villagers, their neighbors, or the great landlords of Caesarea. 20 Overall, the economy of rural settlements in Caesarea’s territory indicates a dispersal of productive capacity, and, from the economic perspective, reinforces the point of Horden and Purcell that the conglomerated city, in this case Caesarea, “should not be presented as conceptually detachable from the remainder of the spectrum of settlement types.” Hamlets are difficult to detect in the survey data, or to distinguish from isolated farms. Wine- and oil-presses and columbaria might characterize either. So far archaeologists have excavated no example of Caesarea’s large- or medium-sized villages, but a detailed surface survey has documented a well-preserved example of the latter in the east of the territory. Umm Riḥan lay on the western slopes of the Samaria hills, extending over 4 ha or so, with 100 dwelling units, a public bath, winepresses, and a population of 500– 800. 21 We must assume several dozen such villages in Caesarea’s territory. Among the small Late Antique villages within it, an important excavation project investigated Tel Tanninim, or Mound of the Crocodiles, on the coast 4 km. north of Caesarea. This settleScholars debate the existence of formal markets in villages; see Decker 2009: 232. Dar, Tepper, and Safrai 1985: 37–38, identified a market building within the village. More generally, Lapin 2001: 133–37. 20 On the villages and their economy see Decker 2009: 65–79, and Bagnall 1990: 110–47. 21 Dar 1993: 1314–16; Dar, Tepper, and Safrai 1985. Dar asserts a decline and virtual abandonment of this site in the Byzantine period, a consequence of the fifth-sixth century Samaritan rebellions (1993 1315– 16; Dar, Tepper, and Safrai 1985: 138). The stratigraphic evidence, however, is slight and not convincing. 19

302

KENNETH G. HOLUM

ment boasted tanks for raising fresh-water fish, two small bath houses, and presumably dwellings for the inhabitants, all grouped around a basilical church. 22 The population numbered, perhaps, in the low hundreds. Some would have tended the fishtanks, while others presumably fished in the sea, and farmers dwelling in the settlement tilled nearby fields and garden plots among the marshes near the neighboring twin outlets of the Crocodile River. Not far away Ḥorvat ʿAqav, the only carefully excavated farmstead in Caesarea’s territory, lay atop the ridge of Ramat ha-Nadiv. This establishment, by no means luxurious, had stables, storage facilities, a wine press, and second-story dwelling quarters for the estate manager and workers. 23 Altogether, we might think of perhaps 20–30 individuals occupying the farmstead and working this estate. These few examples represent for us several hundred similar settlements within Caesarea’s territory, many of them – though by no means the majority – the property of landlords who dwelled in Caesarea or its immediate environs. 24 That the landlord-tenant relationship linked Caesarea tightly with its territory emerges from the same Procopius passage cited above, a critical text documenting the system of agricultural exploitation and the links between city and countryside. The historian, who on this point has no axe to grind, declares that the Emperor Justinian’s soldiers caused serious harm to the Christian landowners of Caesarea when they suppressed an uprising of Samaritans in 529/30. “Ten times ten thousand” rebels fell, Procopius asserts, comprising all of the peasant farmers (geōrgoi) who had who tilled the land, and thus Caesarea’s fields were left derelict (Secret History Stieglitz et al. 2006. To my mind, the church and the baths indicate a village-size population, perhaps relatively dispersed. 23 Hirschfeld et al. 2000: 725) estimates 10–14 occupying the Byzantine dwelling, but this seems to me low for the number needed to cultivate vineyards requiring a dedicated winepress. 24 Bagnall 1990: 148–49, asserts that two-thirds to three-quarters of village land in Egypt was owned by villagers, some of them quite prosperous. I see no reason why Caesarea should have been different, except that in Palestine farmers, whether proprietors or not, apparently lived more frequently than in Egypt in hamlets or isolated farmsteads. 22

THE ECONOMY OF CAESAREA PALAESTINAE

303

11.27–30). Archaeological evidence confirms that Samaritans did till fields and did occupy villages in this landscape, including perhaps Umm Riḥan, 25 and many of them will indeed have owed both rents and taxes to Caesarea landlords, but Procopius surely exaggerates Samaritan numbers and the war’s impact. On another point also we may take the historian at his word. Although the landlords henceforth received no rents, he continues, Justinian still required them to return the full amount of taxes due from the liquidated Samaritan peasants. The text thus displays lucidly that the state still tasked Caesarea’s elite with tax collection in the earlier sixth century, and indeed the ties that linked Caesarea with its territory in one economic unit. That some Caesarea landlords owned huge tracts is the implication of another Procopius passage – written, I emphasize, by a native of the city who possessed detailed knowledge. We learn that a man of Caesarea’s elite named Evangelus, an advocate active in the governor’s court at Caesarea, had amassed a large fortune and was able to purchase an entire village, Porphyrion, north of Caesarea on the coast, for the huge sum of three kentenaria of gold, or 21,600 solidi (Secret History 30.18–19). Although this village lay outside of Caesarea’s territory, 26 the pattern of elite landholding must have applied within the territory as well. Hence Caesarea’s elite drew their wealth from agricultural estates and indeed entire villages – though perhaps villages smaller than Umm Riḥan.

Dar 2014 catalogues a number of Samaritan sites in the Carmel and Menasseh Hills region, most of which (Sindiana, H. Samara, Umm Zinat) apparently lay within Caesarea’s territory. 26 On Porphyreon see Tsafrir, Di Segni, and Green 1993: 204. PLRE 3A, 454, “Euangelus,” misidentifies the village in question, while Decker 2009 mistakenly asserts that Evangelus made his fortune in Constantinople. Procopius explicitly numbers him “among the rhetors in Caesarea [en kaisareiai].” Di Segni 1995: 577, interprets Porphyrion as a fishing village, but that is not necessary or even likely. A fishing village would not have cost so much! 25

304

KENNETH G. HOLUM

Figure 4. Caesarea, Areas CC, KK, and IAA area in Southwest Zone, north at right. Left to right are an urban mansion (or bath?), the Area KK warehouses, and the governor’s palace (praetorium). Drawing by Anna Iamim.

Archaeological discoveries inside the walls of Late Antique Caesarea correspond with this evidence and confirm that, although he exaggerated for invective purposes, Procopius nevertheless represented correctly the economic relationship between city and countryside. In the 1990s Joseph Patrich excavated a complex of six warehouses in Caesarea’s Southwest Zone, area KK, designed for storage of grain in subterranean grainbins as well as oil, wine, and other commodities (figs. 4–5). 27 These warehouses were relatively modest in dimensions and of light construction, in contrast with an evidently public or state warehouse discovered north of the Inner Harbor, and this along with clear division into six separate units Patrich 1996: 168–73 suggested earlier that these warehouses belonged to the state or to the church, but in 2011: 138, he accepted private ownership; also Patrich et al.1999: 75–81. 27

THE ECONOMY OF CAESAREA PALAESTINAE

305

indicate plural ownership, not by the state or church but by private individuals (Stabler et al. 2008: 1–8). Hence the warehouses are physical evidence of elite economic behavior. Caesarean landlords, like their counterparts elsewhere, employed these facilities as engines of profitability, knowing that prices for commodities extracted as rents from their estates, or purchased in the grain market, and stored up in these warehouses, would rise with the passage of time after the harvest – or when famine struck in consequence of drought or other insufficiency. 28 The warehouses also document with hard archaeological evidence the Finley model of a consumer city that extracted its wealth from the surrounding countryside. Their location in the city plan increases their evidential force, for these warehouses occupied an insula near the shore adjacent on the north to one of the urban mansions of the Caesarea elite, and to the south of this mansion, separated from it by another warehouse, lay another luxury dwelling, and on its south still another mansion, poorly preserved, it too equipped with grainbins (figs. 1, 4). 29 The arrangement of luxury dwellings, on a ridge where they would receive cooling landward breezes, interspersed with facilities for storage of marketable commodities, conveniently located with easy access to shipping, indicates to my mind mansions that were the See Holum 2011: 25, note 67, citing van den Ven 1953: 12–13, where a landlord (ktētor) stored up grain in his warehouse (apothēkē) until a shortage raised prices. This was typical elite behavior. See also Patlagean 1977: 80; Erdkamp 2005: 147–67; and Brown 2012: 14: “Not surprisingly, therefore, granaries emerge as the economic villains of the ancient world.” 29 This is area I of the IAA excavations conducted 1992–1994, published so far only in Porath 1996: 116–19l; 1998; 2000: 37*–39*, 42–45; 2008: 1660–61. For the distribution of warehouses, cf. the map in Patrich 1996: 147. Porath and his colleagues interpret the establishment adjacent to the area KK warehouses as a public bath, but the resemblance of this elegant building to an urban domus is much closer, as Patrich has pointed out 2011: 138. Furthermore, an inscription on a broken slab found within the building, CIIP 2: 1348, may refer to it as an oikos, certainly more appropriate for a mansion than for a bath or, as CIIP suggests, for a church. We await the final report. 28

306

KENNETH G. HOLUM

property of the warehouse owners, wealthy landlords, members of Caesarea’s elite.

Figure 5. Caesarea, Area KK warehouses, looking west. Photo by J. J. Gottlieb.

Indeed, now we may focus more intently upon the rhetor Evangelus as an exemplar of that elite. In Procopius’ account, he was a personage of “some distinction” (ouk asēmos) upon whom “a favorable breeze of good fortune had blown.” The text documents a man of relatively modest means amassing wealth through advocacy, 30 who then emplaced that wealth in land, always the most attractive strategy of the ancient investor. This too was elite behavior. 31 Equally instructive is the scale of the man’s new wealth, for in On the wealth of advocates see Liebeschuetz 2001: 50, 177. In the fourth century Libanius, advocate of Antioch, amassed many estates and huge wealth, having begun from a situation of some financial distress; see Liebeschuetz, 2001: 43–48. 31 Modifying often-quoted words of Peter Brown makes them à propos for Evangelus: “Patronage” – in this case advocacy in the lawcourts – “was the only way in which men whose careers lay on the fringe of the 30

THE ECONOMY OF CAESAREA PALAESTINAE

307

purchasing a village for 21,600 solidi this person invested a very large sum, apparently equal to one-fourth to one-third of the taxes collected annually in all of First Palestine, the province in which Caesarea was located. 32 Hence the emperor forced Evangelus to sell his village at a fraction of the cost, declaring (Procopius asserts) that since “he was only a rhetor it was unsuitable that he should be lord (kyrios) of such a village.” Still another Procopius text points in the same direction, toward the massive wealth of Caesarea’s elite. Attacking Justinian for his greed, Procopius reports that he seized the entire inheritance of an unnamed woman, heiress to her husband Mamilianus, of “a highly distinguished family” (oikias epiphanous agan) of Caesarea, and heiress also to her father Anatolius of Ascalon, leaving the woman only a single solidus a day, “so that she would not be reduced to the ranks of the beggars” (Secret History 29.17–25). Here Procopius distorts not the amount but the facts about the emperor’s action, for it appears that Justinian was simply enforcing a law he had issued recently (Novel 38) ordaining that in such cases three-fourths of an elite estate would revert to the state. 33 We may thus suspect that 365 solidi per annum actually represented one-fourth of the yield of the two estates that the anonymous woman inherited, and that the late Mamilianus had therefore taken in annually roughly half of 1,460 solidi or 730 solidi. If we calculate that only six to nine solidi would keep a family of four or five for an entire year above the poverty-line, providing basic food, clothing, and shelter, 34 the woman’s 365 per annum was still a very large income, traditional landed aristocracy could gain access to the one source of wealth and prestige in the ancient world, to the land” (1971: 86). 32 Calculated as 9–12 kentenaria, 64,800 to 86,400 solidi, in Di Segni, Patrich, and Holum 2003: 282, n. 24; cf. Jones 1966: 1.462–63. 33 On this family and its fate see also Holum 1995: 621, 625; Laniado 2002: 59, 70. 34 My rough estimate is based on a variety of documented prices assembled by Jones 1966: 1.447–49, Patlagean 1977: 380–96, and Morrisson 1994. Cf. Kelly 2004: 66, 141. From a different perspective, Durliat 1990: 541–42, n. 151, interprets Justinian Digest 48.2.10 setting the property qualification for bearing witness in lawsuits at 50 solidi. Durliat estimates

308

KENNETH G. HOLUM

alleged by Procopius as sufficient to keep her from beggary, but actually enough for 40 families of four, while the estate of Mamilianus would have supported 80 families or more. Returning to Evangelus, if we suppose that he expected a 5% annual return on his investment in the village, a reasonable estimate for this period, his projected income would likely have been about 1,000 solidi, somewhat more than that of Mamilianus. 35 These two individuals, Evangelus and the late Mamilianus, possessing wealth on such a scale, most of it investments in land, represented the notables, a decreasingly numerous but increasingly wealthy elite in the time of Procopius and Justinian. Also among them were the Christian “landlords” (kyrioi) who suffered loss, as Procopius alleged, in the 529/30 Samaritan rebellion (HA 11.30). Let us suppose only 50 or so Caesarea families at the pinnacle of the elite, hence counting among the “notables,” 36 possessing properties across a spectrum of value but on the average (let us say) that this would be the value of 5–15 ha of cropland that would yield 15– 20 solidi per annum, sufficient for the annual needs of a peasant farmer and his family, hence a kind of poverty line. In my view, this calculation assumes an impossibly high annual return of 30–40 percent of capital. At 5 percent, the estimate that I adopt here, the return would be only 2 1/2 solidi per annum. 35 For comparison, the Apions of Oxyrhynchus, one of the richest families in the eastern provinces, extracted 18,000 solidi per annum from their properties in Egypt; see Laniado 2002: 157. See Duncan-Jones 1982: 33–59, for 5–6 % return on agricultural investment in first-century Italy, somewhat higher for vineland. 36 In the fourth century the curial elite of Timgad in Algeria numbered 282. See Chastagniol 1978: 22–39, 89–90; also Holum 2014: 64–69, and 2005: 108–9. Caesarea was a larger city than Timgad, but by the sixth century the numbers of the local elite had declined significantly, while the wealth of the notables, individual and aggregate, had increased. For numbers on wealth in the sixth century see e.g. Laniado 2002: 154–60. For comparison, we know from land registers that at Hermopolis in Upper Egypt two centuries earlier the top 10% of landowners living within the city owned 78% of the land in the entire nome (territory). See Bagnall 1990: 68–70; 1992: 142.

THE ECONOMY OF CAESAREA PALAESTINAE

309

worth half that of the single village that Evangelus purchased, or 10,000 solidi per family. Their aggregate properties would thus have been worth somewhere near 250,000 solidi, and at a modest 5% return per annum the 25,000 solidi yield would have provided, at a subsistence level, for roughly 3,000–4,500 families of four or five, or approximately half the entire population of Caesarea within its fortification walls. 37 Obviously, these are rough estimates of scale, but even rough estimates make it clear that Caesarea’s elite families dominated its economy by means of the wealth extracted from their rural estates. In my opinion, this speculation alone makes it difficult not to understand Caesarea in Late Antiquity as a Finley-type consumer city. Moreover, the projected 50 or so families at the pinnacle of the elite represent only a fraction of Caesarea landowners, for several times 50 families of relatively modest wealth must likewise have lived in the city as rentiers, drawing rents from property in the territory. 38 Like Evangelus, and like their counterparts in other cities of the East in Late Antiquity, the elite of Caesarea much preferred investment in land, but the Southwest Zone warehouses indicate For comparison, and to confirm that my estimate is realistic, I refer to Bagnall 1990: 70, calculating that in the Hermopolite nome during the fourth century 88 percent of land held by landowners in both city and countryside produced an aggregate surplus beyond the owners’ subsistence needs sufficient to supply a landless population in both city and countryside of 80,000 persons. Similarly, Alston 2002: 343, calculates that ca. 350 the metropolite landlords of Hermopolis took home 18% of the total yield of the nome, sufficient to feed most of the city populace. 38 I have in mind of course, from the perspective of scale, a modest peristyle dwelling in Caesarea’s northwestern quarter where in 1993 archaeologists discovered a hoard of 99 solidi, closing date 395, concealed beneath a tessellated pavement. See Lampinen 1999; Evans 2006: 53–61. We do not, of course, know the source of the wealth represented – whether its owner was e.g. a soldier, a state official, or a rentier living from the proceeds of rural investment. The hoard does, however, suggest the surplus of a moderate income put away for safekeeping as a hedge against famine or other unforeseen disaster. 37

310

KENNETH G. HOLUM

that they adopted more than one strategy for creating wealth. Investment in cabotage, coastwise shipping, would have enhanced the long-term profitability of the warehouses by enabling shippers to bring the stored grain, wine, and oil onto markets where prices were higher, and long-distance shipping of commodities and even of more expensive products like locally-produced glass and textiles might also attract investment. Indeed, during his service abroad Procopius encountered a merchant in Syracuse who was a native of Caesarea and a “friend from childhood” (Wars 3.14), perhaps in fact the agent of a Caesarea notable who invested in trade. Typically, the economic behavior of Late Antique elites included moneylending both to friends and inferiors who needed cash and to small-scale entrepreneurs. 39 We may extract other kinds of elite behavior from the Caesarea evidence. The Greek inscriptions record numerous acts of munificence at Caesarea of the “storing up treasures in heaven” type, though none appears to have been on a large scale, or likely to have been the donation of a notable. 40 In other cities such benefactions of city notables are recorded in the range of 20–40 pounds of gold or 1,440–2,880 solidi. 41 Further, deriving steady profits from their investments, Caesarea’s notable families did indeed have the means to construct towering status by building elegant urban and suburban mansions, like the three excavated in the Southwest Zone, located along the shore adjacent to the warehouses described above, richly ornamented with marble columns, pavements, and wall revetments, figured opus sectile and mosaic wall and floor decoration. One mansion (fig. 4) featured a spacious private bath suite and two See Liebeschuetz 1972: 41–51, finding that the “characteristic economic activities” of the upper class in the fourth century were “land owning, corn-selling and money-lending,” while accepting some elite engagement in trading. Also Bagnall 1990: 37, 73–78, 129, documenting elite Egyptians as moneylenders, shipowners, and proprietors of production facilities (pottery, oil, grain mills) both in the city and on country estates. These patterns likely persisted in the sixth century. 40 CIIP 2: 1139–43 (Jewish), 1148, 1150–52, 1168 (an orphanage), 1179 (of a deacon). 41 See Laniado 2002: 158, for references. 39

THE ECONOMY OF CAESAREA PALAESTINAE

311

semicircular stibadium-type dining rooms, in which the master and his elite guests ate reclining around a sigma-shaped dining table, displaying their wealth and status in a theater-like setting (Ellis 1997) while in another mansion not far to the south the archaeologists discovered a triconch (triple-stibadium) dining suite and a luxurious marble-clad pergola on a lower level near the shore. 42 In a mosaic pavement the owner of suburban villa (proasteion) situated to the city’s northeast, asserted proudly the sources of his wealth, quoted at the beginning of this essay: From their harvests of grain, wine, and oil were they enriched. The wealth of which he boasted came from his agricultural estates! Nearby in the villa of the “Birds” mosaic, the archaeologists recovered an elegant and costly sigma table from a stibadium dining room, decorated with gilded glass, recently exhibited in the Metropolitan Museum, New York (Porath, GorenRosen, and Neguer 2005–2006). Such domestic arrangements also speak to the scale of economic activity, financed mainly by income from rural properties, and tend to render plausible another side of the Finley approach, that the economy of Caesarea was embedded in a moral system, the quest of the elite for status. More generally, wealth that the elite extracted from the countryside showered solidi, droplets of gold, upon the inhabitants of Caesarea. The cost of constructing and decorating elite domiciles, of feeding the elite sumptuous banquets, and of clothing them in expensive garments of silk – not only the notables on whom I focus here but also those who positioned below them on a declining scale of elegance and wealth – obviously provided a large part of the purchasing power that worked its way within the city down through the social levels, into the building trades, of course, and into the service and manufacturing sectors. Hence, from the building sector, the inscriptions document a “worker in marble” and (perhaps) a “cutter of wood,” 43 and in one of the mansions of the Southwest Zone the archaeologists discovered the workshop where opus sectile wall decoration was apparently crafted and stored awaiting installation (Gendelman and Gersht 2010). We do not know See the reports of Porath cited above, n. 29. CIIP 2: 1495, 1536. This are epitaphs probably affixed to unpretentious built tombs. 42 43

312

KENNETH G. HOLUM

whether the craftsmen were itinerants or domiciled permanently in the city, but certainly the mansion owner, one of the elite landlords, provided the materials and paid for their labor, in coin spent at least in part within the city. In the services sector we encounter, in the inscriptions from Late Antiquity, a doctor, and an archiperipolos, perhaps chief of the night watchmen. 44 The anonymous fourthcentury Expositio totius mundi et terrarum (22) extolls Caesarea’s pantomimes, theatrical performers famous at home and abroad. Representing the transport sector is an acclamation wishing good luck for “porters” (ōmophoroi) who carried goods on their shoulders, presumably within the city, but many more must have conducted goods from city to countryside and countryside to city employing carts and pack animals, and occasionally wagons for heavier loads (CIIP 2: 1417). These few examples represent a much larger population in the service sector, domiciled within Caesarea, that were fed, clothed, and housed at the expense of other productive sectors, in large part – though we cannot estimate the scale – from profits extracted from the countryside. Also evident in the sources is the manufacturing sector. The Expositio numbers Caesarea among several cities producing purpledyed fabric of superior quality. 45 Mentioned in an epitaph, a babylonarios perhaps crafted Babylonia-style shoes or garments. 46 Another inscription mentions a tax on “perfumers,” who probably both manufactured and retailed their products. From a number of texts in the Jerusalem Talmud, Lee Levine extracted evidence that at Caesarea craftsmen engaged in glass-blowing, silk-processing, and manufacture of pottery and of a special type of chair or bed. 47 We should be cautious, however, in drawing conclusions about Caesarea itself as an industrial center, because in every case, except the babylonarios and perhaps the perfumers, production may equally CIIP 2: 1446, 1534, both epitaphs. For a much fuller reconstruction of a city’s economy drawn from epitaphs, see Trombley 1987. 45 C. 31 with Rougé’s comment ad loc., p. 252. 46 CIIP 2: 1513 now in Haifa, perhaps from Caesarea. 47 Levine 1975: 52–53, with reference (among other passages) to Shabbat vii.4.10d, Bava Mezia iv.2.9c, Bava Qama ii.5.3a (cf. Levine, p. 185, n. 75), Moʾed Katan ii.2.81b, Qiddushin i.4.60b. 44

THE ECONOMY OF CAESAREA PALAESTINAE

313

have been located in the villages, hamlets, and even at isolated farmsites. Products coming to market in Caesarea would likely have been identified with the city, the mistress of the territory. Indeed, Julian of Ascalon, in his sixth-century treatise on architecture, giving rules for settling legal disputes about use of urban space, banishes pottery kilns from cities because of the heats they generated, yet curiously one of the few industrial enterprises we know of within Caesarea’s city walls was an apparent potter’s workshop identified by the discovery in the governor’s praetorium of a large quantity of ceramic wasters as well as numerous molds for lamps, ceramic vessels, and animal figurines. 48

Figure 6. Caesarea, Governor’s Palace (praetorium) in Area CC, looking southwest. The Revenue Office is in the center foreground. Drawing by I. Rabinovitz, courtesy of J. Patrich.

