Crossing Jordan: North American Contributions to the Archaeology of Jordan 9781845532680, 9781845532697

Table of contents :
Cover
Half Title
Title Page
Copyright Page
Table of Contents
Prefaces
His Royal Highness Prince El Hassan Bin Talal
Dr Fawwaz al-Khraysheh, Director General, Department of Antiquities of Jordan
Dr Barbara A. Porter, Director, ACOR
Regional Archaeological Site Maps of Jordan-From Prehistory to the Islamic Periods
Theory
1. Thinking Globally and also Locally: Anthropology, History and Archaeology in the Study of Jordan's Past
History
2. The History of North American Archaeological Research in Jordan
Methodologies
3. Archaeological Site Surveying in Jordan: The North American Contribution
4. Controlling Space at the Regional Level: Geographic Information Systems (GIS) and JADIS in Jordan
5. On-Site GIS Digital Archaeology: GIS-based Excavation Recording in Southern Jordan
6. High Precision Radiocarbon Dating in Jordan
7. The Evolving Landscape of Jordan: The Contributions of Geoarchaeology and Paleoecology
8. Past Environments of the Jordan Plateau from the Paleolakes of the Eastern Desert
9. Conservation and Preservation of Archaeological Sites in Jordan: Archaeology Initiatives of the American Center of Oriental Research
Regional Archaeology-Deep-Time Studies Across Jordan
10. Ancient Metal Production and Social Change in Southern Jordan: The Edom Lowlands Regional Archaeology Projectand Hope for a UNESCO World Heritage Site in Faynan
11. How Crossing Jordan made the Difference: The Case of the Madaba Plains Project, 1967-2007
12. Tall Hisban: Palimpsest of Great and Little Traditions of Transjordan and the Ancient Near East
13. Tall al-'Umayri through the Ages
14. Highlights from the Heights of Jalul
15. The 'Iraq AI-Amir and Dhiban Plateau Regional Surveys
16. Investigating 5,000 Years of Urban History: The Tall Madaba Archaeological Project
17. Crossing Jordan: by way of the Karak plateau
18. Four Archaeological Surveys in Southern Jordan
Prehistoric Perspectives
19. Searching for Neanderthals and Finding Ourselves: Research at Tor Faraj
20. Neanderthals at the Crossroads: Middle Paleolithic Sites on the Madaba Plateau, Jordan
21. Midnight at the Oasis: The End of the Pleistocenein Wadi al-Hasa
22. Microliths and Mortuary Practices: New Perspectives on the Epipalaeolithic in Northern and Eastern Jordan
23. Crossing the Boundary to Domestication Economies: A Case Study from West-central Jordan
24. The Rediscovery of the Neolithic Period in Jordan
25. Late Prehistory in Wadi Ziqlab, al-Kura, Jordan: From Sedentism to Olive Oil Factories
26. Living on the Edge: Settlement and Abandonment in the Dead Sea Plain
27. Is Big Really Better? Life in the Resort Corridor-Ghwair I, a Small but Elaborate Neolithic Community in Southern Jordan
Bronze Age-Earliest Urbanism
28. Life in the Earliest Walled Towns on the Dead Sea Plain: Bab adh-Dhra' and an-Numayra
29. Death and Dying on the Dead Sea Plain: Fifa, Khirbat al-Khanazir, and Bab adh-Dhra' Cemeteries
30. Life at the Foundation of Bronze Age Civilization: Agrarian Villages in the Jordan Valley
31. Khirbet lskander: A City in Collapse at the End of the Early Bronze Age
32. A Landscape Approach to Craft and Agricultural Production: Tracking the Location of Early Bronze Age Manufacturing at al-Lajjun, Jordan
33. The Early Bronze Age City States of the Southern Levant: Neither Cities nor States
Early States and The Iron Age
34. Independent and Well-Connected: The Ammonite Territorial Kingdom in Iron Age II
35. Shepherds and Weavers in a 'Global Economy': Moab in Late Iron Age 11-Wadi ath-Thamad Project (Khirbat al-Mudayna)
36. The Power of Place: The Dhiban Community through the Ages
37. A Place In-Between: Khirbat al-Mudayna al-'Aiiya in the Early Iron Age
The Edge of Empire-Hellenistic and Roman Period
38. In Search of Hellenistic Petra: Excavations in the City Center
39. Shifting Places, Changing Faces: The Civic Statuary of Roman Jordan
40. Projecting Power on the Periphery: Rome's Arabian Frontier East of the Dead Sea
41. Beyond Frankincense and Myrrh: Reconstructing the Economy of Roman 'Aqaba
Nabatean Civilization and its Jordanian Heartland
42. Gods and Vineyards at Beidha
43. Nabataean Landscape and Power: Evidence from the Petra Garden and Pool Complex
44. Surprises at the Great Temple from 1993 to 2006
45. Beyond the Nabataean and Roman City: Surveying the Central and Southern Wadi 'Arabah
46. Luxury in the Desert: A Nabataean Palatial Residence at Wadi Ramm
47. Torn Asunder: Earthquakes at Qasr at-Tilah
Theocratic Empire-The Byzantine Period
48. Bioarchaeology of North Jordan: A Decade of Cooperative American and Jordanian Student Research
49. Beyond the Rock: Petra in the Sixth Century ce in the Lightof the Papyri
50. Petra's Churches: The Byzantines and Beyond
Islamic Civilization in Jordan
51. From Nabataean King to Abbasid Caliph: The Enduring Attraction of Hawara/al-Humayma,a Multi-cultural Site in Arabia Petraea
52. From Residence to Revolutionary Headquarters: The Early Islamic Qasr and Mosque Complex at al-Humayma and its 8th-century Context
53. Paradox of Power: Between Local and Imperial at Umm Al-Jimal
54. Peasants, Pilgrims, and the Body Politic: The Northern Jordan Project and the Landscapes of the Islamic Periods
Geographical Index
Subject Index
Index of Personal Names