Saliou 1996: 36. Patrich et al., 2008: 296–97 (Patrich and S. Pinkas), with references. As the authors point out, the site of production is unknown, but we may assume it was inside the city. Cf. Adan-Bayewitz 1995. 48

314

KENNETH G. HOLUM

Obviously the city domiciled an unknown but large population engaged in commerce, whose productivity consisted in essence of buying cheap and selling dear. The epitaphs record three vintners who must have traded within the city, as well as a husband and wife team identified as “shopkeepers” (kapēloi) (CIIP 2: 1491, 1495, 1530, 1563). Vintners and shopkeepers like these may have plied their trades in a row of shops, one of them a tavern, uncovered at ground level on the north side of the street in the Southwest Zone labeled Decumanus S2 (fig. 4). 49 The attested babylonarios and the perfumers may have sold goods in similar streetside shops where they produced them, but Caesarea also boasted a central marketplace (agora) of traditional type that the archaeologists have not yet located. 50 In a city of this size we would expect one or more formal markets where the inhabitants could find necessities and villagers might come for goods not available in the villages. A nexus of land and sea routes, 51 and a city with a relatively large population, Caesarea also attracted commerce from abroad in goods both luxurious and mundane available for purchase in the markets at a markup that meant economic gain. Vast though unquantifiable numbers of locally-made jars contained wine and oil brought in from estates in the territory, destined – after storage, perhaps, in the warehouses – for local sale or for export. The more mundane imported goods included tableware from Africa, Cyprus, or western Asia, frequent in the archaeologists’ ceramic inventories. 52 As for personnel, a Patrich et al. 1990: 87–90. The tavern, fitted with reclining benches, is vault 13. For identification as a tavern, see Patrich et al., 2008: 8. 50 The clearest reference is Choricius Laudatio Stephani et Aratii 39, eds R. Foerster and E. Richtsteig (Leipzig: Teubner, 1929), extolling the governor Stephanus for calming a panic in the agora induced by a fire. A crowd filled the marketplace one day toward evening, but there is no mention of buying and selling. On urban markets in general see Lapin 2001: 130–33. 51 See Roll 1996 on Caesarea in the network of highways, and on the city’s harbor Reinhardt and Raban 2008. 52 Johnson 2008 is the most recent and satisfying large inventory. For the challenge of quantifying ceramic evidence see Blakely, “Toward the Study of Economics.” 49

THE ECONOMY OF CAESAREA PALAESTINAE

315

merchant (emporos) attested in a fragmentary epitaph represents the commercial sector of the city’s economy, and it appears that Caesarea’s rabbis, having business connections abroad, traded in luxury goods that included silks and women’s jewelry. 53 Again, we have no way to estimate scale, but clearly the luxury markets depended on the wealth of the elite, drawn mainly from their rural estates. So far all seems to be in order at Caesarea for the Finley model, but two other wealth-producing sectors tended to distance the city from this model, and to make Caesarea’s overall economy more complex. The first was the government sector, located within the city and clearly an internal producer of significant wealth – or, perhaps more accurately, a sector that drew wealth from the entire province. Adjacent on the north to the warehouses in the Southwest Zone (figs. 4–6) lay the praetorium of the provincial governor consisting of the audience hall or lawcourt surrounded by various offices, among them in a wing on the northeast what we call the “imperial revenue office” identified as such by a series of mosaic inscriptions set into its pavements. These rooms housed the secretaries and their supervisors whose duty was to see to it that the full amount of taxes due was paid into the imperial treasury, not only from Caesarea but from cities in the entire province of First Palestine. 54 One might, for example, have found registers of all the land owned throughout the cities’ territories, even in the most remote hamlets and farms, which could be used for checking off tax receipts. 55 Like the Procopius passage, these archaeological remains thus testify to continuity between city and countryside that Horden and Purcell had in mind. Furthermore, the government establishment headquartered in the praetorium not only accounted for and disbursed tax revenues but attracted hundreds of litigants to the city annually to sue or be sued before the governor in his audience CIIP 2: 1608; Levine 1975: 54–56, citing the Jerusalem Talmud Bava Mezia iv.9.9d on the commercial activity of Rabbi Abbahu (died early fourth century). 54 Patrich 2014 is the most recent study. The inscriptions are CIIP 2: 1334–42. 55 See Kelly 2004: 117–19, for a lucid evaluation of the Egyptian land registers. 53

316

KENNETH G. HOLUM

hall. They needed to be housed and fed. The entire establishment numbered 200 or more officiales who handled the administration of justice as well as keeping account of taxes and disbursements. In an edict of 536, promoting the governor of Palestine (Novels 103.1), Emperor Justinian allocated 22 pounds of gold per annum, 1,584 solidi, for the salaries of the governor and his staff, which, divided equally, half for the governor and half for a staff of 200, would have come to a paltry 4 solidi or so for each officialis. 56 But salary was just a fraction of staff income. In the 1990s we found, inscribed on a stone slab once displayed on an interior wall of the praetorium, an edict of the praetorian prefects that specified the fees or tips (sportulae) that these persons charged for legal proceedings brought before the governor in the adjacent audience hall (Di Segni, Patrich, and Holum 2003; Kelly 2004: 138–45). It appears that the court cost to the plaintiff or defendant for a single lawsuit averaged 5 solidi, most of which went to increasing the earnings of the secretaries, law clerks, accountants, messengers, and attendants who populated the governor’s staff. An annual caseload of 500 lawsuits might have brought in another 3,000–5,000 solidi in fees, nor was this the bulk of the wealth that the governor’s presence created. If Evangelus put together much of his vast wealth by working in the courts, as Procopius indicates, he and his advocate colleagues must have extracted further thousands of solidi annually from litigants, wealthy and not so wealthy, who arrived in Caesarea from throughout Palestine. Hence the governor’s establishment generated significant wealth, perhaps as much as 15,000 solidi per annum, but this was only a fraction of the projected 25,000 solidi that 50 notable families extracted from their estates, not to mention the incomes of hundreds of smaller landowners resident in the city. Furthermore, of course, the litigants had likewise by and large extracted their riches from the land. Secondly, there was the church, not present at all in Finley’s model, which for understanding Caesarea’s economy obviously has its weaknesses. Recently I have tried to demonstrate that ca. 500 an Hendy 1985: 178–80, for governors’ salaries of 700–800 solid in Justinian’s legislation, and Jones 1966: 2.593, for the number of officiales, assuming Caesarea’s officium to be half that of the proconsul of Africa. 56

THE ECONOMY OF CAESAREA PALAESTINAE

317

unknown bishop promoted a large-scale project to restore the deteriorated breakwaters of Caesarea’s famed harbor in order to attract to Caesarea’s quays ships that might otherwise offload their cargoes somewhere along the coast (fig. 7) (Holum 2014). The bishop enlisted Emperor Anastasius, who approved using funds from the treasury, and he undertook also to construct above the harbor to the east the magnificent octagonal martyr church that we excavated in the 1990’s. The bishop’s motivation, as I see it, was twofold. First, he expected that Christian pilgrims arriving by ship would mount the staircase that joined harbor and church and profit from the spiritual riches of contact with a powerful saint. Second – and by no means contradicting his first purpose – he intended that the church profit financially from the commerce that the ships brought with them – traffic that led in short order to the development of a commercial quarter on the quay between harbor and church equipped with workshops, warehouses, market stalls, and even commercial bathhouses (dēmosia). Thus the bishop would enhance his ability to bring succor to the poor. In Caesarea as elsewhere we witness the church developing into an economic powerhouse with its own extensive money-making properties, constituting its ousia or endowment, estates in the countryside as well as commercial baths, workshops, and harbor facilities in the city. Again, the wealthproducing capacities of the church bridged the divide between city and countryside that Finley alleged. We have no evidence for the scale of church properties, but surely they increased sharply during the fifth and sixth centuries as Caesareans stored up “treasures in heaven.”

318

KENNETH G. HOLUM

Figure 7. Caesarea, Octagonal Church in Area TP, harbor quays and shops in Area I, looking east. Drawing by Anna Iamim.

Thus in my view the evidence from Caesarea – of which there is much more than I have been able to present here – adequately confirms the Finley model of a consumer city that was parasitic on its countryside; of an economy much unlike that of a modern city that might well be labeled “primitive,” if one wishes to use a culturally biased descriptor; of an economy, moreover, that was certainly embedded in the elite quest for status. Nevertheless, viewed from the perspectives of structure and scale, the Caesarea economy, as much as the economy of Aila, clearly departs from the model in significant respects. Manufacturing and commerce flourished in both city and countryside, Caesarea’s government sector had a larger than usual economic impact, and more and more the church inserted itself as a major economic player. Thus the Caesarea evidence reinforces the advisability, where possible, of examining each Mediterranean-type city on its own terms, the Ailas as well as the Caesareas. As Paul Veyne put it, “… every single city of the ancient world was unlike any other, and there were even towns, Aquileia and Palmyra, which were commercial centres (Veyne 1990: 51.)” We may well number Aila among the latter, but in my view certainly not Caesarea. Nevertheless, Tom Parker’s work at Aila has rewarded us with a richness and variety of data, and has contributed much to a more profound understanding of the ancient Mediterranean economy.

THE ECONOMY OF CAESAREA PALAESTINAE

REFERENCES

319

Adan-Bayewits, D. (1995). A Lamp Mould from Sepphoris and the Location of Workshops for Lamp and Common Pottery Manufacture in Northern Palestine. In: J. H. Humphrey, ed., The Roman and Byzantine Near East: Some Recent Archaeological Research, JRA Supplementary Series, no. 14. Ann Arbor: MI, pp. 177–82. Alston, R. (2002). The City in Roman and Byzantine Egypt. London and New York: Routledge. Bagnall, R. (1990). Egypt in Late Antiquity. Princeton: Princeton University Press. Bagnall, R. (1992). Landholding in Late Roman Egypt: The Distribution of Wealth. JRS 82, 128–49. Beyer, G. (1936). Das Gebiet der Kreuzfahrerherrschaft Caesarea in Palästina siedlungs- und territorialgeschichtlich untersucht. ZDPV 59, 1–91. Blakely, J. A. (1996). Toward the Study of Economics at Caesarea Maritima. In: A. Raban and K. Holum, eds., Caesarea Maritima, Retrospective After 2,000 Years: A Symposium of Scholars Held at Caesarea, Israel, January 3–11, 1995, Leiden: E. J. Brill, pp. 327– 45. Broshi, M. (1980). The Population of Western Palestine in the Roman-Byzantine Period. BASOR 236, 1–10. Brown, P. (1971). The Rise and Function of the Holy Man in Late Antiquity. JRS 61, 80–101. Brown, P. (2012). Through the Eye of a Needle: Wealth, the Fall of Rome, and the Making of Christianity in the West 350–550 AD. Princeton: Princeton University Press. Chastagnol, A. (1978). L’album municipal de Timgad. Antiquitas, Series 3, volume 22. Bonn: Rudolf Habelt Verlag. Dar, S. (2014) Samaritan Communities on Mt. Carmel and Ramot Menashe. In: G. C. Bottini, L. Daniel Chrupcala, and J. Patrich, eds. Knowledge and Wisdom: Archaeological and Historical Essays in Honour of Leah Di Segni, Studium Biblicum Franciscanum, Collectio Maior 54. Milan: Edizioni Terra Sancta, pp. 283–91.

320

KENNETH G. HOLUM

Dar, S. (1993). The Survey of Western Samaria. NEAEHL, 4:1314–16 Dar, S, Tepper, Y., and Z. Safrai. (1985). Um-Riḥan: Ayarah miTekufat ha-Mishnah (“Umm Riḥan: Village from the Time of the Mishnah”). Tel Aviv: Ha-Kibutz ha-Me’uḥad.(Hebrew). Decker, M. (2009). Tilling the Hateful Earth: Agricultural Production and Trade in the Late Antique East. Oxford: Oxford University Press. Di Segni, L. (1999). Epigraphic Documentation on Buildings in the Provinces of Palaestina and Arabia, 4th–7th c. In: J. Humphrey, ed., The Roman and Byzantine Near East, vol. 2: Some Recent Archaeological Research. JRA Supplementary Series 31. Portsmouth, RI: Journal of Roman Archaeology, pp. 149–78 Di Segni, L. (1995). The Involvement of Local, Municipal, and Provinicial Authorities in Urban Building in Late Antique Palestine and Arabia. In: J. Humphrey, ed., The Roman and Byzantine Near East: Some Recent Archaeological Research. JRA Supplementary Series 14. Ann Arbor, MI: Journal of Roman Archaeology, pp. 312–32 Di Segni, Leah. (1995). Metropolis and Provincia in Byzantine Palestine. In: A. Raban and K. Holum, eds., Caesarea Maritima: A Retrospective after Two Millenia. Leiden: E. J. Brill., pp. 575–89. Di Segni, L, Patrich, Y., and K. Holum. (2003). A Schedule of Fees (Sportulae) for Official Services from Caesarea Maritima, Israel. ZPE 135, 273–300. Duncan-Jones, R. (1982) The Economy of the Roman Empire: Quantitative Studies. Cambridge and New York: Cambridge University Press. Duncan-Jones, R. (1990). Structure and Scale in the Roman Economy. Cambridge and New York: Cambridge University Press. Durliat, J. (1990). De la ville antique à la ville byzantine: le problème des subsistances. Collection de l’École Française de Rome, no. 136. Rome: l’École Française de Rome. Ellis, S. P. (1997). Late Antique Dining: Architecture, Furnishings, and Behaviour. In: R. Laurence and

THE ECONOMY OF CAESAREA PALAESTINAE

321

A. Wallace-Hadrill, eds., Domestic Space in the Roman World: Pompeii and Beyond. JRA Supplementary Series, no. 22. Portsmouth, RI: Journal of Roman Archaeology, pp. 41–51 Erdkamp, P. ( 2005). The Grain Market in the Roman Empire: A Social, Political and Economic Study. Cambridge and New York: Cambridge University Press. Erdkamp, P. (2012). Urbanism. In: W. Scheidel, ed., The Cambridge Companion to the Roman Economy, Cambridge and New York: Cambridge University Press, pp. 241–65. Evans, J. D. (2006). The Coins and the Hellenistic, Roman, and Byzantine Economy of Caesarea. In: R. J. Bull and O. J. Storvick., eds., The Joint Expedition to Caesarea Maritima: Excavation Reports, volume 6. Boston: The American Schools of Oriental Research. Rougé, Jean. (1966), Expositio totius mundi et terrarium. Sources chrétiennes, no. 124. Paris: Les Éditions du Cerf. Finley, M. I. (1999). The Ancient Economy. Updated with foreword by Ian Morris. Berkeley and Los Angeles: University of California Press. Gadot, Y., and Tepper, Y. (2009). Archaeological Survey of Israel: Regavim (49). Jerusalem: Israel Antiquities Authority. Available at: http://www.antiquities.org.il/survey/new/default_en.aspx. Gendelman, P. and Gersht, R. (2010) Crafts and Craftsmen in Roman and Byzantine Caesarea. Michmanim 22, pp. 27–48, 3*-34* (Hebrew with English abstract). Hendy, M. (1985). Studies in the Byzantine Monetary Economy c. 300– 1450. Cambridge and New York: Cambridge University Press. Hirschfeld, Y. (1997). Farms and Villages in Byzantine Palestine. Dumbarton Oaks Papers 51, 33–71. Hirschfeld, Y. et al. (2000). Ramat Hanadiv Excavations: Final Report of the 1984–1998 Seasons. Jerusalem: The Israel Exploration Society. Holum, K. G. (2011). Caesarea Palaestinae: A Paradigmatic Transition. In: K. G. Holum and H. Lapin, eds., Shaping the Middle East: Christians, Jews, and Muslims in an Age of Transition 400–800 C.E. CDL Press: Bethesda, MD, pp. 11–31. Holum, K. G. (2005). The Classical City in the Sixth Century: Survival and Transformation. In: M. Maas, ed., The Cambridge

322

KENNETH G. HOLUM

Companion to the Age of Justinian. Cambridge and New York: Cambridge University Press, 87–112. Holum, K. G. (2014). ‘Lord, Establish the Orthodox’: Two Neglected Late Greek Inscriptions and the Harbour Church at Caesarea Palaestinae. In: G.C. Bottini, L. D. Chrupcata, and J. Patrich, eds., Knowledge and Wisdom: Archaeological and Historical Essays in Honour of Leah Di Segni, Studium Biblicum Franciscanum, Collectio Maior, no. 54. Milan: Fondazione Terra Santa, Edizioni Terra Santa, pp. 87–112 Holum, K. G. (2014). Mediterranean Cities in the Fifth Century: Elites, Christianizing, and the Barbarian Influx. In: M. Maas, ed., Cambridge Guide to the Age of Attila. Cambridge and New York: Cambridge University Press, pp. 61–79 Holum, K. G. (1995). The Survival of the Bouleutic Class. In: A. Raban and K. Holum, eds., Caesarea Maritima, Retrospective After 2,000 Years: A Symposium of Scholars Held at Caesarea, Israel, January 3–11, 1995. Leiden: E. J. Brill, pp. 615–27. Horden, P., and N. Purcell. (2000). The Corrupting Sea: A Study of Mediterranean History. Oxford and New York: Oxford University Press. Johnson, B. (2008). The Pottery. In Archaeological Excavations at Caesarea Maritima, Areas CC, KK, and NN, Final Reports, volume 1: The Objects. Jerusalem: Israel Exploration Society, pp. 14– 206 Jones, A. H. M. (1966). The Later Roman Empire 284–602: A Social, Economic, and Administrative Survey. Oxford: Basil Blackwell. Kelly, C. (2004). Ruling the Later Roman Empire. Cambridge, MA: Harvard University Press. Lampinen, P. (1999). The Gold Hoard of 4th-c. Solidi Found in 1993. In: K. G. Holum, A. Raban, and J. Patrich, eds., Caesarea Papers 2: Herod’s Temple, The Provincial Governor’s Praetorium and Granaries, the Later Harbor, a Gold Coin Hoard, and Other Studies. JRA Suplementary Series, no. 35. Portsmouth, RI: Journal of Roman Archaeology, pp. 369–388. Laniado, A. (2014). From Municipal Councillors to ‘Municipal Landowners’: Some Remarks on the Evolution of the Provinicial Elites in Early Byzantium, In: M. Meier and S. Patzold,

THE ECONOMY OF CAESAREA PALAESTINAE

323

eds. Chlodwigs Welt: Organisation von Herrschaft um 500. Stuttgart: Fritz Steiner, pp. 369–88. Laniado, A. (2002). Recherches sur les notables municipaux dans l’empire protobyzantin, Travaux et mémoires, Monographies 13, Paris: Association des Amis du Centre d’Histoire et Civilisation de Byzance, 2002. Lapin, H. (2001). Economy, Geography, and Provincial History in Later Roman Palestine. Tübingen: Mohr-Siebeck. Levine, Lee I. (1975). Caesarea Under Roman Rule. Leiden: E. J. Brill. Liebeschuetz, J. H. W. G. (1972). Antioch: City and Imperial Administration in the Later Roman Empire. Oxford: Oxford University Press. Liebeschuetz, J. H. W. G. (2001). The Decline and Fall of the Roman City. Oxford: Oxford University Press. Morrisson, C. (1994). Monnaie et prix à Byzance du Ve au VIIe siècle. In: C. Morrisson, ed. Monnaies et finances à Byzance: analyses, techniques. Aldershot: Ashgate, pp. 239–60. Ne’eman, Y. (1990). The Archaeological Survey of Israel: Map of Ma’anit (54). Jerusalem: Israel Antiquities Authority, 1990. Ne’eman, Y., Sender, S., and E. Oren. (2000). The Archaeological Survey of Israel: Map of Mikhmoret (52); Map of Ḥadera (53). Jerusalem: Israel Antiquities Authority. Olami, Y. (1981). The Archaeological Survey of Israel: Daliya Map (31). Jerusalem: Archaeological Survey of Israel. Olami, Y., S. Sender, and E. Oren. (2005). The Archaeological Survey of Israel: Map of Binyamina (48). Jerusalem: Israel Antiquities Authority. Orni, E., and E. Efrat. (1973). Geography of Israel. 3rd. ed. Philadelphia: The Jewish Publication Society. Parker, S. T., and A. Smith. (2014). The Roman Aqaba Project, Final Report. Vol. I: The Regional Environment and the Regional Survey. American Schools of Oriental Research Archaeological Reports 19. Boston: American Schools of Oriental Research. Patlagean, E. (1977). Pauvreté économique et pauvreté sociale à Byzance, 4e7e siècles. Paris: Mouton.

324

KENNETH G. HOLUM

Patrich, J. (1996). Warehouses and Granaries in Caesarea Maritima. In: A. Raban and K. G. Holum, eds. Caesarea Maritima: A Retrospective After Two Millenia. Leiden: E. J. Brill, pp. 146–176. Patrich, J. (2011). Several Aspects of Commerce and Economy in Late Antique Caesarea. In: J. Patrich ed. Studies in the Archaeology and History of Caesarea Maritima: Caput Iudaeae, Metropolis Palaestinae. Leiden and Boston: Brill, pp. 117–39. Patrich, J. (2014). The Architectural Evolution of the Late Antique Revenue Office at Caesarea Maritima. In: G. C. Bottini, L. Daniel Chrupcala, and J. Patrich, eds. Knowledge and Wisdom: Archaeological and Historical Essays in Honour of Leah Di Segni, Studium Biblicum Franciscanum, Collectio Maior 54. Milan: Edizioni Terra Sancta.pp. 65–87. Patrich, J. et al. (1999). The Warehouse Complex and Governor’s Palace (Areas KK, CC, and NN, May 1993–December 1995). In: K. G. Holum, A. Raban, and J. Patrich, eds. Caesarea Papers 2: Herod’s Temple, the Provincial Governor’s Praetorium and Granaries, the Later Harbor, a Gold Coin Hoard, and Other Studies. JRA Supplementary Series, no. 35. Portsmouth, RI: Journal of Roman Archaeology, pp. 70–107. Patrich, J. et al. (2008). Archaeological Excavations at Caesarea Maritima, Areas CC, KK, and NN, Final Reports, vol. 1: The Objects. Jerusalem: Israel Exploration Society. Porath, Y. (1993). Herod’s Circus and Related Buildings, Part 1: Architecture and Stratigraphy. Volume 1 of The Israel Excavation Authority Excavation Project at Caesarea 1992–1998. Jerusalem: Israel Antiquities Authority. Porath, Y. (1996). The Evolution of the Urban Plan in Caesarea’s Southwest Zone: New Evidence from the Caesarea Excavations. In: A. Raban and K. G. Holum, eds. Caesarea Maritima: A Retrospective After Two Millenia, Leiden: E. J. Brill, pp. 105– 20. Porath, Y. (1998). Expedition of the Antiquities Authority. ESI 105, pp. 39*–45*. Porath, Y. (2000). Caesarea 1994–1999. ESI 112, pp. 34*–40*, 38– 46 (Hebrew). Porath, Y. (2008). The Israel Antiquities Authority Excavations. NEAEHL 5:1656–65.

THE ECONOMY OF CAESAREA PALAESTINAE

325

Porath, Y., Dar, S., and S. Apelba’um. (1985). Kadmoniyot ‘Emeq Hefer (“Antiquities of Emek Hefer”). Tel Aviv: Ha-Kibutz haMe’uḥad. (Hebrew). Porath, Y, Golri-Rosen, Y., and J. Neguer. (2005–2006). The ‘Birds Mosaic’ Mansion in the Suburbs of Caesarea Maritima, Israel. Musiva & Sectilia 2/3, 171–89. Reinhardt, E. G., and A. Raban. (2008). Site Formation and Stratigraphic Development of Caesarea’s Ancient Harbor. In: K. G. Holum, J. A. Stabler, and E. Reinhardt, eds. Caesarea Reports and Studies: Excavations within the Old City and the Harbor 1996– 2007. BAR-IS 1784, Oxford: British Archaeological Reports, pp. 155–82. Roll, I. (1996). Roman Roads to Caesarea.Iin A. Raban and K. G. Holum, eds. Caesarea Maritima: A Retrospective After Two Millenia. Leiden: E. J. Brill., pp. 550–58. Safrai, Z. (1994). The Economy of Roman Palestine. London and New York: Routledge. Saliou, C. (1996). Le traité d’urbanisme de Julien d’Ascalon (VIe siècle). Ed. C. Saliou. Travaux et mémoires monographies 8. Paris: de Boccard. Scheidel, W. (2012). Approaching the Roman Economy. In: W. Scheidel, ed. The Cambridge Companion to the Roman Economy. Cambridge and New York: Cambridge University Press, pp. 1–21. Stabler, J. et al. (2008). The Warehouse Quarter (Area LL) and the Temple Platform (Area TP), 1996–2000 and 2002 Seasons. In: K. G. Holum, J. A. Stabler, and E. Reinhardt, eds. Caesarea Reports and Studies: Excavations within the Old City and the Harbor 1996–2007, BAR-IS 1784. Oxford: British Archaeological Reports, pp. 1–39. Stieglitz, R. et al. (2006). Tel Tanninim: Excavations at Krokodeilonpolis 1996–1999. American Schools of Oriental Research, Archaeological Reports 10. Boston: American Schools of Oriental Research. Trombley, F. (1987). Korykos in Cilicia Trachis: The Economy of a Small Coastal City in Late Antiquity (Saec. V-VI) – A Précis. The Ancient History Bulletin 1, 16–23.

326

KENNETH G. HOLUM

Tsafrir, Y., Di Segni, L., and J. Green. (1994). Tabula Imperii Romani, Iudaea-Palaestina: Eretz Israel in the Hellenistic, Roman, and Byzantine Periods. Jerusalem: The Israel Academy of Sciences. van den Ven, P. (ed). 1953. La légende de S. Spyridon évêque de Trimithonte. Louvain: Publications universitaires, Institut Orientaliste. Veyne, P. (1990). Bread and Circuses: Historical Sociology and Political Pluralism, abridged with introduction by O. Murray, trans. B. Pearce. London: Penguin. Whittow, M. (1990). Ruling the Late Roman and Early Byzantine City: A Continuous History. Past and Present 129, 3–29.

“SO DEARLY DO WE PAY FOR OUR LUXURY AND OUR WOMEN”: WOMEN AND THE MARGINS OF THE ROMAN WORLD ELIZABETH ANN POLLARD 1 It was in Tom’s far-reaching “Ancient World to 180 AD” class, during my very first semester as a freshman majoring in Chemical Engineering at North Carolina State University (in 1989), that I first realized that someone could follow their heart and study what one loves. Tom’s enthusiasm for antiquity and the excitement and allure of archaeology that he shared with his students hooked me immediately and I became a History major. Tom’s dedicated teaching and selfless mentoring, in the classroom and in my work as his research assistant, prepared me for success in a graduate program in ancient history at the University of Pennsylvania. While in my graduate program, Tom provided me with the opportunity to participate in two seasons of the Roman Aqaba Project --- as a trench supervisor (1994 and 1996) and pottery registrar (1996) --- where I gained deep respect for, and appropriate suspicion of, archaeological evidence and its interpretation. I began my graduate program at UPenn intending to investigate Christianity’s spread on the margins of empire in the Transjordan region; but my “feminist epiphany” in graduate school launched me in a different direction. Nevertheless, what I had learned from Tom about studying “boundaries” prepared me well to explore other liminal topics such as Roman women accused of witchcraft and Mediterranean spice trade with India. Thank you, Tom, for the opportunities you have provided me and for setting me on this career path. What I offer here in celebration of Tom’s career is an attempt to blend my inter1

327

328

ELIZABETH ANN POLLARD

Rome’s trade with the east was thriving by the end of the first century CE. 2 By that time, Romans were sending significant amounts of currency eastward in exchange for spices (such as pepper, cinnamon, myrrh, and nard) and jewels such as emeralds and pearls. 3 During this period of flourishing and arguably unbalanced trade, the natural historian Pliny the Elder moralized about the drain that imports from China, India, and Arabia were placing on the Roman est in misbehaving women (be they political, fierce, witchy, or whore-like) with Tom’s focus on frontier studies. 2 Charlesworth (1924 and 1951), Warmington (1928), Wheeler (1955), and Miller (1969) laid the groundwork for much of the scholarship that would follow. Sidebotham (1986 and 2011) and Young (2001), among others, have explored the economic policies related to Rome’s trade along the eastern frontier. Sources ranging from a seafarer’s guide to Red Sea trade (the oft invoked Periplus Maris Erythraei) to a customs document have been explored by Casson (1986 and 1989). Close studies of the evidence for Romano-Indian trade have been offered by Tomber (2005 and 2008), essays on coins, glass, and pottery in Begley and de Puma (1991) and in studies by various contributors in Cimino (1994) and de Romanis and Tchernia (1997). The meaning and significance of Roman trade with India has been explored by Parker (2002 and 2008) and Pollard (2009). Other elements of Rome’s eastern trade have been usefully explored by Thorley (1969 and 1971), Ferguson (1978), Raschke (1978), and Schmitthenner (1979). 3 For the specific goods available through trade with India, see Periplus 39, 49, 56 and 63; discussed in Miller 1969: 34–109 and Warmington 1928: 226–28 and 180–234; and another detailed list of goods from India mentioned in a later tariff recorded in Justinian’s Digest 39.4.16.7 and discussed by Parker 2002: 41–42. It is worth noting here briefly the debate over the use of the term luxury goods to refer to various high value spices that were heavily consumed in Roman cooking and religious ritual. See Woolf 1990: 51–55; Sidebotham 1986: 15 and 45 and 2011: 249–251 and Young 2001: 18. Whether technically classed as luxuria or more generally as high value prestige goods, the implicit rationale behind the feminizing of the consumption of such spices and silks demands investigation.

WOMEN AND THE MARGINS OF THE ROMAN WORLD

329

economy. 4 He associated this drain directly with women’s consumption of luxury goods, when he opined of the deficit, “so dearly [it is that] we pay for our luxury and our women” (Pliny, Natural History 12.41). The biographer Plutarch, writing a generation after Pliny about the funeral of the vicious dictator Sulla more than a century and a half before, picked up on a similar thread of feminine excess and luxury trade, describing that “women contributed such a vast quantity of spices for [Sulla’s funeral], that, apart from what was carried on two hundred and ten litters, a large image of Sulla himself… was molded out of costly frankincense and cinnamon” (Plutarch, Sulla, 38.2). At his wife Poppaea’s funeral, the grieving Nero was said to have burned more incense than Arabia could produce in a year (Pliny, Natural History 12.83). And apart from those spices burned extravagantly in her honor, Poppaea was said to have been embalmed with spices in the way of foreign kings (Tacitus, Annals 16.6). From Pliny opining on women and luxury and reporting on the extravagant use of spices in Nero’s mourning of Poppaea to Plutarch connecting women’s extreme mourning and eastern spice trade, a rhetoric emerged that linked women’s excess with luxury and long-distance trade. From the arguable beginnings of Roman imperialism in the third century BCE through the late Roman empire, Roman literature and historical narrative articulated permeable, contested, and threatened boundaries and the trade and products that crossed these boundaries as associated with uncontrollable women and womanly vice. 5 This contribution explores the epistemology underThe accuracy and meaning of Pliny’s figures – 100 million sesterces a year drained from the economy toward eastern trade (Natural History 12.41.84) and 50 million a year to India alone (Natural History 6.26.101) – is problematized by Sidebotham 1986: 38–45 and Young 2001: 203–204. 4

For the connection of Roman luxuria with goods from the East, see Parker 2002: 45 and Young 2001: 205–07. For the relationship between “odor and power,” and the idea that the foreign source of all scent ingredients meant that they were always good for rhetoric “connecting the decline of Roman morality with foreign ways,” see Potter 1999: 177. 5

330

ELIZABETH ANN POLLARD

girding this feminizing of long-distance trade and luxury goods in the Roman world. Why blame women when there are so many other suitable candidates for culpability – among whom we might number the developing phenomenon of the emperor and his retinue, Senators inappropriately dirtying their hands in business dealings, generals, merchants, politicians, provincials attempting to curry favor in the capital, or even the increasing numbers of foreigners living in Rome? While some have attributed the Roman linkage of women with luxury (and vice versa) to a general misogyny, this chapter suggests that the Roman medical and philosophical understanding of women as permeable, leaky, and difficult to contain – much like the ever-expanding limits of empire – contributed to this association of long-distance trade and its products as feminized.

CONTAINING WOMEN WITH SUMPTUARY LAWS

One starting place for an examination of Roman linking of luxury with women who know no bounds are some of the earliest legal formations of this association, namely the sumptuary laws of the mid-Republic and the texts that recount them. Sumptuary laws sought to minimize extravagant shows of wealth by aristocrats, especially extravagance related to food, but also with respect to clothing and other luxury goods (Rosivach 2006: 1). A quick review of legislation that falls within this category demonstrates these laws’ arguable goal of balancing male aristocratic power. The lex Orchia (181 BCE), proposed by the plebeian tribune Gaius Orchius, limited the number of guests at entertainments (Macrobius, Saturnalia Gorman and Gorman (2014: 422–23) note that “[w]ith Rome’s expansion and conquest of the Mediterranean world came an influx of wealth and opulence that had a debilitating effect on the manhood and moral fiber of the Roman people [sic].” Their analysis points to two outcomes of luxury in the Roman mindset: 1) that which “leads to softness and the corruption of the military spirit” and 2) that which can drive one to violence in an attempt to get the resources required to satisfy one’s need for selfindulgence. While Gorman and Gorman emphasize the latter, it is the former – luxury as softening/feminizing – with which this chapter is concerned.

WOMEN AND THE MARGINS OF THE ROMAN WORLD

331

3.17.2). The lex Fannia (161 BCE), proposed by the year’s consuls, limited the amount of money that could be spent on entertainment (Macrobius, Saturnalia 3.17.4; Aulus Gellius, Noctes Atticae 2.24.2–6; and Rosivach 2006: 6, for a summary of its provisions). The lex Didia in 143 BCE clarified that the provisions of the lex Fannia applied to all of Italy and made attendees, in addition to hosts, liable to its penalties (Macrobius, Saturnalia 3.17.6). The lex Licinia (ca. 100 BCE), limited per capita food consumption on certain market days as well as on the Kalends and the Nones of each month (Macrobius, Saturnalia 3.17.9). The Sullan lex Cornelia (81 BCE) rearticulated the much-ignored lex Fannia and its related laws and extended the legislation to encompass extravagant funerals. That a funeral effigy of Sulla was molded from women’s contributions of frankincense and cinnamon certainly offers a whiff of irony to his attempt to limit such excess (Plutarch, Sulla 38.2). Julius Caesar was so vexed that these sumptuary laws were roundly ignored by his contemporaries that he enacted his lex Julia that placed officers in markets to seize foods that were in violation of the laws and to raid banquets for infractions (Suetonius, Julius Caesar 43.2). To be sure, this abbreviated recounting of the provisions of leges sumptuariae demonstrates that they were not primarily targeted at women’s wealth or consumption thereof. These laws sought to maintain the delicate balance of rival male aristocrats’ political power and influence by limiting their ability to sway the masses with conspicuous displays of wealth. Yet the reporting on these laws subtly hints at a connection with outrageous women’s luxury. One of the main sources for the provisions of various sumptuary laws of the Republic is Macrobius’s Saturnalia, written in the fifth century CE. 6 It is striking that the chapter that recounts these laws culminates with Macrobius’s refusal to report on a sumptuary edict by Antony due to Antony’s relationship with the notoriously luxury-consuming Cleopatra. Instead of describing the provisions of Antony’s law, Macrobius delights in telling the tale of Cleopatra’s wager with Antony that she could consume 10 million sesterces in a single meal, a feat she acThe other main source that discusses the sumptuary laws is Aulus Gellius’s Noctes Atticae 2.24, composed in the mid second century CE. 6

332

ELIZABETH ANN POLLARD

complished by dissolving a huge pearl in vinegar and then drinking it down (Macrobius, Saturnalia 3.17.14–18). By Pliny’s reckoning of the purported trade imbalance with India about a century after Cleopatra’s consumption of this pearl (an Indian commodity), that was one fifth of a year’s alleged 50 million sesterces monetary drain to India in a single gulp! Cleopatra’s outrageous flouting of sumptuary laws aside, possibly the best-known sumptuary law – and the one most relevant for the argument at hand – is the lex Oppia, famous for the elaborate debate and women’s political involvement the occasion of its repeal instigated. The lex Oppia had been promulgated in the context of the devastating defeats that Rome suffered at the start of the second Punic war in the late third century BCE. Predating the sumptuary laws already discussed, the Oppian law, named for the tribune Gaius Oppius who proposed it, was passed ca. 215 BCE in the aftermath of Roman losses to Hannibal’s invading forces at Trebia (218), Lake Trasimene (217), and Cannae (216). Unlike the other sumptuary laws just discussed, this law specifically targeted women’s displays of wealth, outlining that “no woman might own more than half an ounce of gold nor wear a multi-colored dress, nor ride in a carriage in the city or in a town within a mile of it, unless there was a religious festival” (Livy, History of Rome, 34.1.3). In speaking against the proposed repeal of this law in 195 BCE, twenty years after its creation, Cato – that stodgy, “Carthago delenda est”-spouting, anti-Hellene defender of Roman morality – expressed shock and outrage that women had deigned to pour out into the streets to persuade men, not of their relation, to repeal the law. Cato described these lobbying matrons as shameless, uncontrollable, and unable to be kept indoors in their desire to have their access to luxuries restored. In the passage describing the debate about the law’s repeal, Livy repeatedly refers pejoratively to womanly luxury (luxuria muliebris) (34.4.6 and 34.6.2). 7 Bauman 1992: 11 highlights Livy’s pejorative use of the term muliebris in this passage, both the description of the event as consternatio muliebris (feminine riot) and the womanish lack of self-control (impotentia muliebris). Bauman stresses that the adjective muliebris, which is peppered throughout this entire episode, is rarely positive. 7

WOMEN AND THE MARGINS OF THE ROMAN WORLD

333

Writing during the reign of Augustus in the late first century Livy recounts Cato’s purported words in harsh rebuke of the protesting matrons. He writes that the matrons (matronae) were not able to be contained at the threshold (contineri limine poterant), the limen, by means of the authority (auctoritate), respect (verecundia), nor imperium (imperio) of their husbands (34.1.5). Cato goes on to proclaim: “now at home our freedom is conquered by female fury (impotentia muliebris)… because we have not contained the individuals [i.e. the women] (quia singulas non continuimus)” (34.2.2). He continues: “Give the reins to their unbridled nature (impotenti naturae) and this unmastered creature (indomito animali), and hope that they will put limits on their own freedom… if they are victorious now, what will they not attempt” (34.2.12)? Cato sums up what women want: “that there should be no limit to [their] expenses and luxury (ne ullus modus sumptibus ne luxuriae sit)” (34.3.9). Livy’s recounting of Cato’s disdain of women’s inability to resist luxuria comes at the beginning of Rome’s expansion across the Mediterranean. In the early second century, Rome by her defeat of Carthage in Spain and in North Africa had taken control of the western Mediterranean and had begun making inroads into the East. This first major connection of women with luxury in the debates about the lex Oppia appears at the beginning of Rome's open access to foreign markets. And this connection continues. For example, the poet Propertius, a rough contemporary of Livy, linked the cost of wooing a woman to the influx of eastern luxuries (Bowditch 2006). In Book III.13, the poet describes how goods from Tyre, Arabia, the Red Sea, and India have flooded Mediterranean markets and essentially fueled and inflamed the economy of desire. Propertius exclaims: “The road of luxury has been made too free (luxuriae nimium libera facta via est)!” This example from Propertius is typical of the discourse of luxury imported via the Red Sea that appears in Augustan poetry, especially elegy (el-Nowieemy 2013). Whether Livy’s depiction of Cato’s opposition to the repeal of the Oppian law or Propertius’s representative anxiety about the impact of free-flowing eastern luxuries on Roman moral and monetary exchange, Roman sources suggest that as a burgeoning impeBCE,

334

ELIZABETH ANN POLLARD

rialist Rome is pushing its territorial limits, uncontainable and insatiable women are pushing the limits of their fathers’, husbands’, and brothers’ imperium over them.