Citation preview

CrossingJordan North American Contributions to the Archaeologyof Jordan

This page intentionally left blank

CrossingJordan North American Contributions to the Archaeologyof Jordan

edited by Thomas E. Levy, P.M. Michele Daviau, Randall W. Younker, & May Shaer

~ ~~o~;~~n~~:up LONDON AND NEW YORK

First published 2007 by Equinox, an imprint of Acumen Published 2014 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN 711 Third Avenue, New York, NY 10017, USA Routledge is an imprint of the Taylor & Francis Group, an informa business ©Thomas E. Levy, P.M. Michele Daviau, Randall W. Younker, and May Shaer 2007

All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Notices Practitioners and researchers must always rely on their own experience and knowledge in evaluating and using any information, methods, compounds, or experiments described herein. In using such information or methods they should be mindful of their own safety and the safety of others, including parties for whom they have a professional responsibility. To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume any liability for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas contained in the material herein. Library of Congress Cataloguing-in-Publication Data A catalogue record for this book is available from the Library of Congress

ISBN 13: 978-1-84553-268-0 (hbk) ISBN 13: 978-1-84553-269-7 (pbk) Typeset and Edited by Forthcoming Publications Ltd

Contents

Prefaces His Royal HighnessPrince El HassanBin Talal Dr Fawwazal-Khraysheh,Director General,Departmentof Antiquities of Jordan Dr BarbaraA. Porter,Director, ACOR RegionalArchaeologicalSite Maps of JordanFrom Prehistoryto the Islamic Periods

XI X Ill

XV

XVII

THEORY I.

Thinking Globally and also Locally: Anthropology, History and Archaeologyin the Study of Jordan'sPast 0ystein S. LaBianca

3

HISTORY 2.

The History of North AmericanArchaeologicalResearchin Jordan Nancy Lapp

15

METHODOLOGIES 3.

4.

5.

6.

ArchaeologicalSite Surveyingin Jordan: The North AmericanContribution Burton MacDonald

27

Controlling Spaceat the Regional Level: GeographicInformation Systems(GIS) and JADIS in Jordan StephenH. Savage

37

On-SiteGIS Digital Archaeology: GIS-basedExcavationRecordingin SouthernJordan ThomasE. Levy and Neil G. Smith

47

High PrecisionRadiocarbonDating in Jordan ThomasHigham

59

vi 7.

8.

9.

CROSSING JORDAN

The Evolving Landscapeof Jordan: The Contributionsof Geoarchaeology and Paleoecology Carlos E. Cordova

69

PastEnvironmentsof the JordanPlateau from the Paleolakesof the EasternDesert Caroline Davies

79

Conservationand Preservationof ArchaeologicalSites in Jordan: ArchaeologyInitiatives of the American Centerof Oriental Research Pierre M. Bikai

87

REGIONAL ARCHAEOLOGY-DEEP-TIMESTUDIES ACROSSJORDAN 10.

II.

Ancient Metal Productionand Social Changein SouthernJordan: The Edom Lowlands RegionalArchaeologyProject and Hope for a UNESCOWorld HeritageSite in Faynan ThomasE. Levy and MohammadNajjar

97

How CrossingJordanmadethe Difference: The Caseof the MadabaPlains Project, 1967-2007 LawrenceT. Geraty

107

Tall Hisban: Palimpsestof Greatand Little Traditions of Transjordanand the Ancient Near East 0.5. LaBiancaand B. Walker

111

13.

Tall ai-'Umayri through the Ages Larry G. Herr and DouglasR. Clark

121

14.

Highlights from the Heightsof Jalul Randall W. Younker

129

IS.

The 'Iraq AI-Amir and Dhiban PlateauRegionalSurveys Chang-HoC.Ji

137

16.

Investigating5,000 Years of Urban History: The Tall MadabaArchaeologicalProject Timothy P. Harrison, Debra Foran, andAndrew Graham

143

17.

CrossingJordan:by way of the Karak plateau GeraldL. Mattingly and JamesH. Pace

153

18.

Four ArchaeologicalSurveysin SouthernJordan Burton MacDonald

161

12.

CONTENTS

vii

PREHISTORICPERSPECTIVES Searchingfor Neanderthalsand Finding Ourselves: Researchat Tor Faraj Donald 0. Henry

171

Neanderthalsat the Crossroads: Middle PaleolithicSiteson the MadabaPlateau,Jordan Michael S. Bisson,April Nowell, Carlos E. Cordova,and ReginaKalchgruber

179

Midnight at the Oasis;The End of the Pleistocenein Wadi al-Hasa Nancy R. Coinmanand DeborahI. Olszewski

187

Microliths and Mortuary Practices: New Perspectiveson the Epipalaeolithicin Northernand EasternJordan Lisa A. Maher

195

Crossingthe Boundaryto DomesticationEconomies:A CaseStudyfrom West-centralJordan Michael P. Neeley and JaneD. Peterson

203

24.

The Rediscoveryof the Neolithic Period in Jordan Gary 0. Rollefson and Zeidan Kafafi

211

25.

Late Prehistoryin Wadi Ziqlab, ai-Kura, Jordan: From Sedentismto Olive Oil Factories E.B. Banning

219

Living on the Edge: Settlementand Abandonmentin the Dead SeaPlain Patricia L. Fall, StevenE. Falconer,and Phillip C. Edwards

225

Is Big Really Better?Life in the ResortCorridorGhwair I, a Small but ElaborateNeolithic Communityin SouthernJordan Alan Simmonsand MohammadNajjar

233

19.

20.

21. 22.

23.

26. 27.

BRONZE AGE-EARLIEST URBANISM 28.

29.

Life in the EarliestWalled Towns on the DeadSeaPlain: Bab adh-Dhra'and an-Numayra R. ThomasSchauband Meredith S. Chesson

245

Deathand Dying on the Dead SeaPlain: Fifa, Khirbat ai-Khanazir,and Bab adh-Dhra'Cemeteries Meredith S. Chessonand R. ThomasSchaub

253

viii 30.

3 I. 32.

33.

CROSSING JORDAN

Life at the Foundationof BronzeAge Civilization: Agrarian Villages in the JordanValley StevenE. Falconer,PatriciaL. Fall, andJenniferE. Jones

261

Khirbet lskander:A City in Collapseat the End of the Early BronzeAge SuzanneRichard andJesseC. Long, Jr.

269

A LandscapeApproachto Craft and Agricultural Production: Tracking the Location of Early BronzeAge Manufacturingat ai-Lajjun, Jordan JenniferE. Jones

277

The Early BronzeAge City Statesof the SouthernLevant: NeitherCities nor States StephenH. Savage,StevenE. Falconer,and Timothy P. Harrison

285

EARLY STATES AND THE IRON AGE 34.

35.

36. 37.

Independentand Well-Connected: The Ammonite Territorial Kingdom in Iron Age II P.M. Michele Daviau and Paul EugeneDion

301

Shepherdsand Weaversin a 'Global Economy': Moab in Late Iron Age 11-Wadi ath-ThamadProject{Khirbat ai-Mudayna) P.M. Michele Daviau and Robert Chadwick

309

The Powerof Place:The Dhiban Communitythrough the Ages Benjamin Porter,Bruce Routledge,Danielle Steen,and Firas al-Kawamlha

315

A PlaceIn-Between:Khirbat ai-Mudaynaai-'Aiiya in the Early Iron Age Bruce Routledgeand Benjamin Porter

323

THE EDGE OF EMPIRE-HELLENISTIC AND ROMAN PERIOD 38. 39. 40.

41.