UNBOUND AMAZONS

The Amazons offer an example of a complex feminizing of the barbarian margins of the classical world. Extending far back into Greek history, Amazons fought on the side of Troy in the Trojan War (Proclus, Chrestomathia 2), harried Theseus in his legendary attempts to found Athens (Plutarch, Life of Theseus 26–27), figured in Herodotean narrative of the Greco-Persian wars (Herodotus, Histories 4.110–117), and appeared in the legends of Alexander the Great, both as hostages from the Median satrap (Arrian, Anabasis 7.13) and through his possible sexual dalliance with their Queen Thalestris (Plutarch, Diodorus and Curtius, discussed in Mayor 2014: 319–338). Amazons particularly fascinated Greeks and later Romans for their shunning of male society and their manly prowess in battle. Strabo summed up Roman fascination with them when he wrote: “As regards the Amazons, the same stories are told now as in early times, though they are marvelous and beyond belief…who could believe that an army of women or a city or a tribe could ever be organized without men” (Strabo 11.5.3). While some scholars have argued that Amazons were not “real” but rather a good way to think about “the other,” more recent scholarship, especially excavations of elaborate burials of women warriors in the Black Sea region have challenged this assertion. 8 For Amazons as a useful “other,” see Stewart 1995: 571–97 and Blok 1995. Typical for this point of view, Walcot writes, “Wherever the Amazons are located by the Greeks, whether it is somewhere along the Black Sea in the distant north-east, or in Libya in the furthest south, it is always beyond the confines of the civilized world. The Amazons exist outside the range of normal human experience” (Walcot 1984: 42). For a picture of “real” Amazons, see Shapiro 1983: 105–11, Mayor 2014, and Penrose (2017). For burials of Scythian warrior women, linked to the Amazons, see Davis-Kimball 2002 and Guliaev 2003: 112–25. 8

WOMEN AND THE MARGINS OF THE ROMAN WORLD

335

Amazonomachies decorated Greek temples, notably the frieze from the mid-fourth century BCE Mausoleum at Halicarnassus. 9 This artistic theme continues in Roman art, with a prominent Roman noblewoman, Eumachia of Pompeii, including Amazons on her tomb in the first century CE (Cooley and Cooley 2004: 141) and even an emperor, Claudius, picking up on the iconography of Achilles and Penthesilea in the depiction of his conquest of Britannia from the first-century CE Sebasteion in Aphrodisias in Asia Minor (Smith 1987: 115–17). Amazons make their appearance in republican Roman imperial narrative, as potential allies of Mithridates VI Eupator, King of Pontus on the Black Sea and the major challenger ca. 100 BCE to Roman hegemony in the East. When Pompey is sent to fight Mithridates, he is diverted in his pursuit by what Plutarch calls “barbarian” Albanians. Plutarch writes: In this battle it is said that there were also Amazons fighting on the side of the Barbarians, and that they came down from the mountains about the river Thermodon. For when the Romans were despoiling the Barbarians after the battle, they came upon Amazonian shields and buskins; but no body of a woman was seen (Plutarch, Life of Pompey 35).

Plutarch goes on to situate these Amazons in the Caucasus region. Appian, also writing in the early second century CE, counts Amazons among the hostages taken by Pompey, although in the passage describing them Appian admits to some confusion as to who exactly these women are (Mithridatic Wars 21.103). Adrienne Mayor has put this evidence together to suggest that Amazons were actually led in Pompey’s triumph celebrating his eastern campaigns in 61 BCE (Mayor 2010: 358). More recently, Mayor has argued that there are three categories of Amazons: 1) “real nomadic horsewomen archers of the steppes,” 2) “Amazon queens Hippolyte, Antiope, and Penthesilea and other Amazons of classical mythology,” and 3) “women warriors in non-Greek traditions from the Black Sea to China” (Mayor 2014: 31). For artistic representations of Amazons, see Lexikon Iconographicum Mythologia Classica, vol. I, s.v. “Amazones” and von Bothmer 1957. 9

336

ELIZABETH ANN POLLARD

What is most important for this chapter, however, is to highlight the physical geography of this tribe of women Mediterranean men imagined as free from male control. Located variously on the Thermodon or the Don River, or more generally from the Steppes, the Amazons were unbound warrior women who operated outside male patriarchies and who were situated directly on inner-Eurasian trade routes that linked the Mediterranean to the Black Sea and beyond, on the very routes which brought eastern “luxury” goods into Roman markets.

WITCHY WOMEN

Another example of this problematic feminizing of the boundaries and of luxury trade is the literary location of some of antiquity’s most notable witches at the far reaches of significant trade-routes. Connected with the route to sub-Saharan and Eastern trade is the drunken bawd-witch Meroe, from Apuleius’s mid-second century CE Metamorphoses. In fact, Meroe shares her name with the capital of Kush, a trading center that controlled trade between Rome and Central Africa. This powerful and royal innkeeper (potens et regina caupona), could do the standard tricks of witches including overturning the natural order, transforming men into beavers and frogs, and sealing the wombs of rival women. But listed near the top of the litany of her evil deeds is: “The fact that she makes men fall madly in love with her – not just local inhabitants but also Indians and both kinds of Ethiopians and even Antichthones (or Antipodeans)” (Apuleius, Metamorphoses 1.8). This witch’s boundless sexual relationships with exotic foreigners – including Indians and Ethiopians (in other words, African Kushites) and even those from the other end of the earth – is central to her deviance. Whether she travels to them for sexual rendezvous, or vice versa, Meroe “gets around,” either literally or figuratively. This memorable boundary-crossing witch highlights the connection of women with excess and trade routes. In origin much earlier than Meroe is Medea, of Greek tragedy, although she occupies a central place in Hellenistic and later Roman imagination, as well. Although much earlier epic accounts of Medea existed in antiquity, the major extant sources for Medea are Euripides, Apollonius of Rhodes, Seneca and Ovid. While Medea is perhaps bestremembered for her skill with pharmaka and her cruel murder of her two children as revenge against her husband Jason’s betrayal, it

WOMEN AND THE MARGINS OF THE ROMAN WORLD

337

is perhaps easy to forget that, like Meroe, Medea hailed from the limits of trade-contact in the north-east. Medea was brought by Jason into the Mediterranean world from Colchis, modern-day Georgia, on the eastern shores of the Black Sea. This was a region with which the Greeks had fostered consistent trade associations through colonization since at least as early as the eighth century BCE. In fact, some of her most efficacious magical ingredients, the protective ointment known as Promethean in particular, are listed as coming from the Caucasus region, collected in a shell from the Caspian Sea (Apollonius, Argonautica 3.838–67). Meroe (along a Nile-Red Sea corridor) and Medea (along an Aegean – Black Sea – Central Asia corridor), offer just two examples of articulating as suspicious, even witchy, feminized boundaries the limits of civilization across which luxury goods flow into the Mediterranean world. 10

WHORE OF BABYLON

While mid-Republican sumptuary laws, Amazons on innerEurasian trade-routes, and witchy women all begin to highlight the suspicious feminizing of luxury trade, perhaps the crowning example of this development is the “Whore of Babylon” in the firstcentury CE Christian apocalyptic text, Revelation. Lynn Huber has discussed in detail the vividly detailed ekphrastic nature of the description of this whore; she is not just a metaphor but rather “something the audience should envision” (Huber 2013: 61). This “city-woman” is a container (Huber 2013: 71), but a faulty one. This creature, “drunk with the blood of saints and martyrs of Jesus” (Rev 17:6), “formless and expansive seated on seven mountains, on a beast with seven heads and ten horns on vast waters” (Rev 17:7–15) is equated with Rome, “a great city that has domin-

See Frank 1991: 9 for the listing of three corridors linking Asia to the Mediterranean: the Nile-Red Sea corridor, the Syria-MesopotamiaPersia corridor, and the Aegean-Black-Sea-Central Asia corridor. Each of these corridors for an influx of luxury trade is straddled by a misbehaving woman: Meroe (Red-Sea Nile), the whore of Babylon (SyriaMesopotamia-Persia), and Medea and the Amazons (Aegean-Black SeaCentral Asia). 10

338

ELIZABETH ANN POLLARD

ion over all the earth” (Rev 17:18). At her fall, an angel cries out that “all nations have drunk the wine of her impure passion, and the kings of the earth have committed fornication with her, and the merchants of the earth have grown rich with the wealth of her wantonness” (Rev 18:3). Merchants weep at her fall, “since no one buys their cargo any more, cargo of gold, silver, jewels and pearls, fine linen, purple, silk and scarlet, all kinds of scented wood, all articles of ivory, all articles of costly wood, bronze, iron and marble, cinnamon, spice, incense, myrrh, frankincense, wine, oil, fine flour and wheat, cattle and sheep, horses and chariots, and slaves, that is, human souls” (Rev 18:11–13). Finally, “shipmasters and seafaring men, sailors and all whose trade is on the sea, stood far off and cried out as they saw the smoke of her burning” (Rev 18:17–18). These passages from Revelation perfectly package the grossly feminized characterization of luxury trade, and a city that lives and dies on that intercourse, as a drunken, boundless, whore.

LEAKY AND UNCONTAINED WOMEN

What then do we make of these women on the margins and the feminizing of trade in luxury goods? What do we make of the concerns about women’s desire for foreign spices and luxury goods: silk to clothe their bodies and spices for cosmetics, perfumes, and ritual practices such as burial? One could of course see this feminizing as merely part of the undeniable Roman (and before that, Greek) program of associating anything bad or suspect with women. As Daniel Mendelsohn has commented with respect to women in Greek tragedy, “women – thought to be irrational, emotional, deceitful, slaves of passion – were themselves ‘other’ to all that the free, rational, self-controlled male citizen was”… “symbolic entities representing everything ‘other’ to that smoothly coherent citizen identity” (Mendelsohn 2003). Of course, feminizing a rival, or any object of derision, was an effective rhetorical tool in antiquity. For example, Cicero famously took this approach in his salvo against Mark Antony. In his scathing mid first-century BCE Second Philippic, Cicero deployed feminizing invective to locate the roots of Antony’s debauchery in his youth. According to Cicero, Antony “took on the man’s toga, which he immediately made into a woman’s toga,” when he reportedly prostituted himself to men until he “had been given the matron’s stola and taken into stable and sure matrimony” by his lover

WOMEN AND THE MARGINS OF THE ROMAN WORLD

339

Curio (Cicero, Philippic 2.44). Almost 150 years later, the biographer Plutarch recalled other phrasing used by Cicero in this same speech, namely that just as Helen caused the Trojan War, so Antony was the cause of the civil war of the first century BCE (Plutarch, Antony 6, drawing on Cicero, Philippic 2.55). But this feminizing invective by Greeks and Romans was not limited to individuals. Whole groups and types of activities could be feminized, in both visual and literary form. David Castriota, for example, explores the “visual rhetoric… in the very heart of Athenian civic space” conveyed by the paintings of the Stoa Poikile in fifth-century Athens. According to Castriota, these paintings – described by the second century CE author Pausanias (Description of Greece 1.15.2) – feminized Persian and Trojan “barbarians” by dressing “weak-spirited, panicky” Amazons in Persian garb and imaging fallen Troy through “captive, powerless Trojan women about to be made sex slaves” (Castriota 2005: 100–01). Whole categories of activity, such as magical practices, could be feminized by Greek and Roman authors. For instance, depictions of who does witchcraft in literary sources exposes writers feminizing inappropriate behavior that was actually conducted by men. Even though 85–90% of all magical spells that have been found written on papyri or etched on metal tablets and gemstones had men as their agents or beneficiaries, almost every witch in the Greek and Roman imagination – from Circe and Medea of Greek epic and drama to Canidia and Erictho of Latin poetry – is female (Ogden 2002: 78; Graf 1997: 185–90; Dickie 2000: 563–83). One could also see the correlating of long distance luxury trade with women through the lens of French theorizing that some things, women included, are just “good to think with.” As LeviStrauss has argued, women are useful signs and objects of exchange (Levi-Strauss 1969: 496 in Rabinowitz 2004). Concerned about the fifth-century BCE Greek polis? Consider reading Euripides’ plays for a better understanding of the “masculinity of Athenian society and its anxiety about female sexuality” (Rabinowitz 1993: 10). Concerned about the impact of eastern religious traditions on secondcentury BCE Rome? Consider reading Livy’s description of the women’s involvement in the Bacchanalian conspiracy of 186 BCE. Ross Shepherd Kraemer has upended much of what she argued in Her Share of the Blessings and numerous articles that explored the potential realities of women’s religious experience, expressing

340

ELIZABETH ANN POLLARD

doubt in her Unreliable Witnesses that the evidence that appeared to tell us something about women was ever evidence for real women at all (Kraemer 1994 and 2011)! Kraemer, in Levi-Strauss fashion, writes: “that far from corresponding easily and usefully to women’s experience and lives, ancient sources deploy ancient ideas about gender, mapped onto female (and male) characters, to explore a range of issues of concern to their largely elite male authors and ancient audiences” (Kraemer 2011: 31). Indeed, many scholars have critically engaged the evidence for women’s experience, wondering if that evidence is genuine reporting or male fantasy, or a complex layering of both. Turning specifically to the issue at hand in this contribution, could there be something about the Greek and Roman conceptualizing of women that makes them a particularly good metaphor for trade and boundary-crossing?

CONCLUSION

I would suggest that the Greek and Roman philosophical understanding of women as leaky, porous, permeable, and needing male control made them a particularly apt vessel for thinking about trade and luxury. Anne Carson has put it well when she noted that “when we focus on Greek attitudes to and treatment of the female, we see anxiety about boundaries from a particular perspective – that of hygiene, physical and moral” (Carson 1991: 135–36). 11 To begin with, the patrilocal marriage of Greek and Roman society made women into a “mobile unit.” A man's location in oikos/domus and polis/civitas was fixed from birth; a woman, however, was always in motion, from her father’s home, to her husband’s home, to, if she outlived her husband, which is likely, the home of her brother or son (Carson 1991: 136). Women, by their very nature, must cross the boundaries set up by and between men, much the same way that luxury trade did. For expediency, this chapter draws heavily from Carson’s discussion of female anatomy. Other helpful, comparable discussions of Greek and Roman interpretations of women’s anatomy include: Rouselle 1988, Sissa 1990 and 1991, Hanson 1991, Clark 1993, Dean-Jones 1994, and King 1998. 11

WOMEN AND THE MARGINS OF THE ROMAN WORLD

341

As far as physical composition is concerned, some scholars have argued that “the major anatomical feature distinguishing women from men is the existence of a hodos, a route extending from the orifices of the head to the vagina: [in other words] woman has ‘an uninterrupted vagina from nostrils to womb,’” making her a thoroughly permeable being – not only are her boundaries permeable but she is easily passed right through (King 1998: 28). Not only do some writers conceive of woman as an easily traversed hodos, but women’s wombs were imagined, by some, as upturned amphorae, that vessel so emblematic of long distance trade. 12 The idea of women’s permeability was furthered by the characterization of their very flesh as having a loose texture, being spongy and wool-like – the idea being that while a man's flesh is like a finely woven garment, female flesh is like un-worked wool. 13 In a world where philosophical and medical understandings of the body promoted the notion that dry and hot were the ideal state for the body, women were incurably wet and cold. 14 As a result of their cold wetness, women were thought to be more susceptible to “gluttony [and] extravagance.” 15 Another consequence of women’s nature is their inability to maintain what the Greeks called sophrosyne, what Foucault called “Care of the Self,” and what we might clarify as the ability to control natural impulses (Carson 1991: 142). Not only are women wet, but their spongy, wool-like wetness is leaky and unbounded, or at least formless. 16 Perhaps the most vivid mythological image for women’s leakiness is that of the Danaidai, whose punishment in For a range of understandings of the womb’s anatomy (vessel, jug, wineskin, etc.), see Dean-Jones 1994: 65–69. Magical gemstone amulets (often of hematite and with imagery that suggests they have something to do with uterine health) contain many examples of a jug- or amphorashaped uterus (Items 335–362, in Delatte and Derchain 1964: 245–57). 13 King 1998: 28–29, drawing on Hippocrates Glands 1 and 16 and Diseases of Women 1.1. 14 Hippocrates, Aristotle, Heraklitos, and Aristophanes, among others, all cited in Carson 1991: 137–38. 15 Carson 1991: 138, citing Aristotle, Empedocles, and Simonides. 16 Plato, Aristotle and Pythagoreans, cited in Carson 1991: 153–58. 12

342

ELIZABETH ANN POLLARD

the underworld for their murder of their bridegrooms on their wedding night was to spend eternity attempting to gather water in sieves. Based on philosophers, legislators, poets and societal conventions, Anne Carson has argued succinctly: “women are formless creatures who cannot or will not, or do not maintain their own boundaries and … are awfully adept at confounding the boundaries of others” (Carson 1991: 155). Carson continues, “in sum, the female body, the female psyche, the female social life, and the female moral life are penetrable, porous, mutable, and subject to defilement all the time” (Carson 1991: 159). In the minds of Greek writers and the Roman thinkers who inherited their point of view, women’s societal dislocation, their nature, and their physical constitution made them a constant source for potential miasma, or pollution – a threat to boundaries that needed protection and control. What better way, then, for Roman writers and thinkers to articulate the threat posed by luxury and exotic trade than leaky, boundless, polluting women. The uncontained limen-crossing women in the streets of Rome lobbying the repeal of the Oppian law, like wild animals, break free of the control of the domus and of their husbands and fathers to push for their right to new luxuries made possible by newly-opened access to foreign goods. 17 The Amazons were nomadic women free from the control of men, uncontrolled “mobile units” on the fringes of Roman interaction along an all-important trade route through Inner-Eurasia. Meroe, There is a clear resonance between Cato’s characterization of the politicking wives as unmastered creatures and magic-medical texts’ reference to wombs as wild animal-like. Second-century CE Greek medical writer Aretaeus referred to the womb as being like an independent animal that might wander through the body attracted to good smells (Aretaeus, Medical Writings 2.11.1–3). A third-fourth century CE Greek magical handbook includes a spell that urges the womb not to “gnaw into the heart like a dog, but remain in your place” (PGM VII.269–70). While the “wandering womb” hypothesis was debated even in antiquity (Soranus’s Gynecology 3.29.5, for one, challenged the interpretation), it is suggestive of how ideas about the way that women’s bodies worked translated into related invective, as I am arguing in this chapter was the case with women’s permeable, leaky bodies informing the notion of luxury trade as a feminized vice. 17

WOMEN AND THE MARGINS OF THE ROMAN WORLD

343

drunk, oversexed, and boundary-crossing in her sexuality and use of witchcraft and Medea, a portable potion-mistress, dislocated by Jason from the Black Sea to the central Mediterranean, lacks sophrosyne in her overemotional response to Jason’s rejection, to the point of killing her own children. These witchy women sit along vital trade routes funneling luxury goods into the Mediterranean. And the whore of Babylon, “drunk on the blood of Jesus and the martyrs” is a textbook case: “For all the nations have drunk of the wine of the wrath of her fornication, and the kings of the earth have committed fornication with her, and the merchants of the earth have grown rich from the power of her luxury” (Rev 18:3). The very language used to describe these women on the margins and their desire for luxury supports this chapter’s assertion that philosophical and medical understandings of women's bodies suggests why, from the beginning of Roman imperialism in the third century BCE through the late Roman empire, Roman literature and historical narrative associated permeable and threatened boundaries and the trade and products that crossed these boundaries with women. Unstoppable, leaky, permeable, uncontrollable, witchy, and excessive women by their very nature were the logical culprits when male writers agonized over the costs – monetary and moral – of the fluid margins of the Roman world and the unquenchable thirst for the exotic goods that flowed inexorably into the Mediterranean.