In Searchof Hellenistic Petra:Excavationsin the City Center David F. Graf

333

Shifting Places,ChangingFaces:The Civic Statuaryof RomanJordan Elise A. Friedland

341

ProjectingPoweron the Periphery: Rome'sArabian Frontier Eastof the Dead Sea S. ThomasParker

349

Beyond Frankincenseand Myrrh: Reconstructingthe Economyof Roman'Aqaba S. ThomasParker

359

CONTENTS

ix

NABATEAN CIVILIZATION AND ITS JORDANIAN HEARTLAND 42.

Godsand Vineyardsat Beidha PatriciaM. Bikai, ChrysanthosKanellopoulos,and Shari Saunders

369

43.

NabataeanLandscapeand Power: Evidencefrom the PetraGardenand Pool Complex Leigh-Ann Bedal and JamesG. Schryver

375

44.

Surprisesat the GreatTemplefrom 1993 to 2006 Martha S. Joukowsky

385

45.

Beyondthe Nabataeanand RomanCity: Surveyingthe Centraland SouthernWadi 'Arabah Andrew M. Smith II

393

46.

Luxury in the Desert:A NabataeanPalatial Residenceat Wadi Ramm DennineDudley and M. BarbaraReeves

401

47.

Torn Asunder:Earthquakesat Qasrat-Tilah Tina M. Niemi

409

THEOCRATIC EMPIRE-THE BYZANTINE PERIOD 48.

49.

50.

Bioarchaeologyof North Jordan: A Decadeof CooperativeAmericanand JordanianStudentResearch JeromeC. Rose,Mahmoudel-Najjar, and Dolores L. Burke

419

Beyondthe Rock: Petrain the Sixth Centuryce in the Light of the Papyri Robert C. Caldwell and TraianosGagos

417

Petra'sChurches:The Byzantinesand Beyond Megan A. Perry and Patricia Maynor Bikai

435

ISLAMIC CIVILIZATION IN JORDAN 51.

52.

From NabataeanKing to Abbasid Caliph: The EnduringAttraction of Hawara/ai-Humayma, a Multi-cultural Site in Arabia Petraea John PeterOleson

447

From Residenceto RevolutionaryHeadquarters: The Early Islamic Qasr and MosqueComplexat ai-Humayma and its 8th-centuryContext RebeccaM. Foote

457

x 53.

CROSSING JORDAN Paradox of Power: Between Local and Imperial at Umm AI-Jimal

Bert de Vries 54.

467

Peasants, Pilgrims, and the Body Politic: The Northern Jordan Project and the Landscapes of the Islamic Periods

BethanyJ. Walker

GeographicalIndex SubjectIndex Index of PersonalNames

473

481 485 491

Preface H.R.H. Prince El Hassan Bin Talal

Ever since the epic explorationsof the JordanRiver and Dead Seaby US Naval LieutenantW.F. Lynch in 1847,North Americanexplorersand archaeologistshavemadea profoundcontributionto environmental,historical, and archaeologicalresearchin the land of Jordan.Following the 1967 Middle East war and the establishmentof the American Centerof Oriental Research(ACOR) in Amman,the North Americancontributionto archaeologicalresearchin Jordanhasgrown exponentially. SinceACOR was foundedin 1968,I havebeenespeciallyimpressedby the many partnerships that haveevolvedover the yearsbetweenthe Departmentof Antiquities of Jordan,the universities of Jordan,and so many North Americanarchaeologistswho work underthe umbrellaof ACOR in Jordan.The many papersin this book, CrossingJordan-North American Contributions to the Archaeologyofjordan, are the bestevidenceof the rich contribution our North American friends have madein revealing,researching,and conservingJordan'srich past. Although Jordanis a fairly small country (ca. 100,000km2 ), as part of the land bridge that links the continentsof Africa and SouthwestAsia, it containsa remarkableamountof climatic, topographic, and ethnographicdiversity. The theme of this year's International Conferenceon the History andArchaeologyof Jordanin Washington,DC-CrossingJordan'-accuratelyreflectsthe diversity of environmentsand culturesthat have traversedthis land from remoteprehistorictimes to the present.Eachof thesepeopleshaveleft important'cultural footprints' on Jordan,and North American researchershave playeda key role in studyingthesesocial changesthrough time. Jordan'stopographicdiversity can bestbe seenby crossingthe land from north to south. The fertile Mediterraneanstrip of land in the northwest(ancientGilead) extendssouththroughcenter of the country (ancientAmmon),following the Transjordanianplateaulike a seriesof islandsinto the desertregionsof ancientMoab, and,finally, to Edamin the remotesouth.TheseMediterranean 'islands' contrast markedly with the black volcanic basalt of the Black 'Harra' desert in the northeast,and the spectacularNubian sandstonedesertsof Moab and Edam to the south. To the southeastis wherethe magnificentNefud desertof Arabia begins,punctuatedby the barrenbrown mud-flat desertsof the Jafr region. To the southeastof the ruggedmountainsof Moab, Jordansits on the edgeof the Arabian Bedouin heartlandof Wadi Sarhan.The JordanValley and Dead Sea transform,with its many geologicalfaults, have helpeddefine Jordantopographically,producing perhapsthe most beautiful wadis and arid zone landscapein the Middle East. Thesewadis and valleys are where 'crossingJordan'took place from prehistorictimes to the present. The recordof thoseculturesand peopleswho havecrossedJordanis remarkablefor its diversity stretchingback to someof the worlds earliestand greatestcivilizations and historically important local states.As shownby the major North Americanarchaeologicalresearchprojectsin Jordan,the country hasbeenhost to indigenousstatesand kingdomsas well as many of the greatcivilizations. Someof thesegreat empiresinclude ancientMesopotamia(Assyria and Babylon), ancientEgypt, ancientPersia,ancientGreece,Rome, Byzantium, Islam (the Umayyads,'Abbasids,Fatimids,and Mamluks),the Mongols, Crusaders,Ottomans,and modernEuropeans.Equally, if not moresignificant for Jordan,are the southernLevant's own indigenouskingdoms and statesthat made their mark on the region'scultural diversity. Someof theseincludeAmmonite,Moabite,Edomite,Phoenician, Israelite,Nabatean,andGhassanidkingdomsandcivilizations. This tapestryof cultureswho

xii

CROSSING JORDAN

have envelopedJordan have also been accompaniedby the sacred heritage of a multitude of Prophetsand saintsincluding Abraham,Moses,Jesus,and the ProphetMohammad(Blessingsand Peacebe uponthemall}-giving the cultural landscapeof Jordana spiritual dimensionthat continues to resonatetoday. This 'deeptime' heritagecontinuesto contributeto Jordan'sremarkablecontrastsin religious, ethnic, and social charactertoday.As a long-time observerof ACOR'sresearchers, it givesme great pleasureto congratulateour North American colleagueson more than 150 years of researchin Jordan.We look forward to even closer partnershipsin the 21st century.

Preface Dr. Fawwaz AI-Khraysheh Director General, Department of Antiquities of Jordan

Jordanhas alwayswelcomedarchaeologistsand researchersfrom all over the world and hasbeen the scenefor the unravelingof remarkablehistoricalepochs,from the prehistoricperiodto Ottoman times. Jordan'sbelief in multiculturalism and cross-culturalinteraction is deeply reflected in the emergingculturesof the country since the dawn of history. North Americantravelersand archaeologistsstartedtheir journeyof exploringJordan150 years ago,contributingto a greaterunderstandingof the manyculturesthat crossedthis country with its rich archaeologicalheritage.Their researchand archaeologicalworks have been numerousand cover vast areas.Jordanianarchaeology,spanningover 150,000yearsof prehistoryand history, tells us aboutthe diversesocietiesthat werecreatedduring different eras,with their respectivearts, architecture,inscriptive materials,and methodsof production;in other words, their way of life. The Departmentof Antiquities of Jordan has always encouragedcooperationand cultural exchangesinceits establishmentin 1928,andtakespride in its successfulpartnershipsand collaborationswith the internationalarchaeologicalinstitutionsand missions.Since1968,it hasbeenclosely collaboratingwith the American Centerof Oriental Research(ACOR), which has hostedNorth Americanarchaeologistsworking in the country. ACOR hassetthe scenefor a numberof archaeological andconservationprojectsin Jordan.With over 140 affiliated projects,it hasgreatlycontributedto enrichingour understandingof our cultural heritageandconservingit for future generations. This book, CrossingJordan-NorthAmerican Contributions to the Archaeologyof jordan, is launchedon the occasionof the Tenth InternationalConferenceon the History andArchaeologyof Jordan,with the sametheme'CrossingJordan'.This internationalgatheringwas initiated in 1980 by His Royal HighnessPrinceEl-HassanBin Talal, patronof the conference,and hasbeenheld at a different host country every three years.It hasbeenan importantvenuefor the exchangeof ideas andthe presentationof invaluableresearchon the history and archaeologyof Jordan.This year, it is organizedby ACOR in cooperationwith the AmericanSchoolsof Oriental Research(ASOR) and is taking place in Washington,DC. CrossingJordan-NorthAmerican Contributionsto the Archaeologyof jordan consistsof 54 chaptersthat highlight the contributionsof North American archaeologiststo the archaeologyof Jordan.With its informative and richly illustratedtext, it will be an invaluableresourcenot only for scholars,but also for all individuals who are interestedin expandingtheir knowledgeof Jordanian and Middle Easternarchaeology.Scholarlypublicationsand the disseminationof information are consideredthe essenceof archaeology,and hencethis book is a significant addition to the rich collection of publishedworks by well-renownedscholars. I would like to take this opportunityto thank all of the North Americanarchaeologists,historians,andscholarswho worked diligently throughoutJordan,contributingto our understandingand safeguardingof the Jordaniancultural heritage,as reflectedthroughoutthis valuablemanuscript.I would like to expressmy sinceregratitudeto ACOR and ASOR for their continuousefforts and supportfor Jordanianarchaeology.Finally, my appreciationgoesto the authorsand editorsof this volume whosededicationand diligenceled to the productionof this up-to-date,authoritative,and informative book.