BIBLIOGRAPHY

Appian (1912). Roman History, Volume II. Edited and translated by Brian McGing. Cambridge, MA: Harvard University Press. Apuleius (1996). Metamorphoses (The Golden Ass), Volume I: Books 1– 6. Edited and translated by J. Arthur Hanson. Cambridge, MA: Harvard University Press. Bauman, Richard A. (1992). Women and Politics in Ancient Rome. New York: Routledge. Begley, V. and R. D. de Puma, eds. (1991). Rome and India: The Ancient Sea Trade. Madison: University of Wisconsin Press. Blok, J. (1995). The Early Amazons: Modern and Ancient Perspectives on a Persistent Myth. Leiden: E.J. Brill.

344

ELIZABETH ANN POLLARD

Bowditch, L. (2006). Propertius and the Gendered Rhetoric of Luxury and Empire. Comparative Literature Studies 43 (3), 306– 25. Carson, A. (1991). Putting Her in Her Place: Woman, Dirt, and Desire. In: F. Zeitlin, J. Winkler and D. Halperin, eds., The Construction of Erotic Experience in the Ancient Greek World. Princeton, NJ: Princeton University Press, pp. 135–70. Casson, L. (1986). P. Vindob G 40822 and the Shipping of Goods from India. Bulletin of the American Society of Papyrologists, 23(3), 73–79. Casson, L. (1989). The Periplus Maris Erythraei: Text, Translation, and Commentary. Princeton, N.J.: Princeton University Press. Castriota, D. (2005). Feminizing the Barbarian and Barbarizing the Feminine: Amazons, Trojans, and Persians in the Stoa Poikile. In: J. Barringer and J. Hurwit, eds., Periklean Athens and its Legacy. Austin, TX: University of Texas Press, pp. 89–102. Charlesworth, M. (1951). Roman Trade with India: A Resurvey. In: P. Coleman-Norton, ed., Studies in Roman Economic and Social History in Honor of Allan Chester Johnson. Princeton, NJ: Princeton University Press, pp. 131–43. Charlesworth, M. (1924). Trade-routes and Commerce of the Roman Empire. New York: Cooper Square Publishers. Reprint 1970. Cimino, R.M., ed. (1994). Ancient Rome and India: Commercial and Cultural Contacts between the Roman World and India. New Delhi: Manohar. Clark, G. (1993). Women in Late Antiquity: Pagan and Christian Lifestyles. New York: Oxford University Press. Cooley, A. and M.G.L. Cooley (2004). Pompeii: A Sourcebook. New York: Routledge. Davis-Kimball, J. and Behan, M. (2002). Warrior Women: An Archaeologist’s Search for History’s Hidden Heroines. New York: Warner Books. Dean-Jones, L.A. (1994). Women’s Bodies in Classical Greek Science. New York: Oxford University Press. Delatte, A. and Ph. Derchain (1964). Les Intailles Magiques GrécoÉgyptiennes. Paris: Bibliothèque Nationale.

WOMEN AND THE MARGINS OF THE ROMAN WORLD

345

de Romanis, F. and A. Tchernia, eds. (1997). Crossings: Early Mediterranean Contacts with India. New Delhi: Manohar. Dickie, M. (2000). Who Practised Love-magic in Classical Antiquity and in the Late Roman World? The Classical Quarterly 50(2), 563–83. Ferguson, J. (1978). China and Rome. Aufstieg und Niedergang der Römischen Welt, II.9.2, 581–603. Frank, A. (1991). A Plea for World System History. Journal of World History 2(1), 1–28. Graf, F. (1997). Magic in the Ancient World. Cambridge, Mass.: Harvard University Press. Gorman, R.J. and V.B. Gorman (2014). Corrupting Luxury in Ancient Greek Literature. Ann Arbor, MI: University of Michigan Press. Guliaev, V. (2003). Amazons in the Scythia: New finds at the Middle Don, Southern Russia. World Archaeology 35(1), 112–25. Hanson, A.E. (1991). Medical Writer’s Woman. In: F. Zeitlin, J. Winkler and D. Halperin, eds., The Construction of Erotic Experience in the Ancient Greek World. Princeton, NJ: Princeton University Press, pp. 309–38. Huber, L.R. (2013). Thinking and Seeing with Women in Revelation. New York: Bloomsbury T&T Clark. King, H. (1998). Hippocrates’ Woman: Reading the Female Body in Ancient Greece. New York: Routledge. Kraemer, R. (1994). Her Share of the Blessings: Women’s Religions among Pagans, Jews, and Christians in the Greco-Roman World. New York: Oxford University Press. Kraemer, R. (2011). Unreliable Witnesses: Religion, Gender, and History in the Greco-Roman Mediterranean. New York: Oxford University Press. Lévi-Strauss, C. (1969). The Elementary Structures of Kinship. 2nd Edition. Trans. by J. Von Sturmer, J. Bell, and R. Needham. Boston: Beacon Press. Livy (1935). History of Rome, Volume IX: Books 31–34. Translated by Evan T. Sage. Cambridge, MA: Harvard University Press. Macrobius (1963). Saturnalia. Edited by Jacob Willis. Leipzig: Teubner.

346

ELIZABETH ANN POLLARD

Macrobius (1969). The Saturnalia. Translated by Percival Davies. New York: Columbia University Press. Mayor, A. (2010) The Poison King: The Life and Legend of Mithradates, Rome’s Deadliest Enemy. Princeton, NJ: Princeton University Press. Mayor, A. (2014) The Amazons: Lives and Legends of Warrior Women across the Ancient World. Princeton, NJ: Princeton University Press. Mendelsohn, D. (2003). The Bad Boy of Athens. New York Review of Books. Vol 50(2). Accessed via http://www.nybooks.com/ articles/2003/02/13/the-bad-boy-of-athens/. Miller, J. (1969). The Spice Trade of the Roman Empire, 29 B.C. to A.D. 641. Oxford: Clarendon Press. Ogden, D. (2002). Magic, Witchcraft, and Ghosts in the Greek and Roman Worlds. Oxford: Oxford University Press. Parker, G. (2002). Ex Oriente Luxuria: Indian Commodities and Roman Experience. JESHO 45(1), pp. 40–95. Parker, G. (2008). The Making of Roman India. Cambridge: Cambridge University Press. Penrose, W. (2017). Postcolonial Amazons: Female Masculinity and Courage in Ancient Greek and Sanskrit Literature. Oxford: Oxford University Press. Plutarch (1961). Lives, Volume V: Agesilaus and Pompey. Edited and Translated by B. Perrin. Cambridge, MA: Harvard University Press. Pollard, E. (2009). Pliny’s Natural History and the Flavian Templum Pacis: Botanical Imperialism in First-Century Rome. Journal of World History 20(3), 309–38. Potter, D. (1999). Odor and Power in the Roman Empire. In: J. Porter, ed., Constructions of the Classical Body. Ann Arbor: University of Michigan Press, pp. 160–89. Propertius (1990). Elegies. Edited and translated by G. P. Goold. Cambridge, MA: Harvard University Press. Rabinowitz, N. (2004). Changing Lenses: The Politics and Discourse of Feminism in the Classics Women, Good to Think With? [Online] Available from: http://www.stoa.org/diotima/essays/fc04/ Rabinowitz.html [Accessed 21 July 2016].

WOMEN AND THE MARGINS OF THE ROMAN WORLD

347

Rabinowitz, N. (1993). Anxiety Veiled. Ithaca: Cornell University Press. Raschke, M. (1978). New Studies in Roman Commerce with the East. ANRW II.9.2, pp. 604–1361. Rouselle, A. (1988). Porneia: On Desire and the Body in Antiquity. New York: Blackwell. Rosivach, Vincent J. (2006) The Lex Fannia Sumptuaria of 161 BC. Classical Journal 102(2), 1–15. Schmitthenner, W. (1979). Rome and India: Aspects of Universal History During the Principate. JRS 69, 90–106. Shapiro, H. (1983). Amazons, Thracians, and Scythians. Greek, Roman, and Byzantine Studies 24(2), 105–14. Sidebotham, S. (1986). Roman Economic Policy in the Erythra Thalassa 30 B.C.-A.D. 217. Leiden: E.J. Brill. Sidebotham, S. (2011). Berenike and the Ancient Maritime Spice Route. Berkeley, CA: University of California Press. Sissa, G. (1990). Greek Virginity. Cambridge, MA: Harvard University Press. Sissa, G. (1991). Maidenhood without Maidenhead: The Female Body in Ancient Greece. In: F. Zeitlin, J. Winkler and D. Halperin, eds., The Construction of Erotic Experience in the Ancient Greek World. Princeton, NJ: Princeton University Press, pp. 339–64. Stewart, A. (1995). Imaging the Other. Poetics Today 16, 571–97. Strabo (1961). Geography, Volume V: Books 10–12. Translated by H.L. Jones. Cambridge, MA: Harvard University Press. Smith, R.R.R. (1987). The Imperial Reliefs from the Sebasteion at Aphrodisias. JRS 77, 88–138. Thorley, J. (1969). The Development of Trade between the Roman Empire and the East Under Augustus. Greece & Rome 16(2), 209–23. Thorley, J. (1971). The Silk Trade between China and the Roman Empire at its Height, Circa a.d. 90–130. Greece & Rome 18(1), 71–80. Tomber, R. (2005). Trade Relations in the Eastern Mediterranean and Beyond: The Egyptian-Indian Connection. In: M. B.

348

ELIZABETH ANN POLLARD

Briese and L. E. Vaag, eds., Trade Relations in the Eastern Mediterranean from the Late Hellenistic Period to Late Antiquity: The Ceramic Evidence. Odense: University Press of Southern Denmark, pp. 221–33. Tomber, R. (2008) Indo-Roman Trade: From Pots to Pepper. London: Duckworth. Von Bothmer, D. (1957). Amazons in Greek Art. Oxford: Clarendon Press. Walcot, P. (1984). Greek Attitudes Towards Women: The Mythological Evidence. Greece & Rome 31(1), 37–47. Warmington, E. (1928). The Commerce Between the Roman Empire and India. London: Curzon Press. Reprint 1974. Wheeler, M. (1955). Rome Beyond the Imperial Frontiers. New York: Philosophical Library. Woolf, G. (1990). World-Systems Analysis and the Roman Empire, JRA 3, 44–58. Young, G. (2001). Rome’s Eastern Trade: International Commerce and Imperial Policy 31 BC – AD 305. New York: Routledge.

HUMAN POPULATION INCREASE AND ITS EFFECTS ON THE ARID LANDSCAPE OF SOUTHERN JORDAN: AN ARCHAEOBOTANICAL STUDY JENNIFER RAMSAY 1

This paper presents and overview of the results of botanical analysis carried-out at five archaeological sites in southern Jordan that were occupied between the Nabataean, Roman, Byzantine and Post Classical periods (1st century BCE thorough the 12th century CE). The sites of Petra (Ramsay and Bedal, 2015), Humayma (Ramsay 2013), Bir Madhkur (Ramsay and Smith 2013), Roman and early Islamic Aila (Ramsay and Parker 2016) and ʿAyn Gharandal (Ramsay in prep.) are all located in the arid desert environments of the al-Sharah Mountains, the Hismâ and Wadi Araba in southern Jordan (Fig. 1). Many of the sites in this region were first documented by Fritz Frank and Nelson Glueck in the 1930’s (Frank 1934; Glueck 1935). More recent work has focused on aspects of the cultural landscape such as land-routes (Graf, 1992; Zayadine, 1992). Surveys of note were conducted in the Southeastern Araba (Smith With appreciation and gratitude I would like to acknowledge the support and encouragement Tom has generously given to me over the years. He is not only a wonderful mentor and colleague but also a true friend. I count myself lucky to have had the opportunity to learn from him and work with him in the field. Tom has always been gracious with both praise and constructive criticism, always making me strive to do better, which I am trying to perfect with my own students. And finally, when all else fails I could always manage to twist his arm for a pop. 1

349

350

JENNIFER RAMSAY

2014; Smith and Neimi 1995; Smith et al. 1997), the region near Petra surrounding Jabal Haroun (Lavento et al. 2004, 2007a, 2007b; Kouki, 2009) and north of Petra (Tholbecq 2001; Knodell and Alcock, 2011; Alcock and Knodell, 2012).

Figure 1. Research sites in Southern Jordan

Data from these studies and several studies related to ancient agriculture in the region have documented features related directly to agricultural production, such as field walls, terraces, water channels

HUMAN POPULATION INCREASE

351

and check dams (Barker 2002, 2012; Evenari et al. 1958, 1961; Mayerson 1960, 1985; Kedar, 1967; Erickson-Gini 2006, 2010, 2012; Hirschfeld and Tepper 2006; Avni et al. 2009; Ashkanazi et al. 2012). Archaeobotanical evidence from the study sites in this paper provides confirmation of local agriculture in the form of cereal grains, chaff and associated weed species. This data points to a much greener desert in antiquity even with climatic conditions that appear to have not changed significantly and indicate a period of greater agricultural production than has been possible in the present day.

BACKGROUND OF STUDY SITES Aila Aila is located at the northern apex of the Red Sea in a coastal oasis that is climatically hyperarid with a mean annual rainfall of approximately 40mm. The site of Aila was excavated and well documented by S. Thomas Parker starting in 1994 and continuing on through 2002 (Parker 1996, 1997a, 1997b, 1998, 2000, 2002, 2003, 2006, 2009a, 2009b, 2009c, 2013). Historical sources note that Aila was occupied during the Nabataean, Roman and Byzantine periods. As early as the 1st century BCE Strabo notes that Aila was at an important trade junction of several overland trade routes that ran through the Negev to Palestine, north to the Transjordan and westward through the Sinai towards Egypt (Geog. 16.2.30). According to Eusebius, by the early fourth century the 10th Roman legion, which was formerly stationed in Jerusalem, was moved to Aila (Eusebius, Onomaticon, 6.17–21) where it remained at least until the beginning of the 5th century. Aila remained a dynamic port with important shipping route for spices from Indian through Late Antiquity. Aila continued to flourish under Muslim rule from the 7th to the 12th CE. With the population concentration at Aila, it would seem likely that there would have been agriculture carried-out in its hinterland. Although hyper-arid in nature, Aila had a shallow water table where fresh water could be attained and if cisterns could capture floodwaters then it is likely that cereal agriculture could have been viable. Likewise there is the possibility that oasis agriculture was practiced at the site, which has been noted to be able to support multiple crops (Tengberg 2012).

352

JENNIFER RAMSAY

Figure 2. Map illustrating the number of sites (mostly military) in the Late Roman/Byzantine period of occupation

The Roman Aqaba Project (RAP), directed by S. Thomas Parker, surveyed the region around ancient Aila (Smith 2014; Smith and Neimi 1995; Smith et al. 1997) but revealed no evidence of an agricultural system, likely due to modern development since the 1960s, which has destroyed much of the archaeological and geological evidence in Aqaba. However, Musil, who visited the region ca. 1900, observed that the people of Aqaba attempted farming in the immediate region after a heavy rainfall (Musil 1907). And an an-

HUMAN POPULATION INCREASE

353

cient soil horizon found during excavations at Aila may provide supporting evidence for local agriculture in the hinterland of Aila (Neimi 2013). Likewise there is possible evidence of ancient field systems seen through Google Earth in the hinterland of Aila that may have been in use during the Nabataean through Byzantine periods (Fig. 2). Bir Madhkur Bir Madhkur, currently being excavated and surveyed by Andrew Smith II was a major caravan station along the ancient spice route that connected Petra with the Mediterranean port of Gaza (Smith 2005a, 2005b, 2010). The main features of the site that have been excavated include a Late Roman/Byzantine army fort (Area A), a domestic settlement surrounding the fort (Area D) and a bath complex (Area B). There were several smaller, likely associated sites in the vicinity of Bir Madhkur including farmhouses and nomadic encampments (Smith 2010). Bir Madhkur lies in on the eastern edge of the Wadi Araba that is hyper-arid and has frequently been characterized as a wasteland with an average rainfall of between 50–100 mm annually, concentrated in the winter months (Ramsay and Smith 2013). In early surveys of the region Glueck (1935) and Frank (1934) noticed evidence of extensive agricultural activity which did not seem in keeping with a hyper-arid environment. David McCreery (n.d.), as Director of the American Center of Oriental Research, visited Bir Madhkur in the 1980’s and collected predominately Nabataean and Late Roman pottery. He was followed by King et al. (1987), whose pottery ranged in date from the Nabataean to Byzantine periods, which provided a general chronological framework for the agricultural installations of the site. In 1994 Bir Madhkur was surveyed by the Southeast Araba Archaeological survey by Smith and Tina Niemi, who also noted extensive ancient field systems in the vicinity of the site (Smith and Niemi 1994; Smith et al. 1997; Smith 2005a, 2010, 2014) and evidence published by Ramsay and Smith (2013) argues for the agriculture during the Late Roman/Byzantine period at Bir Madhkur.

354

JENNIFER RAMSAY

Humayma Humayma, Ancient Hawar, is located in the Hismâ desert in Jordan and lies between Roman Aila on the Red Sea and Petra. It was originally excavated by John Oleson between 1992–2004 and is currently being excavated by Barbara Reeves. Humayma was founded by the Nabataean King Aretas III in the 80’s BCE probably to serve as a center for sedentarization of the nomadic Nabataean pastoralists who occupied the area (Oleson 1997b, 2001, 2004; Oleson et al. 2008). Ancient Hawar was the only substantial settlement in the Hismâ Desert in southern Jordan. The Roman Road from Syria to the Red Sea, the Via Nova Traiana, passed by the site. The environment is a bleak desert characterized by sparse desert vegetation. However, at Humayma excellent water-catchment field systems for cultivation were documented. Oleson (1988, 1990, 1995, 2008, 2010) assessed that if the whole field system around Humayma absorbed as much as 50% of the current average of 80mm of annual precipitation, 40,000 cubic meters would still be available, which he noted would be enough water, if stored, to sustain a population of 11,000 for a year. Through careful management of the meager spring water and precipitation, the resulting community was able to enjoy a settled existence based on agriculture, stock-raising, and the servicing of caravans. In 106 CE the Nabataean Kingdom became the Roman Province of Arabia and shortly thereafter a 500 person Roman fort was built at Hawar. The modest settlement prospered into the Byzantine period as indicated by the 5–6 churches identified on the site (Oleson and Schick 2014). Humayma’s notoriety rose during the Islamic period as it was the village where the Abbasids plotted the overthrow of the Umayyad dynasty. Excavations at the site uncovered three buildings associated with the Abbasid family including the qasr and a small mosque (Oleson 1997a). Petra The ancient city of Petra is situated about 80 km south of the Dead Sea and 275 km southwest of modern Amman in Jordan at an elevation of 825 m. Although Petra experienced slightly greater precipitation than sites in the Wadi Araba and in the Hismâ, its role as the capital of a trading empire in the ancient Near East during the late first century BCE, influenced the landscape in the region. The

HUMAN POPULATION INCREASE

355

city was the principal city of ancient Nabataea and is located in an arid environment that has no rivers and few natural springs. In antiquity the region received less than 100 mm of rainfall annually but generally this ranged between 25–75mm due to the high evaporation rate of exposed water (Rosen 2007; National Water Master Plan of Jordan 1977). Although this may seem like an inhospitable landscape today, Petra likely had a very significant population, perhaps 30,000 (Joukowsky 2001:2). It is clear from this that the Nabataeans rapidly learned how to collect the small amounts of precipitation that were available to them. Oleson (1995) has documented a wide variety of hydraulic engineering techniques that were utilized by the Nabataeans as their population grew due to rapid urbanization. The archaeobotanical material that was analyzed from the Petra Garden and Pool Complex provide an indication of the agricultural economy during the difference phases of occupation of Petra. The results for the cereal crops identified establish that there was crop processing in the region of Petra (Ramsay and Bedal, 2015). ʿAyn Gharandal The site of ʿAyn Gharandal is located on the eastern edge of the Wadi Araba approximately 50 km north of Aila. The ongoing ʿAyn Gharandal Archaeological Project under the direction of Robert and Erin Darby established dates for the occupation of the site to between the Nabataean through the Islamic periods (Darby and Darby 2012). The presence of an artesian spring in the mouth of the nearby Wadi Gharandal, which still flows abundantly today, presumably served as the reason for human occupation at the site. To date a Roman period fort, bath house and an aqueduct have been documented (Darby and Darby 2012). The analysis of the archaeobotanical assemblage for ʿAyn Gharandal, is still in the preliminary stages of analysis but the assemblage recovered thus far provides potential evidence for local agriculture being practiced nearby and the likely use of dung as a fuel source. It is difficult to determine the scale to which agriculture was being practiced; however plant husbandry likely was practiced in the hinterland of the site due to the presence in the assemblages of crop by-products and weed species common in crop fields.