This page intentionally left blank

Preface Dr Barbara A. Porter Director, ACOR

When informed by ArtemisJoukowsky,the Presidentof the Board of ACOR, in February2005 that I had beenselectedto be PierreBikai's successor,I was awarethat with that responsibilitycamethe role of organizingthe 1Oth InternationalConferenceon the History and Archaeologyof Jordan (ICHAJ) initiated by Pierreat ICHAJ 8 in Sydneyin 2001. Both co-chairsof the North American Committee,Arte and Doug Clark, let me know what the nearfuture would bring. When I cameto Amman in June2005 to checkout ACOR (and be checkedout), I was able to meetwith Doug on the ACOR patio-a spacewell known to most authorsof this book-andlearn of the progress made thus far. Doug and I have been the co-organizerscovering both sides of the Atlantic. Preliminary planning continuedapaceduring the fall of 2005 and has been a central focus ever since, both before and after 1 April 2006, the date I officially becamedirector. My first meetingwith many North American archaeologistsactive in Jordantook place at the Dig Directors' meetingin Philadelphiain November2005. It was there that Tom Levy presented the idea of this importantpublication,which was to be in keepingwith the tradition of the volume producedfor ICHAJ 8 in Australia. The realizationof this effort with its shortturn-aroundtime has beenthe productof hard work on the part of the editorsand contributors.As I review all of their names,it pleasesme that in one year I have met more than two-thirds of them due to the central role ACOR playsin assistingNorth Americanarchaeologistsin their work in Jordan.ICHAJ 10 has also given ChristopherA. Tuttle, the AssistantDirector in Amman,and me an excellentopportunity to connectwith colleaguesin Jordan.We are proud of ACOR's efforts to ensurethat manyof them are able to come to the United Statesfor the conference. In Nancy Lapp'schapteron history, with the story that startsin 1968,sheoutlinesin greatdetail the history and importanceof ACOR and invokes the namesof those who have madethis center what it is today. In the future yearsof my directorship,when the conferencewill itself be part of A COR'shistory, it will be a pleasureto look back andreflect on our role in bringing ICHAJ 10 to Washington. This dreamhasbeenrealizedsix yearslater, in large part due to the efforts and sponsorshipof AmbassadorEdward Gnehm,the Kuwait Professorof Gulf and Arabian PeninsulaAffairs at the Elliott School of InternationalAffairs of GeorgeWashingtonUniversity, the actual settingfor the conference.Five monthsfrom the writing of thesecomments,ICHAJ 10 will be over and therewill be many peopleto thank for makingit happen-frompatronsto caterers.We shall do so with great pleasurein Washington,DC, wheresometwo hundredpeoplewill conveneto shareand learn from eachother and to interweavedifferent generations,nationalities,and areasof expertise.

This page intentionally left blank

Regional Archaeological Site Maps of Jordan-From Prehistory to the Islamic Periods All mapsin this sectionwereproducedby StephenSavage,Geo-Archaeologica/Information Applications (GAIA) Lab, ArizonaStateUniversity

----

0

100

200----·300

Archaeological Sites in Jordan Paleohthte through Kebaran Perlods

400 Kilometers

+ s

xviii

CROSSING JORDAN

..

. .,

",

I . Jawa \ )

\3.,

..

"

0

100

200

400 Kilometers

300 Eiova~~Jx,"'":'l

Archaeologica l Sites in Jordan Neolithic through Chalcolithic Periods

~

1200.,1500mol 900to 1200msl eclOk> 900mol 300.,600mol 010300 msl

·30011)0mol < ~300 msl

xix

REGIONAL ARCHAEOLOGICAL SITE MAPS OF JORDAN

1 Jawa

0

100

200

.