356

JENNIFER RAMSAY

REGIONAL SUPPORTING EVIDENCE IN SOUTHERN JORDAN FOR POPULATION INCREASE

Agricultural intensification across southern Jordan does not appear to have happened concurrently at all sites. A contributing factor to the increase or initiation of local cereal agriculture around some sites is the population increase that likely occurred by an influx of Roman military personnel in southern Jordan during the Roman and Byzantine period. The fortification structures that have been documented in in the Wadi Araba, such as Mezad Dafit, Rujm Taba, Yotvata, ʿAyn Gharandal, Khirbet es-Saʾidiyeen, Mezad Beʾer Menuha, Qasr et-Tayyiba, Moyet ʿAwad and Mezad Hazeva support this increase beginning in the fourth century (Fig. 2) (Smith and Niemi 1994; Smith et al. 1997; Smith 1998; Niemi and Smith 1999; Smith 2005). These structures speak to the importance of the region during the Late Roman period as is documented historically (Eusebius, Onomasticon, 210; Notitia Dignitatum Or. 34.30).

Figure 3. Threshing floor and field walls near Bir Madhkur (Photo by J. Ramsay)

Further, archaeobotanical remains recovered from a farm house in the hinterland of Bir Madhkur show local cereal cultivation and date to at least the Late Roman/Byzantine period. Further verification can be seen for more intensive agricultural practices in south-

HUMAN POPULATION INCREASE

357

ern Jordan in the form of ancient field walls (Fig. 3), catch dams (Fig. 4) and diversion walls in wadis (Fig. 5) that are scattered across the landscape.

Figure 4. Catch dam near Petra (photo by J. Ramsay)

358

JENNIFER RAMSAY

The following discussion will examine archaeobotanical remains as a means of addressing agricultural production in arid landscapes. This data will also aid in illuminating changes to the natural environment and landscape of southern Jordan caused by cereal agriculture. Specifically, evidence will be presented that relates to issues of local cereal cultivation, crop choice, trade and the timing and scale of agricultural intensification.

Figure 5. Diversion walls in wadi north of Aqaba (courtesy Google maps).

METHODOLOGY AND RESULTS

The quantity of archaeobotanical material recovered from the study sites in southern Jordan varied and was dependent on the skill of who was collecting samples, the method of recovery (flotation vs. sieving), the archaeological contexts being sampled, and the length of time the site was excavated; for example Aila was excavated over six seasons and the material recovered from Bir Madhkur was only from one season. The methodology related to the recovery, analysis and interpretation of the archaeobotanical remains are covered in the publications related to each site and will not be covered in depth in this paper (see Ramsay and Parker 2016; Ramsay and Smith 2013; Ramsay 2013; Ramsay and Bedal 2015). There were

HUMAN POPULATION INCREASE

359

also differences in the contexts that were being excavated between sites; for example, bath houses, forts, a farmhouse and domestic structures. These factors all lead to variability in the quantities of samples analyzed. For clarity related to sampling, table 1 states the number of samples taken from each site, the number of samples that had identifiable remains in them and the number of plant taxa that were identified from each site. Table 1. Archaeobotanical sample data from study sites

Sites Number of Samples Analyzed to Date Number of Samples that Contained Botanical Remains Number of Identified Remains Number of Taxa Identified

Aila

Humayma

Petra

Bir Madhkur

‘Ayn Gharandal

Totals

204

152

277

62

16

711

148

92

127

40

10

417

4255

1888

1354

1093

172

8762

58

52

38

35

22

As there were multiple occupation periods on each site, there was also inconsistency between what periods were emphasized for sample collection (Table 2). As illustrated in Table 2, each period has between 54–148 samples jointly that were used for analysis. Table 3 shows the number of identified taxa from each site. The number of taxa identified from each category of plant species is generally similar between sites, specifically cereals, legumes and fruit/nut taxa identified, with the exception of the small number of taxa identified from ʿAyn Gharandal due to the paucity of samples recovered thus far (Table 3).

360

JENNIFER RAMSAY Table 2. Number of samples recovered with plant remains present by period for the study sites.

Aila Sites Samples from the Nabataean/ Early Roman 8 Period Samples from the Roman/ Late Roman 33 Period Samples from the Late Roman/ Byzantine 37 Period Samples from the PostClassical – 70 Medieval

‘Ayn Gharandal

Total

0

0

54

56

0

0

107

32

39

40

0

148

5

13

0

10

98

Humayma

Petra

29

17

18

Bir Madhkur

Table 3. Number of taxa identified from the study sites by plant group.

Sites Number of Cereal Taxa Number of Legumes Number of Fruit and Nut Number of Wild Species

Aila

Humayma

Petra

Bir Madhkur

‘Ayn Gharandal

5

5

6

4

1

6 9

38

6 6

35

3 4

25

4 6

21

4 2

15

It is difficult to interpret archaeobotanical assemblages from a single site but to compare the results from numerous sites is fraught with difficulties. Trying to find a statistical method to compare samples that will provide meaningful results is a challenge. For example, in archaeobotanical samples, the assumption is that all things being equal, larger sediment samples would have more plant remains present. Density is used as a measure to correct for this

HUMAN POPULATION INCREASE

361

assumption. If the volume of sediment sampled is standardized, other variables can be measured against it, like preservation, recovery rates and deposition (Miller 1988: 73). Unfortunately, the volume of sediment processed for each sample from the study sites was not known as the samples had been collected and processed in the field over several excavation seasons by different people. Another method of analysis is species diversity measures, which account for the total number of taxa present in a population and the abundance of each species (Pielou 1969: 221–235). The Shannon Weaver index, an example of a diversity measure, was introduced to archaeobotany by Pearsall (1983: 130–31) as quantifying the plant assemblages by absolute counts may not accurately reflect the pre-depositional assemblage due to taphonomic processes but these counts can be used further in quantitative and statistical analysis. The formula used by Pearsall (1983: 137) is as follows: H = Sum (Nj/N) log (Nj/N) Where H = diversity index

N = total number of seeds in the period

Nj = total number of seeds of taxon j in the phase

However, Pielou (1977: 292) states that caution must be used as diversity depends on two independent properties of an assemblage and as a result, ambiguity is unavoidable. Therefore, these results may be useful for looking at general trends in plant assemblages but according to Popper (1988: 68), we must look to the data themselves to understand the nature of the trends. Popper (1988: 69) also warns that the Shannon-Weaver index requires high counts for each taxa, which limits its applicability to the archaeobotanical assemblages in this study. Another method is ubiquity, which is a measure of the number of samples in which a particular taxon appears as a percentage of a group of samples. By converting the raw counts of number of different taxa recovered from various contexts to percentages of the total taxa recovered on the site, it is possible to examine general trends in the number of taxa represented during each period. Therefore, applying ubiquity analysis to common economic crop species and crop by-products is useful as this measure disregards

362

JENNIFER RAMSAY

the absolute count of a taxon and instead looks at the number of samples in which the taxon appears (Popper 1988: 60–61). Comparing ubiquity scores from taxa at the study sites shows that it may be a better indicator than diversity when making comparisons between sites, however Pearsall (2000: 212) cautions that taphonomic processes that are site specific can cause varying results between sites. Nevertheless, by replacing diversity with ubiquity, a better understanding of the trends in the local agricultural economy and as a result landscape can be examined.

DISCUSSION

One of the major questions that need to be addressed when considering how the population increase during the Nabataean through Byzantine periods affected the landscape in southern Jordan is whether agricultural products identified in the archaeobotanical assemblages were the result of trade or local production. To address this question cereal species, their associated weed taxa and fruit and nut remains were examined from the study sites. For cereal crops, generally the presence of the by-products of crop processing, the chaff and weed and wild species common in cultivated fields, would be less likely to be present in an assemblage recovered from sites if local cereal agriculture was not being carried-out. The inclusion of weed species and chaff (culm nodes and rachis segments) in an assemblage would be unlikely if the material was a trade item since transportation cost would prohibit the addition of extra weight that represented non-economic goods. In some cases, even today, cereal crops are transported still in spikelet form (with intact chaff), however it is more likely that local agriculture was taking place if there is a significant percentage of cereal chaff and weeds in the samples. The results of the analysis of cereal crops from the study sites show strong support for local cultivation and evidence of trade during various periods of occupation. Figures 6a, 7a, 8a and 9a below show the ubiquity of cereal grains and chaff in archaeobotanical samples recovered from the samples dated to the Nabataean/Early Roman period, the Roman/Late Roman period, the Late Roman/Byzantine period and the Post Classical period of occupation at each site, which will be addressed below. Likewise, fruit and nut species are also good indicators of regional trade as many of the species identified at the study sites are not suited to grow in arid environments typical at the study sites. The following

HUMAN POPULATION INCREASE

363

discussion will examine the archaeobotanical evidence from each of the study sites with the aim of identifying when local cultivation of cereal crops was necessary and when trade was relied on to feed the populations. Petra during the Nabataean period possessed a large urban population that would have required significant quantities of cereal grain and other agricultural commodities. Figures 6a-b illustrates the changes in evidence of cereals cultivation of fruit and nut production or importation through time. Figure 6a clearly indicates that there were cereal grains in over 80% of the samples that dated to the Nabataean period from the Pool and Garden Complex at Petra. This data does not support cereal cultivation in the town center as there no evidence of the by-products of crop processing in this period. This is reasonable and likely reflects cereal production of this major center being carried out in its hinterland instead of directly on site. Hydraulic engineering features like catch-dams (Fig. 4) and cisterns are found scattered through the adjacent hinterland of Petra and support the Nabataean advancement in water technology.

Percentage of Samples with Taxa Present

Figure 6a. Ubiquity of cereal crops and by-products by occupation period at Petra Ubiquity of Cereals and By-Products from Petra through Time

90 80 70

Nabataean - Roman 17 Samples

60 50 40 30

Late Roman - Byzantine - 45 Samples

20 10 0

Post-Classical - 13 Samples

364

JENNIFER RAMSAY Figure 6b. Ubiquity of Fruit and Nut species by occupation period at Petra Ubiquity of Fruit and Nut Taxa from Petra through Time 90.0

80.0

Percentage of Samples with Taxa Present

70.0 Nabataean - Roman 17 Samples

60.0

50.0

Late Roman - Byzantine - 45 Samples

40.0 Post-Classical - 13 Samples

30.0

20.0

10.0

0.0

Ficus carica - fig

Olea europaea olive

Vitis vinifera grape

Phoenix dactylifera

Nutshell indet.

In the Late Roman/Byzantine period at Petra there appears to be a decline in agricultural production or importation as cereal grains occurred in only ca. 20% of the 45 samples that dated to this period. The context of these samples were from the Pool and Garden Complex site in central Petra, which has been documented to have undergone abandonment and destruction during this period which is supported by the archaeobotanical remains as there is a sharp decline in cereal grains recovered (Ramsay and Bedal 2015). The Post Classical period at Petra also supports a decline in population and abandonment of the site with a continued decrease in both cereals and fruits identified in the assemblages. Agriculture however, must have been practiced in the hinterland of Petra as the percent distribution of weed and wild species found on the site overwhelmingly (94.08%) represent species found in cultivated contexts (Fig. 6c). Therefore, like in urban environments today, you would not expect to find agricultural fields in the city but rather in the

HUMAN POPULATION INCREASE

365

region surrounding the city and the archaeobotanical data from Petra reflects this pattern. In the Pool and Garden Complex, the ubiquity of fruit and nut species in samples from the Nabataean contexts shows that in over 75% of samples olive (Olea europaea) was present, close to 50% of the samples had grape (Vitis vinifera), and there were also sparse quantities of fig (Ficus carica), date (Phoenix Dactylifera) and indeterminate nutshell (Fig. 6b). Olive would have been a staple commodity at Petra as it would have been used for oil, fuel and consumption. Therefore the presence of olive in such a large quantity of samples is not surprising. Grape is also not surprising, in terms of viticulture, as it is well attested to by the presence of wine presses and vineyards likely of Nabataean origin in the Beidha region, near Petra (al-Muheisen 1990; Bikai 2004). Figure 6c. percent distribution of weed and wild species by habitat at Petra

Petra - Percent Distribution of Weed and Wild Species By Habitat

Rocky 0.59%

Sandy 4.73%

Deserts/Fields 0.59%

Agricultural 94.08%

In the Late Roman/Byzantine period there is a slight increase in the fruit and nut that were already noted in the Nabataean period. The increase may be due to the greater number of archaeological samples taken from these layers, 45 as opposed to 17, or it may illustrate greater production or trade of these items. Unlike the other more arid sites, Petra was and still is able to successfully cultivate

366

JENNIFER RAMSAY

grapes and figs, as well as many other fruit species which have not been recovered archaeobotanically, such as pomegranate and Prunus species (e.g. apricots, peaches and plums). Olives likely grew in the highland regions to the north and west where the climate was more suited to olive production. In the Post-Classical period when scholars have argued Petra would have been mostly abandoned there is still a significant presence of olive, grape and nuts. This may be the result of small Islamic period settlements in and around Petra that maintained tree and vine crops or these inclusions may represent small family gardens where grape, olive and nuts could have been grown for local consumption. However, the results could also be an artifact of sampling as there were only 13 Post-Classical period samples, or an artifact of context as all samples were from the Garden and Pool Complex when the area had long since been converted to small agricultural fields. Figure 7a. Ubiquity of cereal crops and by-products by occupation period at Aila

Percentage of Sample with Taxa Present

Ubiquity of Cereal Grains and By-Products from Aila through Time

70 60 50 40

Nabataean/Early Roman - 4 Samples

30

Late Roman - 16 Samples

20 10 0

Early Byzantine - 20 Samples

Post Classic - 42 Samples

HUMAN POPULATION INCREASE

367

Figure 7b. Ubiquity of Fruit and Nut species by occupation period at Aila Ubiquity of Fruit and Nut Taxa from Aila through Time

Percentage of Samples with Taxa Present

80 70 60

Nabataean/Early Roman - 4 Samples

50 40 30

Late Roman - 16 Samples

Early Byzantine - 20 Samples

20 10

Post Classic - 42 Samples

0

According to the existing archaeobotanical data (although many samples are still being analyzed) there is no evidence of local cereal production during the Nabataean period at Aila; although cereal grains have been identified, no chaff has yet been noted (Fig. 7a). Data from the Roman period at Aila points to a slight increase in agricultural production with the identification of wheat (Triticum sp.), barley (Hordeum vulgare), rye (Secale sp.) and their associated byproducts (Fig. 7a). The following Byzantine period shows an increase in the presence of cereal grains identified but a notable decrease in the by-products of cultivation. These differences likely points to the importation of cereal grain that would have been necessary to supplement the local sources when the Roman military population increased with the arrival of the 10th Roman legion at Aila. The most dramatic shift that the archaeobotanical illustrates is the significant increase in cereal grains but not cereal by-products during the Post-Classical period likely attributed to its success in trade under Muslim rule between the 7th and 12th centuries (Whitcomb, 1994). Probable explanations for this increase in grain but lack of chaff could be larger importation of grain or that grain was

368

JENNIFER RAMSAY

being grown and processed in Aila’s hinterland, hence the lack of by-products. Looking at the percent distribution of weed and wild species identified, it is clear that there is strong support for agriculture being carried-out in the hinterland of Aila as 55% of the species recovered were only found in cultivated field systems (Fig. 7c). Fruit and nut taxa at Aila during the Nabataean and early Roman periods are biased by scant samples but do show fig, olive, grape, date and nuts being part of the local diet, as evidence of trade items or both (Fig. 7b). During the Late Roman period there strong evidence of fig, which has excellent storage properties for long-distance transport, so it is not surprising to find it throughout Aila. Grapes are commonly transported in a dried state as raisins and were likely a common import item and possibly even grown in small house gardens and eaten fresh, as they are today. Historical sources mention Aila’s plantations of date palms, which grow well in arid conditions with high soil salinity and are therefore likely a local product and not an import item (Khouri 1988, 140–141; Whitcomb 1994, 7; Zayadine 1994, 501). Figure 7c. percent distribution of weed and wild species by habitat at Aila

Aila - - Percent Distribution of Weed and Wild Species by Habitat Rocky 4%

Hydrophillic 7%

Sandy 21%

Field/Fallow 55%

Deserts 13%

The few olive remains likely indicate that these were imports as olive trees require a significant amount of water. The olive stones recovered were fragmentary and likely were used for fuel once the

HUMAN POPULATION INCREASE

369

fruit had been consumed as oil or food. Olive has also been recovered from the hyper-arid environment of Berenike where it was similarly interpreted as a likely import (Cappers 2006). Byzantine samples from Aila show an increase in nuts recovered but the frequency of fruit remains found in samples decreases. This may be an artifact of sampling as only 20 samples are dated to the Byzantine period and therefore only four samples need to have nut shell present to account for a 20% ubiquity. In the 42 samples that date to the Post-Classical period, 25% or over 10 samples had dates identified in the sample and almost 30% of the samples had nut shell. Date, as noted above was likely to have been grown locally, however, nut would have been an import as the trees require more precipitation that would have been available anywhere in the hinterland of Aila. Zizyphus spina-christi (Christ’s Thorn) was identified and was known through biblical and classical (Theophrastus, Inquiry into Plants, 4.7–8) sources to have been commonly grown in the valleys and lowlands of Israel (Dafni et al. 2005). The fresh and the dried fruits of the plant are edible and highly valued locally today by Arabs as well as Bedouins (Dafni et al. 2005). The presence of several of these fruit can also be accounted for if oasis agriculture was being utilized at Aila in antiquity. The date palm is the keystone species for this oasis agrosystem and its groves today can host fig, pomegranate, Christ’s thorn, grape, banana, mango, papaya and several citrus fruits (Tengberg 2012) for both human and animal consumption. Oasis plots are irrigated periodically to grow a wide variety of crops: cereals, legumes, vegetables, oil plants, medicinal plants and grasses (Tengberg 2012). With the documented date plantations at Aila, this oasis agricultural system provides a feasible means for local agriculture. The site of Humayma has two reservoirs and more than 50 rock-hewn and built cisterns, wadi barriers and a dam. It also had an aqueduct that once brought water from three natural springs in Ras al-Naqab to an open reservoir at the northern edge of the site (Oleson 1997b; 2010). As a result it is not surprising that Humayma has evidence of cereal agriculture during all phases of occupation (Fig. 8a). The samples from the Nabataean and early Roman periods indicate a fairly even distribution of wheat and barley grains and by-products, this evidence together with the presence of agricultural fields that surrounded the site in antiquity support significant cereal agriculture being practiced at this time. Moving into the

370

JENNIFER RAMSAY

Late Roman period there is a similar pattern of cereal remains that likely indicates continuity in crop cultivation at the site. There is a significant increase in barley grains and rachis as well as culm nodes in the Post-Classical period when Humayma was prospering as the village where the Abbasids plotted the overthrow of the Umayyad dynasty. Wheat is also represented but barley, being drought tolerant, appears to be the dominate cereal species, which may reflect a more passive agricultural system in the Post-Classical period with less investment in hydraulic engineering features like catch dams and wadi barriers. The weed species identified as those related to agricultural environments predominate in the assemblage from Humayma with 74% of all weeds identified being from this group (Fig. 8c). This data also lends support to active cultivation of cereal crops being carried out in the hinterland of Humayma (Ramsay 2013). Figure 8a. Ubiquity of cereal crops and by-products by occupation period at Humayma

Percentage of Samples with Taxa Present

Ubiquity of Cereal Grains and By-Products from Humayma through Time

70 60 50

Nabataean /Early Roman - 24 Samples

40 30

Early Roman /Late Roman - 18 samples

20 10

Late Roman/ Byzantine - 38 Samples

0 Post Classic - 19 Samples

HUMAN POPULATION INCREASE

371

Fruit and nut species are also good indicators of regional trade as many of the species identified at the study sites are not suited to grow in arid environments typical at the study sites. The data recovered from Humayma indicates that fig was present during all periods of occupation, which is not surprising as fig (noted above) stores and transports well. Grape, when dried, also stores well for trade, therefore its appearance during all phases at Humayma is reasonable. What is interesting is the distribution of fruit and nut species in the Late Roman and Byzantine period. In this period the significant rise in olive, grape and the appearance of walnut (Juglans regia) might relate to an increase in trade to the site or to Roman dietary preference. Olive and walnut would not have been economically feasible to produce at Humayma for trade due to the water requirements of these species. This does not however, rule out the presence of house gardens that may have had a small number of trees or vines that would have been irrigated on a small scale. Also, of note is the lack of species diversity during the Post-Classical period when all that was identified in the samples were grapes and dates, both of which preserve and transport well. Figure 8b. Ubiquity of Fruit and Nut species by occupation period at Humayma

Percentage of Samples with Taxa Present

Ubiquity of Fruit and Nut Taxa from Humayma Through Time

70 60 50 40 30 20 10 0

Nabataean /Early Roman 24 Samples Early Roman /Late Roman - 18 samples Late Roman/ Byzantine 38 Samples Post Classic - 19 Samples

372

JENNIFER RAMSAY

Bir Madhkur has clear evidence of local cereal cultivation from not only the archaeobotanical assemblage but also the archaeological features (Fig. 3) surrounding the site that relate to agriculture. However, the data analyzed to date is only from the Late Roman and Byzantine periods and as such will only be able to support arid agriculture in the region during these periods. Therefore, instead of ubiquity having been used for the analysis of the Bir Madhkur material, percent distribution was used (Fig. 9a). It is clear from the data that there was a significant proportion of not only cereal grains but also the by-products of crop processing, which supports the claim of local cultivation near Bir Madhkur. The weed and wild species also support a local agricultural economy as 61% of the assemblage was only found in cultivated field habitats (Fig. 9b) (Ramsay and Smith II 2013). Figure 8c. Percent distribution of weed and wild species by habitat at Humayma Humayma - Percent Distribution of Weed and Wild Species by Habitat

Rocky 16%

Hydrophyllic 3%

Sandy 7% Agricultural 74%

HUMAN POPULATION INCREASE Figure 9a. Percent distribution of cereal crops and by-products from the Late Roman/Byzantine period at Bir Madhkur Percent Distribution of Cereal Grains and Chaff from Bir Madhkur

Culm nodes 15%

Awn Fragments 3%

Barley 7%

Barley Rachis 13%

Wheats 18%

Ceral Grain Indet. 29% Wheat Rachis 14%

Millet 1%

Figure 9b. Percent distribution of weed and wild species from Bir Madhkur Percent Distribution of Weed and Wild Species by Habitat from Bir Madhkur

Open Dry Habitats 1%

Sandy Habitats 33%

Rocky places 5%

Field weeds 61%

373

374

JENNIFER RAMSAY

The remains of fruit and nut species were sparse with the strongest evidence of fig, which was common on all of the arid study sites, as well as very small quantities of olive, grape, date and nutshell. This data supports Bir Madhkur as an agricultural production area in the Wadi Araba, likely with a small population, but not a major trade center, although future evidence may draw a different picture. Figure 10. Percent distribution of cereal crops and by-products from ʿAyn Gharandal

Percentage of Samples with Taxa Present

Ubiquity of Cereal Grains and By-Products from ʿAyn Gharadnal 40.0 35.0 30.0 25.0 20.0 15.0 10.0 5.0 0.0

Finally, although very preliminary and not dated securely, archaeobotanical evidence from ʿAyn Gharandal does support the possibility of local cultivation occurring in the hinterland of the site. The ubiquity of cereal grain and especially cereal by-products like rachis fragments (i.e. 37.5% wheat and 31.25% barley) and culm nodes (37.5%) points to local crop processing (Fig. 10). The early assessment of weed species also supports local cultivation as the major of weeds identified to date are only found in agriculturally related contexts. As for fruit, only fig and date have been recovered to date and both are known to be commonly found as components of arid archaeobotanical assemblages in the region. With the hopes that more samples will be recovered from chronologically secure con-

HUMAN POPULATION INCREASE

375

texts in the future, material from ʿAyn Gharandal will likely aid in painting a picture of a much greener hinterland in antiquity.

CONCLUSIONS

There is a good deal of support for cereal cultivation being practiced in southern Jordan between the Nabataean and Islamic periods of occupation but it is clear that it is not a uniform pattern of agricultural intensity chronologically. The archaeobotanical remains provide direct evidence, which is supported by archeological remains of field walls, farm houses, threshing floors, catch dams and wadi barriers. It is notoriously difficult to determine the date of use for many of these features but recently Ramsay and Smith (2013) connected a farm house, threshing floor and field walls with ceramics in the farm house that dated to the late Roman/Byzantine period of occupation at the site of Bir Madhkur. This is significant as it may illustrate the wider regional use of these agricultural features at other sites, like Aila, Humayma and ʿAyn Gharandal, which were all occupied by similar cultures at the same time. Evidence for local cultivation of cereals (by-products and weeds) from the Nabataean and Early Roman periods to date was only found at Humayma. The natural springs near the site and the fact it was located on an early trade route may indicate that the Nabataean population of Humayma had become more sedentarized by this time, which would have allowed them to invest more in local agriculture that would not have been possible if they were semi-nomadic. The lack of archaeobotanically evidence of local cereal cultivation in the Petra civic center likely stems from it not being a processing area since there is ample evidence of cereal production in the hinterland of Petra that would have been required to support the civic population. There was a population increase in southern Jordan likely beginning in the Nabataean period corresponding to the areas importance in the incense trade and increasing into the Roman period with the surge in the ranks of the Roman army in the second though fifth centuries (Smith et al. 1997, 66–67; Parker 1997b, 190– 191). The sites of Aila, Humayma, Bir Madhkur and ʿAyn Gharandal had military forts installed during this period whose inhabitants would have needed augmented quantities of food, as a result of this, an increased demand for agricultural commodities would have been likely. Local cultivation on any scale may well

376

JENNIFER RAMSAY

have been a necessity at these sites to support the great sedentary population. This is not to say the populations of southern Jordan were relying solely on local agriculture, specifically cereal grains, to supply all of their needs however the data cited above (weeds, chaff, and field systems) points to at least some local cultivation. Strong support for this comes specifically from the weeds recovered from Petra, Humayma, Aila and Bir Madhkur like Lolium temulentum (rye grass), Malva sp. (mallow), Euphorbia sp. (spurge), Chenopodium sp. (goosefoot/fat hen), Silene sp. (catchfly), Medicago sp. (medick) and Trifolium sp. (clover), which are all common in cropped or fallow fields, gardens, pastures and waste areas. It is highly unlikely that the presence of agricultural weed species would represent the majority of the assemblages at all the study sites if cereal agriculture was not being carried out in the region. Clearly the presence of cereal grains found on sites through the different periods of occupation but more importantly the presence of the by-products of crop processing support the conclusion that there was a local agricultural economy. It is clear that wheat and barley were the most dominate cereal species during all periods of occupation. At later dates, millet (Panicum sp.; Setaria sp.) and rye also show up as components of the archaeobotanical assemblage. Although this data provides evidence for local agriculture supporting the subsistence regime of these sites, other factors that may account for the presents of cereal grains and chaff do need to be taken into account. For example, modern Bedouin populations practice passive cereal agriculture in the high land region above Petra and Humayma, which may have also been carried-out in the past as these areas would have been more suitable for cereal crops due to a slightly higher annual precipitation. Likewise, the inclusion of the by-products of crop processing may have occurred in samples as the result of their presence in animal dung, which may have been imported to sites as fuel. However, when considering the features related to water management or agriculture that have been documented on and surrounding these arid sites it does seem plausible that cereal agriculture was a part of the subsistence strategy of populations living in southern Jordan from the Roman through Islamic periods. As a result the landscape of southern Jordan in antiquity would have been significantly altered from barren desert to a more verdant and anthropomorphic environment reflective of

HUMAN POPULATION INCREASE

377

a greater sedentary population that required greater quantities of agricultural commodities to support their existence.

BIBLIOGRAPHY

Alcock, S., Knodell A. (2012). Landscapes North of Petra: The Petra Area and Wadi Silaysil Survey (Brown University Petra Archaeological Project, 2010–2011). In: L. Nehmé, Wadeson, L., eds, The Nabataeans in Focus: Current Archaeological Research at Petra. Oxford: Archaeopress, pp. 5–16. Al–Muheisen Z (1990). Exemples d’installations hydrauliques et des techniques d’irrigation dans le domaine nabatéen (PétraJordanie méridionale). In: Geyer B., ed.,Techniques et pratiques hydro-agricoles traditionnelles en domaine irrigué. Actes du Colloque de Damas 1987. Paris: Librairie Orientaliste P. Geuthner, pp 507–13. Ashkenazi, E., Y. Avni and G. Avni. (2012). A comprehensive characterization of ancient desert agricultural systems in the Negev Highlands of Israel. Journal of Arid Environments 86, 55–64. Avni, G., Avni, Y., and N. Porat. (2009). Ancient agriculture in the Negev Highlands – a reexamination. Cathedra 133, 14–44. (Hebrew). Barker, G. (2002). A tale of two deserts: Contrasting desertification histories on Rome’s desert frontiers. World Archaeology 33(3), Ancient Ecodisasters, 488–507. Barker, G. (2012). The desert and the sown: Nomad-farmer Interactions in the Wadi Faynan, southern Jordan. Journal of Arid Environments 86, 82–96. Bikai P M (2004). Beidha al-Amti. In Archaeology in Jordan, 2003 Season, by S.H. Savage, K.A. Zamora, and D.R. Keller. American Journal of Archaeology 108(3), 429–45. Cappers, R.T. J. (2006). Roman Footprints at Berenike: Archaeobotanical evidence of subsistence and trade in the eastern desert of Egypt. Los Angeles: Costen Institute of Archaeology. Dafni A., Levy, S and E. Lev. (2005). The ethnobotany of Christ’s Thorn Jujube (Ziziphus spina-christi) in Israel. Journal of Ethnobiology and Ethnomedicine 1, 8.

378

JENNIFER RAMSAY

Darby, R. and E. Darby. (2012). The ʿAyn Gharandal Archaeological Project: A Preliminary Report on the 2010 and 2011 Seasons. ADAJ 56, 405–22. Erickson-Gini, T. (2006). ‘Down to the Sea’: Nabataean colonisation in the Negev Highlands. In: P. Bienkowski and K. Galor, eds., Crossing the Rift. Resources, Routes, Settlement Patterns and Interaction in theWadi Arabah. Oxford: Oxbow, pp. 157–66. Erickson-Gini, T., (2010). Nabataean settlement and self-organized economy in the Central Negev. BAR-IS. 2054. Oxford: Archaeopress. Erickson-Gini, T., 2012. Nabataean agriculture: Myth and reality. Journal of Arid Environments 86, 50–54. Eusebius. Onomasticon. Ed. E. Klosetermann. 1904. Das Onomastikon der biblischen Ortsnamen. Hildsheim: G. Olms. [Reprint 1966]. Evenari, M., Aharoni, Y., Shanan, L., and N. Tadmor. (1958). The ancient desert agriculture of the Negev. III. Early Beginnings. IEJ 8, 231–68. Evenari, M., Shanan, L., Tadmor, N., and Y. Aharoni. (1961). Ancient agriculture in the Negev. Science 133 (3457), 979–96. Frank, F., (1934). Aus der Araba I: Reiseberichte. ZDPV 57, 191– 280. Glueck, N. (1935). Explorations in Eastern Palestine II. BASOR 15, 15–202. Graf, D. 1992. Nabataean Settlements and Roman Occupation in Arabia Petraea. SHAJ 4, 253–56. Hirschfeld, Y. and Y. Tepper. ( 2006). Columbaria towers and other structures in the environs of Shivta. Tel Aviv 33, 83–116. Joukowsky, M. (2001). Nabataean Petra. BASOR 234, 1–4. Kedar, Y. (1967). Ancient agriculture in the Negev mountains. Jerusalem: Mosad Bialik. [Hebrew]. Khouri, R. G. (1988). The Antiquities of the Jordan Rift Valley. Amman: Al-Kutba. King, G.R.D, Lenzen, C. J, and G.O. Rollefson. (1987). Survey of the Byzantine and Islamic sites in Jordan: Third season preliminary report (1982): The Wadi ‘Arabah. ADAJ 31, 199–215.

HUMAN POPULATION INCREASE

379

Knodell, A. and S. Alcock. (2011). Brown University Petra Archaeological Project: The 2010 Petra Area and Wadi Silaysil Survey. ADAJ 55, 489–508. Kouki, P. (2009). Archaeological evidence of land tenure in the Petra region, Jordan: Nabataean-Early Roman to Late Byzantine. Journal of Mediterranean Archaeology 22, 29–56. Lavento, M., Siiriäinen, A., Jansson, H., Kouki, P., Mukkala, A., Silvonen, S., an T. Tenhunen. (2004). The Intensive Survey around Jabal Harûn: Settlement History and Land Use in the Area. SHAJ 8, 225–35. Lavento, M., Kouki, P. Silvonen, S., Ynnilä, H. and M. Huotari. (2007a.) Discussions on Terrace Cultivation and its Relationship to the City of Petra, in Southern Jordan. SHAJ 9, 145–56. Lavento, M., Kouki, P, Eklund, A., Erving, A., Hertell, E., Junnilainen, H., Silvonen, S., Ynnilä, H. (2007b). The Finnish Jabal Harun Project Survey: Preliminary Report of the 2005 Season. ADAJ 51, 289–302. Mayerson, P. (1960). The ancient agricultural remains of Nessana and the Central Negeb. BASOR 160, 27–37. Mayerson, P. (1985). The Wine and Vineyards of Gaza in the Byzantine Period. BASOR 257, 75–80. Musil, A. (1907). Arabia Petraea II. Edom I. Vienna: in Komm. bei A. Hölder. National Master Water Plan of Jordan. (1977). Amman: Natural Resources Authority. Niemi, T. M. (2013). Regional Environmental Setting, Seismic History, and Natural Resources of Aqaba, Jordan. In: Parker and Smith. (2013). The Roman Aqaba Project Final Report. Volume I: The Regional Environment and the Regional Survey, pp. 33–79. Oleson, J. P. (1988). The Humayma Hydraulic Survey: Preliminary report of the 1987 season. ADAJ 32, 157–169. Oleson, J. P. (1990). The Humayma Hydraulic Survey: Preliminary report of the 1989 season. ADAJ 34, 285–312. Oleson, J.P. (1995). The Origins and Design of Nabataean WaterSupply Systems. SHAJ 5, 707–19.

380

JENNIFER RAMSAY

Oleson, J. P. (1997). Humaima. In: E. M. Meyers, ed., The Oxford Encyclopaedia of Archaeology in the Near East. Oxford: Oxford University Press, pp. 3.121–22. Oleson, J. P. (1997a). Landscape and cityscape in the Hismâ: the Resources of Ancient Humayma. SHAJ 6, 175–88. Oleson, J.P. (2001b). King, Emperor, Priest, and Caliph: Cultural Change at Hawara (Ancient al-Humayma) in the First Millennium AD. SHAJ 7, 569–80. Oleson, J.P. (2004). ‘Romanization’ at Hawara (Humayma)? The Character of ‘Roman’ Culture at a Desert Fortress. SHAJ 7, 353–60. Oleson, J.P. (2008). Social and Technological Strategies for the Design of Nabataean Water-Supply Systems in Hyper-Arid Environments. In: E. Hermon, ed., L’eau comme patrimoine; de la Méditerranée à l’Amerique du Nord. Québec: Les Presses de L’Université Laval, pp. 301–14. Oleson, J.P. (2010). Humayma Excavation Project, 1: Resources, History, and the Water-Supply System. Boston: American Schools of Oriental Research. Oleson, J. P., Reeves, M. B., Baker, G. S., de Bruijn, E., Gerber, Y., Nikolic, M., and A.N. Sherwood. (2008). Preliminary Report on Excavations at Humayma, Ancient Hawara, 2004 and 2005.ADAJ 52, 309–42. Oleson, J.P. and R. Schick. (2014). Humayma Excavation Project, 2: Nabataean Campground and Necropolis, Byzantine Churches, and Early Islamic Domestic Structures. Boston: American Schools of Oriental Research. Parker, S. T. (1996). The Roman Aqaba Project: The 1994 Campaign. ADAJ 40, 231–57. Parker, S. T. (1997a). Preliminary Report on the 1994 Season of the Roman Aqaba Project. BASOR 305, 19–44. Parker, S. T. (1997b). Human Settlement at the Northern Head of the Gulf of Aqaba: Evidence of Site Migration. SHAJ 6, 189– 93. Parker, S. T. (1998). The Roman Aqaba Project: The 1996 Campaign. ADAJ 42, 375–94.

HUMAN POPULATION INCREASE

381

Parker. S. T. (2000). The Roman Aqaba Project: The 1997 and 1998 Campaigns. ADAJ 44, 373–94. Parker, S. T. (2002). The Roman ʿAqaba Project: The 2000 Campaign. ADAJ 46, 409–28. Parker, S. T. (2003). The Roman ʿAqaba Project: The 2002 Campaign. ADAJ 47, 321–33. Parker, S. T. (2006). Roman Aila and the Wadi Arabah: an economic relationship. In: P. Bienkowski and K. Galor eds, Crossing the rift: Resources, routes, settlement patterns and interaction in the Wadi Arabah. Oxford: Oxbow, pp. 223–30. Parker, S. T. (2009a.) The Foundation of Aila: A Nabataean Port on the Red Sea. SHAJ 10, 685–90. Parker, S. T. (2009b. The Roman Port of Aila: Economic Connections with the Red Sea Littoral, in Red Sea IV: Connected Hinterlands. In: Blue, L, Cooper, J, Thomas, R, and J. Whiteright, eds., Connected Hinterlands: Proceedings of the Red Sea Project IV. BAR-IS 2052. Oxford: Archaeopress, pp. 79–84. Parker, S. T. (2009c). The Roman Frontier in Southern Arabia: A Synthesis of Recent Research. In: W. S. Hanson, ed., The Army and Frontiers of Rome: Papers offered to David Breeze on the occasion of his sixth-fifth birthday and his retirement from Historic Scotland, Journal of Roman Archaeology, Supplementary Series 74. JRA: Portsmouth, pp. 142–52. Parker, S. T. (2013). Introduction. In: Parker and Smith, eds., The Roman Aqaba Project Final Report. Volume I: The Regional Environment and the Regional Survey. ASOR Archaeological Reports 19. Boston: American Schools of Oriental Research, pp. 1–31. Parker, S. T. and A.M. Smith II, eds. (2013). The Roman Aqaba Project Final Report. Volume I: The Regional Environment and the Regional Survey. ASOR Archaeological Reports 19. Boston: American Schools of Oriental Research. Pearsall, D. (1983). Evaluating the Stability of Subsistence Strategies by Use of Paleoethnobotanical Data. Journal of Ethnobiology 3(2), 121–37. Pielou, E. (1969). An introduction to Mathematical Ecology. New York: Wiley & Sons.

382

JENNIFER RAMSAY

Popper, V. (1988). Quantitative Measurements in Palaeoethnobotany. In: C. Hastorf and V. Popper. eds., Current Palaeoethnobotany: Analytical methods and Cultural Interpretations of Archaeological Plant Remains. Chicago: The University of Chicago Press, pp. 53–71. Ramsay, J. H. (2013). Plant Remains. In: J. P. Oleson and R. Schick, eds., Humayma Excavation Project, 2: Nabatean Campground and Necropolis, Byzantine Churches, and Early Islamic Domestic Structures. Boston: American Schools of Oriental Research, pp. 353–60. Ramsay, J.H. and L. Bedal. (2015). Garden Variety Seeds: Botanical Remains from the Petra Garden and Pool Complex. Vegetation History and Archaeobotany,24 (5), 621–34. Ramsay, J.H. and S.T. Parker. (2016). A Diachronic Look at the Agricultural Economy at the Red Sea Port of Aila: An Archaeobotanical Case for Hinterland Production in Arid Environments. BASOR 376, 101–20. Ramsay J H., and A.M. Smith II. (2013). Desert Agriculture at Bir Madhkur: The First Archaeobotanical Evidence to Support the Timing and Scale of Agriculture in the Hinterland of Petra. Journal of Arid Environments 99, pp. 51–63. Rosen, A. M. (2007). Civilizing Climate: Social Response to Climate Change in the Ancient Near East. Lanham, MD: Altamira Press. Smith II, A. M. (2005a). Pathways, Roadways, and Highways: Networks of communication and exchange in Wadi Araba. NEA, 68(4), 180–89. Smith II, A. M. (2005b). Bir Madhkur Project: A Preliminary report on recent fieldwork. BASOR 340, 57–75. Smith II, A. M. (2010). Wadi Araba in Classical Late Antiquity: An historical geography. BAR-IS 2173. Oxford: Archaeopress. Smith II, A.M. (2014). The Southeast ʿAraba Archaeological Survey. In: S.T. Parker and A.H. Smith II, eds., The Roman Aqaba Project Final Report. Volume I: The Regional Environment and the Regional Survey. ASOR Archaeological Reports 19. Boston: American Schools of Oriental Research, pp. 103–289. Smith II, A. M. and T. M. Niemi. (1995). Results of the Southeast ʿArabah Archaeological Reconnaissance. ADAJ 38, 469–83.