- ~

~ ~

ltolf-:---:--~-t--

1 Ain al-Asad(Kelso and Rollefson1989) 2 Wadi ash-Shallalah(Cordova2007) 3 Wadi Kafrein (Vita-Finzi andDimbleby 1971) • 4 Wadi al-Wala(Cordova2007) 5 Wadi Faynan(Hunt et al. 2004) 6 Beidba(Fish 1989) 7 Petra-WadiMusa(Falll990) 8 JabalQalkba(Emery-Barbier1995) 9 Ain Abu-Nukhaylah(Scott-Cummings2001) 10 Ammanarea(Al-Eisawi and Dajaoi 1987, 1988) 11 SouthernHighlandsto Wadi Araba(Daviesand Fall 2001) --112 WestJordantransect(El-Moslimany 1983) Figure 3. Location of pollen and phytolith studies

74

CROSSING JORDAN

deposits,which havebeenthe depositspreferredby traditional palynologists,are virtually absentin Jordan.However, the few pollen studiescarried out in Jordanshow that pollen can be obtained from a variety of depositsand moderncontexts(Fig. 2). The works by Emery-Barbier(1995) in rock shelterdepositsof southJordan,and by Hunt eta/. (2004) in alluvial depositsin the Wadi Faynan region prove that pollen is sometimespreservedin hyperaridregions.The modernpollen transect that Daviesand Fall (2001) laid alongthe altitudinal gradientfrom Wadi 'Arabahto the Highlands aboveDana,showhow pollen taxacanbe linked to different environments,despiteall the problems of contaminationof exotic pollen grains and the effectsof plants that overproducepollen. Pollen recordsfrom alluvial depositsand paleosolsin Wadi al-Wala, nearthe Early Bronzesite of Khirbat Iskandar,and Wadi ash-Shallalah,nearthe Chalcolithic-EarlyBronzesite of Khirbat azZayraqun,provide two separatelocal sequences of vegetationchangeand agriculturaldevelopment during the middle and late Holocene,which correlatewith regional vegetationchangein pollen diagramsfrom lakes elsewherein the Levant (Cordova2007). Opal phytoliths, often referredto as phytoliths, are silica cells and bodiesformed in the tissues of someplants.Although the organicpartsof plantsdecayafter they die, opal phytoliths survive to becomeone of the multiple microscopicparticles of soil. The contribution of opal phytoliths is more promisingthan pollen, merely becausephytolithswithstanddeteriorationbetterthan pollen, particularly in arid lands. Although a techniquedevelopedmore recently than pollen research, phytolith researchin the Near Easthasgainedmomentum,particularlywhenappliedto the studyof wild and domesticatedgrasses.In Jordan, phytoliths have been obtained from archaeological depositsof prehistoricsites in Wadi Ziqlab and various sites in southJordan(Fig. 3). Applicationsof phytolith research,however,shouldincludethe distributionof cool season(C3) and warm season(C4) grasses,as well as phytoliths diagnostic of certain plants. In addition to grasses,a large numberof woody plantsin the Levant producediagnosticphytoliths (Albert and Weiner 2001), which are of great use in environmentswhere plant macro-remainsare not preserved.Motivated by the potentialof opal phytoliths, I haveinitiated a studyof silica productionin the flora of the four vegetationregionsof Jordan.One of the main objectivesof this researchis to tracevegetationchangein relation to climate and herbivoresover time. One of the long-termgoals of this project is the assessment of the influence of African savannahvegetationin the southern Levant, particularly during warm and wet periods.Throughthis approachI hopeto gatherenough datato testhypothesesregardingthe possibleinfluencesof summerrain resultingfrom an enhanced monsoonover theIndian Oceanduring certain times in the Pleistoceneand Holocene. On the issueof plant-herbivorerelations,two new techniquesare currently appliedto ecofacts and soils from the Pre-PotteryNeolithic site of 'Ayn Abu-Nukhaylain southJordan.Oneof themis the useof spherulites,which are microscopiccalciumconcretionsformed in dungof animals(Albert and Henry 2005). This is an importantissuerelatedto soils and sedimentsthat held domesticated bovinesand ovines. Another techniquecurrently applied to the site and the region is the study of oxygen, carbon and strontium isotopes, currently carried out by Marie Balasse from the La SorbonneUniversity in Paris.