HUMAN POPULATION INCREASE

383

Smith II, A. M, Niemi T. M. and M. Stevens. (1997). The Southeast Araba Archaeological Survey: A Preliminary Report of the 1994 Season. BASOR 305, 45–71. Tholbecq, L. (2001). The hinterland of Petra from the Edomite to the Islamic Periods: the Jabal ash-Sharah survey (1996–1997). SHAJ 7, 399–405. Tengberg, M. (2012). Beginnings and early history of date palm garden cultivation in the Middle East. Journal of Arid Environments 86, 139–47. Whitcomb, D. (1994). Ayla: Art and Industry in the Islamic Port of Aqaba. Chicago: Oriental Institute Museum. Zayadine, F. (1992). L’espace urbain du grand Pétra, les routes et les stations caravaniéres. ADAJ 36, 217–40. Zayadine, F. (1994). Ayla – ‘Aqaba in the Light of Recent Excavations. ADAJ 38, 485–505.

INDEX ʿAmr, K. 13–15 ʿAyn Ghazal. 5, 24 ʿAyn Jammam South. 18–19 ‘Ayn Gharandal. 349, 355–356, 359, 360, 374–75 ‘En Tamar. 83–84 363 CE earthquake. 122–123, 125, 129, 140, 142, 228, 280 551 CE earthquake. 128, 144–45, 175 abandonment. 12, 25, 51, 115, 125, 128, 131–39, 143, 250, 310n, 364 Abbahu, Rabbi. 298, 315n Abu Hudhud. 7, 9 actus. 249, 254, 257–259, 261– 67 AEP finds registry. 132 Agriculture. 5, 42–43, 55–56, 291, 293, 351, 353, 355–56, 362, 364, 369, 372, 376, Aila. xi, 77, 83, 125n, 151–161, 225–26, 228, 233, 243, 283, 291–93, 298, 318, 349, 351– 355, 358–60, 366–69, 375. Also see Aqaba. Aila, glass. 173-179 ‘Ain Lejjūn. 56, 227 ala. 245, 276, 279 Al-Basit. 13–14 Alexander Jannaeus. 197 Amaz Ascent. 81 Amazons. 334–37, 339, 342 Amphora, -ae. 159, 341

Anastasius, Emperor. 233, 282, 317 Anatolius of Ascalon. 307 Anonyma, heiress, wife of Mamilianus. 307–8 antechamber. 93, 98, 103–8, 111, 114–15, 119, 120-139, 142–43 Antioch on the Orontes. 130n, 228, 230, 277, 306n Antiochus III. 207 Antiochus XII Dionysos Epiphanes. 196, 207 Apion family, notables of Oxyrhynchus (Egypt). 308n Aqaba. xi, 23, 151, 153–54, 156, 158–59, 161–62, 173–74, 179, 225, 228n, 291, 327, 352, 358 Arabia. 54, 58, 63, 77–79, 82, 93, 141, 151, 190–94, 203, 208, 225–28, 238–40, 276, 278n, 279, 281, 328–29, 333, 354 Archaeological Survey of Israel. 299 Aretas III. 196–97, 199, 204–5, 354 Aretas IV. 194, 200, 204–5 Arsinoe I. 198 Arsinoe II. 200, 203 Arsinoe III. 198 assemblage. 11, 22, 48, 91, 93, 95–96, 103, 114, 116, 120–

385

386

ESSAYS IN HONOR OF S. THOMAS PARKER

24, 128, 130–43, 335, 360– 64, 370, 372, 374, 376 Assurbanipal. 52, 58 Assyrian. 51, 53–4, 58–59 Atargatis. 104n, 135 Augustus. 199, 208, 229, 333 Avdat. 118, 194, 197–98, 209– 10, also see Oboda Ayl to Ras an-Naqab Archaeolical Survey. 7, 18–19 Baʿja I. 15, 17 Baʿja II. 15–17 Baʿja V. 15 Babatha archive. 63 Baidha/Beidha. 13, 15–18, 24 199, 365 Bar Sauma. 280–81 Barley. 14, 22, 25, 367, 369–70, 373–74, 376 Basta. 13–14, 19 Beʾer Menuha. 129n, 356 Berenike, port. 228–29, 369 Berenike, Queen. 198 Bir Madhkur. 185–188, 349, 353, 356, 358-60, 372–75 Black Sea. 209, 334–37, 343 black-painted ware. 119–20 Bostra. 77, 257, 262, 274, 276– 78 Bouleutai=curiales. 294 boundaries. 298n, 327n, 329, 336–37, 340–43 By-products. 355, 361–63, 366– 376 Byrd, B.F. 15–16 Byzantine. 14, 63, 72, 75, 77, 79, 83, 94–95, 97–98, 103–4, 122, 128, 132–35, 138, 165, 173–75, 177, 185–86, 188, 194, 228, 273–74, 279-283, 293, 299, 349, 351–54, 356, 360, 362–67, 369–375 Cabotage. 310

Caesarea, 157, 291–318; Church, octagonal. 317–18; Commercial quarter. 317; Harbor. 293, 304–5; “Imperial revenue office.” 315; Mansions. 305, 310–11; agora. 314; Praetorium. 304, 313, 315–16; Proasteion. 311, Southwest Zone. 304, 309– 11, 314; Tavern (emporos). 314; Potters’ workshop. 313; Warehouses. 304–6, 309–10, 314–17 cage cup. 177, 179 Carmel, “Inner”. 298 Cereals. 359, 363, 369, 375 Chora=territory. 298–99 Christian. 63, 151, 155, 157, 161, 173, 230, 280–81, 302, 308, 317 Christianity. 230–31, 280, 327n, 337 city coins. 144 Civitas=Polis. 293–94, 340 Clark, G.A. 10–11 Claudius. 208, 335 clay lamp. 151–61, 172 Cleopatra III. 198, 201, 205 Cleopatra. 331–32 Clermont-Ganneau, Ch. 193, 198 cohors. 278 Constantius II. 145, 232 Consumer city. 305, 309, 318 Curiales, see bouleutai Da‘janiya. x, 258, 265 Damascus. 197, 207 Dana-Faynan-Ghuwayr Date, crop. 63, 365, 368–69, 371, 374 Dead Sea. 6, 19, 63–65, 67, 69, 75–83, 196, 354 Dead Sea Rift Valley. 14, 24–25 Decapolis. 204

INDEX decempeda. 254, 259 Dhiban. 36, 38–39, 47, 52, 57 Dhraʿ. 19 Dionysos, god. 198–201, 206 Dura-Europos. 207 Dushara. 195, 199–200 Dushares. 199 early 2nd c. earthquake. 118, 125, 139–141 Early Islamic. 95, 175, 349 earthquake. 78, 91, 118, 125, 139–141, 143–145, 175, 178, 186, 245, 280 Edom. 53–54, 63–64, 70, 80–81, 192 Egeria. 173 Egypt. 100n, 151–63, 167, 174, 199–209, 228–30, 238, 254, 260, 267, 270, 299, 302n, 308n, 310n, 315n, 351 Elagabalus. 130n, 277 el-Lejjūn. 42, 56, 151, 156–57, 163, 225–28, 233 ʿEn Yotvata. 118, 141, 186, 356 epidemic. 124–125, 141, 144 Epipalaeolithic. 8, 10, 12 esh-Sharif (Parker’s site 477), 37, 38, 46–47, 51, 58 Euergetism. 295 Evangelus. 303, 306n, 307–9, 316 Expositio totius mundi et terrarum. 312 Fajj el-‘Useikhir. 47 Fig, crop. 14, 22, 25, 364–71, 374 figurine. 24, 100n, 104n, 129, 135, 202, 313 fineware 116–125, 132, 135, 140, 144 Finley, M. 291–92, 295, 298, 305, 309, 311, 315–18 foot, Roman. 238, 255

387 fort, auxiliary. 237–45, 256–58, 261–67 “frog” lamp. 151–63 Gaia, suburb of Petra. 195, 199 Gebel, H. 13, 16–18 Gerasa (Jerash). 175, 203–4, 276 Ghor es-Safi. 63, 77, also see Zoara Ghuwayr I. 20–23 glass, vessels. 96, 120, 124, 165– 68, 171–172; Achaemenid style, 170–71, blown, 171; colorless. 174, 177–180; core-formed. 167; lamp. 171–74, 178–79; lenses. 165, 178–180; molded. 171; opaque 165–66; transparent 165, 167–171, 172–175, 177–180; window 165, 178, 180 Glueck, N. 35, 37, 42, 46–47, 53, 59, 64, 77, 140n, 349, 353 Gordion, glass. 170 grape. 119, 364–71, 374 groma. 248–50, 254 gromatici. 250, 259 Gruppe 7, Schmid. 121, 125, 132 Gruppes 8-9, Schmid. 121, 132 Hadrian. 130n, 227, 276, 277 Harpocrates. 104n, 135 Hasmonean. 80, 197 Hauarra. 237–42, 245–48, 253– 54, 259, 260, 265, 279, also see Hawara and Humayma Hauptmann, A. 23–25 Hawara, 241, 243, 250, also see Hauarra and Humayma Hellenistic. 73, 74, 80, 83, 95, 171, 193–209, 336 Hermansen, B.D. 16–17 Hermopolis, Egyptian nome. 308n, 309n

388

ESSAYS IN HONOR OF S. THOMAS PARKER

Herod Antipas. 208 Hinterland. 298, 300n, 351, 353, 355, 356, 363–64, 368–70, 374–75 Holocene. 6, 20, 21, 23 Ḥorvat ‘Aqav. 302 Horvat Dafit. 118 Horvat Hazaza. 118, 129n Household Excavations at Petra. 91 Huldu, Queen. 200 Humayma. 173, 175, 177–78, 237–253, 258, 265, 279, 349, 354, 359–60, 369–72, 375 also see Hauarra and Hawara Humayma, glass. 175, 177–78 incense altar. 100 Incense Road. 119, 121, 123, 124 India. 327–29, 332 interior lighting. 165, 171–79 international trade. 143–44, 327–330 Iron Age I. 36, 39, 43–45, 50, 57, 59 Iron Age II. 14, 39, 43, 46–59 Isis. 100n, 198, 200–5 Islamic. 38, 50, 52, 57, 58, 72, 95, 175, 191, 283, 349, 354, 355, 366, 375–76 Jalame, glass. 175 Jerash, see Gerasa. Jordan River. 6, 42 Josephus 195, 197, 244n, 252– 53, 273 Julia Domna. 130n Julian of Ascalon. 313 Jurf ad-Darawish. 12–13, 25 Justinian, Emperor. 233, 282, 295, 302–3, 307–8, 316, 328 Kafafi, Z. 6, 9 Karak. 35–59, 175 Karak Resources Project. 41, 56

Karanis, glass. 153, 155, 157, 161–62 Kastron Zadacathon. 279 Khirbat al-Hammam/Abu Ghrab. 7–9, 12 Khirbat edh-Dharih. 11 Khirbet ‘Arbid (Parker’s site 464), 37, 38, 48, 49–53 Khirbet el-Fityan. 46, 56–57, 264 Khirbet Medeinet ‘Aliya (Parker’s site 149). 37, 38, 43–45, 50 Khirbet et-Tannur. 119 Khirbet Thamayil (Parker’s site 456). 38, 48, 49, 51–53, 55 Kirkbride-Helbaek, D. 15 Kuniyeh Ascent. 66–70, 72, 75, 78 lamp, clay. 151–163, 172 Late Byzantine. 83, 95, 175 Late Nabataean. 122, 127, 134, 137, 274 Later House, Petra. 98, 103, 144–45 legio III Cyrenaica. 238, 276 Lejjūn. vii, x, 42, 56, 151, 153, 156–59, 161, 163, 175, 178, 225–28, 233, 244, 254, 258, 264–65 limen. 342 Limes, frontier. 3n, 35, 185n, 225, 228, 239, 273 Limes Arabicus Project. ix–x, 35– 39, 42–52, 55–56, 173 Lisan. 19, 76–78 Livia. 208 Lot’s monastery. 63 Luhit ascent. 64 Luxury/luxuria. 168, 170, 177, 302, 305, 314–15, 327–34, 336–43 Madā’in Ṣāliḥ. 190, 277 Madaba Map. 63, 65

INDEX Malichos I. 205 Malichos II. 190, 204 Mampsis. 81–82, 141, 258, 266 Marble Workshop, Petra. 106, 113, 119, 126, 128 Marcus Agrippa. 208 Mark Antony. 209, 338 Masada. 72, 77–78, 117–18 mensor/mensores. 250, 255, 259 Mezad ʿEn Rahel. 118, 141 Mezad Hazeva. 123, 356 Mezad Mahmal. 124 Mezad Neqarot. 124 Middle House, Petra. 94, 115, 142, 143 Middle Roman pottery. 115, 121, 133 Miller, J. 37, 42–43, 46–47, 49– 53 Moab. 35–36, 38, 42–43, 46, 50–55, 58–59, 63–66, 70, 75n, 80–81, 130n, 192, 197 Moyat ʿAwad. 119, 122, 124, 137, 356 Musil, A. 64–65, 75, 77–78, 352 Nabataean/Nabataeans. 14, 15, 17, 18, 72, 80, 83, 93, 116– 24, 126–34, 134–43, 173, 185–210, 238, 241–45, 264, 273–77, 281, 349, 351, 353– 55, 360–71, 375 Najjar, M. 21–22 Naqeb Amaz. 81 Naqeb Buwieb. 81 Naqeb ez-Zuweirah. 79–80 Neeley, M. 9, 12–13, 20 Negev, A. 142n, 193–94, 197 Negev, desert in southern Israel. 54, 63, 76n, 82, 141, 197, 351 Neolithic. 3–25 Nero. 179, 209, 329 Notitia Dignitatum. 238, 240, 279, 281, 356

389 numerus. 281 Nysa-Scythopolis. 203 Oboda. 118, 121–24, 129n, 135, 141–43, 194, also see Avdat Oboda pantry. 122–23, 135, 142 Obodas I. 195–98, 202 Obodas II. 195 , 202, 205 Obodas III. 195 Obodas the God. 193–210 Octavian. 208, also see Augustus Olive. 172, 299, 364–71, 374 Osiris. 155, 198, 202 Ousia (church endowment). 317 painted fine ware. 117, 119, 131, 135, 137–38 Painters’ Workshop, Petra. 93, 103, 105, 112, 114, 122, 124, 126–28, 135, 143 Palaestina Salutaris. 281 Parker, S. Thomas. vii–xii, 3n, 7, 23, 35–37, 39, 42–53, 57, 63n, 91n, 151, 161, 185n, 225n, 237n, 244, 273–74, 291–93, 318, 351–52 pastophorion. 175–77 Pella, glass. 175 perfumed oils. 102, 144 Perry, M. xii, 125 pertica. 254, 259 pes monetalis. 238, 253, 255, 259–267 Peterson, J. 9 Petra. xi–xii, 83, 185–86, 190, 191, 194–99, 202–4, 209, 232, 254, 260n, 265, 349–50, 353–55, 357, 359–60, 363, 375–76; Church. 228; edDeir. 195; ez-Zantur. 228, 232, 274; ez-Zantur, glass. 175; military. 273-83; North Ridge 124, 144; Papyri. 278; Pool and Garden complex. 363, 365–66; Qasr al-Bint.

390

ESSAYS IN HONOR OF S. THOMAS PARKER

278–79; region/hinterland 15–18, 26, 375; titles. 276– 77; Upper Market. 125; Urn Tomb (Jason’s Church). 281; also see Temple of the Winged Lions and Painters’ Workshop petrography. 151, 159 Philip I, Emperor. 121, 123, 130 photoamulet. 155, 157 Pleistocene. 12–13, 23 Pliny the Elder. 328–29 Polis see Civitas Polyani, K. 292n Polybius. 250–52, 255, 256, 259, 261 polycandela. 175 Porphyrion, village. 303 post-annexation period. 93, 116–24 Post-Classical period. 360–71 Procopius of Caesarea. 231, 293, 299, 302–10, 315–16 Provincia Arabia. 93, 239 Prudentius. 173 Pseudo-Hyginus. 242–45, 249, 255, 257, 259, 261 Ptolemy II Philadelphus. 198, 200, 203 Ptolemy III Euergetes. 198 Ptolemy IV Philopater. 198, 200 Ptolemy IX Soter II. 198 Ptolemy XII Auletes. 198 Qaseir et-Tamrah. 47 Qasr Abu el-Karaqa. 39 Qasr Abu el-Kharaqa (Parker’s site 110), 37, 38, 43 Qasr ed-Daba’a (Parker’s site 194), 36, 38, 48–49, 51–52 Qasr el-‘Al (Parker’s site 124). 37, 38 Qasr Saliya (Parker’s site 2). 37, 38, 44, 48, 51, 58 Qasra. 124

Qedarite Arabs. 53, 58–59 Rabbath-Moaba. 130n Rabbel I. 196–97, 204 Rabbel II. 195, 199–200, 205–6, 274 rainfall. 6, 41–42, 45, 56, 351–55 Ras al-Siq. 11 Rassam Cylinder. 54, 58 Red Sea. 151, 157, 228, 328n, 333, 337 Rollefson, G.O. 5–6, 9–10 Romans. 76, 171–72, 252, 254, 275–76, 328, 334, 339 Roman annexation. 93–94, 138, 141, 144, 194, 210 Roman Aqaba Project. xi, 173, 179, 288n, 327n, 352 Roman army. 63, 78, 144, 225, 231, 273, 275, 281, 375 Roman period. 35, 74–75, 77, 79, 82–83, 100, 102–3, 116n, 124–27, 131–34, 137–38, 141, 203, 274–79, 355–56, 362, 367–370, 375 Routledge, B. 37, 43–45, 49, 51– 52, 55–56 Sabra 1. 15, 18 Samaritans, Uprising. 302–3 Sauer, J. viii–ix Scythopolis, see Nysa- Scythopolis Sedentarization. 354 Septimus Severus. 130n Shaʿar Ramon. 122, 124 Shammakh to Ayl Archaeological Survey. 3, 7, 13–15 Shaqarat Mazyad. 15, 17–18 Shaqilat, Queen. 194 Simmons, A. 21–22 Smith II, A.M. 7, 23, 185, 353 soil-resistivity survey. 93 Southern Ghors. 3, 7, 19–20, 25 Southern Ghors and Northeast ʿArabah Archaeological Sur-

INDEX vey. 19–20 Spices (pepper, incense, cinnamon, myrrh). 100, 273, 328– 29, 331, 338, 351, 375 Sportulae. 316 Sumptuary laws. 330–34, 337 Tafila-Busayra Archaeological Survey. 3, 12–13 Tall Wadi Faynan. 20, 22 Tamar Pass. 81–82 Temple of the Winged Lions. 91-145 Tel Tanninim. 301 Tell Madaba. 36, 39 Thamudic. 191 The Wadi al-Hasa Archaeological Survey. 7–12 titulum. 245 Trajan. 114, 126, 132–33, 130, 227, 238, 242, 265, 274, 276 Transjordan Plateau. 3, 7, 20, 21, 25 Tsurim Ascent. 81 TWLCRM project. 97 Udhruh. 232–33, 279, 283 Umm al-Tawabin. 72 Umm el-Jimal. x, 225–26, 233, 262 Umm Rihan. 301, 303 unguentarium. 125, 130, 134 Upper Paleolithic. 10 Uranius. 194, 196, 198 Vabalathus. 120, 130n, Vegetius. 255–56 Veyne, P. 318 via nova Traiana. 57, 77, 227, 229, 239, 243, 245, 354 Wadi ʿAfra. 8

391 Wadi ʿArabah. 3, 7, 19, 23, 25, 76, 83, 118, 186, 189, 349, 353–56, 374 Wadi al-Hasa. 7–12, 25 Wadi ath-Thamad. 36, 38, 57 Wadi Faynan. 4, 20–25 Wadi Fidan. 19–20 Wadi Fifa. 19–20 Wadi Isal. 64, 75 Wadi Kuniyeh. 67–69 Wadi Laʿban. 11 Wadi Mujib. 39, 43–46, 54, 57 Wadi Musa. 13–16, 93, 145, 186, 195, 199 Wadi Ramm. 190 Waheeb, M. 18 Walnut. 371 Weber, M. 292n weeds. 362, 370, 373–76 wheat. 14, 22, 25, 338, 367, 369–70, 373–76 Wheeler-Kenyon method. 95 Whore of Babylon. 337–38, 343 Witches (Meroe, Medea). 336– 37, 339, 342–43 witchcraft. 327, 339, 343 Women. 154, 315, 327–343 Yarmoukian. 5–6, 17 Yarmuk River. 42 Zahrat adh-Dhra 2. 19 Zantur Phase 3. 116 Zeron Ascent. 79–80 Zerqa River. 42 Zoar Ascent. 64, 66–68, 70, 75, 78–79 Zoara. 63–66, 70, 74–79, 81–84, also see Ghor es-Safi