Research Issuesand Questions In addition to poor preservationof certainbiological remains,anotherbig problemthat we still face in geoarchaeologyand paleoecologyis the lack of a tight chronology. In part this is due to the scarcity of radiocarbondates,coupledwith the scarcity of sequencesat all. In part this problem ariseswhen the applicationof radiocarbondating is difficult in areaspoor in carbon,as is the case of dryland deposits.However, the use of Optically StimulatedLuminescenceand other forms of luminescence(Thermal Luminescenceand Infrared StimulatedLuminescence)offer an immense

THE EVOLVING LANDSCAPE OF JORDAN

75

potentialin mostdepositsin Jordan.The work of Munro, Morgan,andJobling (1997) in southern Jordan and the recent work in Middle Paleolithic sites (Nowell et al., this volume), and the aforementionedsite of 'Ayn Abu Nukhayla, showsthat the luminescencetechniquesare gaining acceptanceand popularity in geoarchaeologicalresearch.The use of Uranium series has been applied to depositsmantling former lake-level benchesin southeastern Jordan(Abed et al. 2000). The questions geoarchaoelogistsand paleoecologistsare trying to answer in Jordanian archaeologyvary dependingon the periodsin question.For prehistoriansworking in the Lower and Middle Paleolithic it refers to the preservationof material and particularly in finding remainsof hominids,a problemthat is thwarteddue to the widespreaddisturbanceof cavedeposits(Nowell et a!., this volume). The problem is that many caveshave yet to be testedfor remains.Also, another important aspecthere is the amount of erosion that has obliterated open sites on the plateau (Cordovaeta!. 2005). In the al-Jafr and al-Azraq basins,evidenceof climatesand vegetationdifferent from today has come from faunal remainsassociatedwith large amountsof Acheulian and Levallois materials. Thereis, however,not yet stratigraphicresolutionto pinpoint the wet vs. dry periods.The suggestion that climate changesbroughta migration of African flora and fauna, including hominids, has beenposed(Tchernov1988; Cordova2007). However,a tight stratigraphicwork with numerous radiocarbondatesis neededto reconstructlandscapechangethat could have beensignificant for hominid migrationsfrom the African continent. Very importantalso in Paleolithicresearchis the transitionfrom the Middle Paleolithicto Upper Paleolithic,which is markedby the apparentdismissalof Neandertalsand the takeoverof modern anatomicalhumans. A greatdeal of researchis beingdirectedto the late Pleistoceneand early Holocene,periodsthat encompasses many cultural changes,from the transitionof foragersto specializedhunter-gatherers to the domesticationof animalsand plants(Cordova2007). Regionalrecordsshowthat thesewere times of intensechangeon a global scale. The other main problem is the questionof what causedthe decline of the Early Bronze Age culture. This issuein particularhasbeenaddressedthroughnumerousgeoarchaeological studieson both sides of the Rift and in the Upper Mesopotamia.Although climate changeseemsobvious, many other changesof social and economicnature obscurethe real impact of the environment. Evidenceof widespreadsynchronousfloodplain deteriorationat the end of the 3rd millennium BCE hasbeenreportedfrom streamvalleys in the Levant, suggestinga radical changein hydrology that may be the resultof climatechange(Cordova2007). Possibilitiesof waterand soil mismanagement, however,could have played an important role in this regional geomorphicevent. Other periods of interest in more recent times include the changesthat occurredduring the secondhalf of the 1st millennium CE, particularlyaroundthe Islamizationof the region,a topic that has creatednumerouscontroversiesdue to the lack of information on climate change.

Concluding Remarks Interestsin reconstructingthe environmentsaroundhumangroupsin the Levanthavecalledupona variety of specialists.Thus, the participationof geomorphologists,geologists,botanists,and other paleoecologistsin archaeologicalresearchhasbroadenedour understandingof the paleolandscapes of Jordanduring different periods. But despitethe rampantincreasein geoarchaeological and paleoecologicalresearchin Jordan, many researchproblemsare still to be solved. One of them is the scarcity of absolutedatesin geoarchaeological stratigraphicsequences, which hasbeenhinderedby the difficulties in applying radiocarbondatingto organic-poordeposits.This problemcan be overcomeby applyingalternative

76

CROSSING JORDAN

datingmethodssuchas thoseof the luminescencegroup. Relativedatingtechniquescanbe useful in building hypothesesto test numeric methods.One of these techniquesis the use of pedogenic development,which could be very useful for prehistoricperiods. Problemsin paleoecologyrangefrom poor preservationto areasin spaceand time for which there are no data.Although Quaternarysedimentsin Jordanare relatively poor in well-preserved pollen, studiesshowthat thereare potentialsin depositsthat havehardly beentested.They include travertines,hyrax middens,and a seriesof cultural deposits.Additionally, the lack of pollen requires the use of techniquessuchas phytoliths, which has a great potential. However, as is the caseelsewhere,the implementationof phytolith researchprotocolswill requirelarge amountsof testingand experiment.

References Abed, A., P. Carbone],J. Collina-Girard, M. Fontugne,N. Petit-Marie,j.C. Reyss,and S. Yasin (2000) Un paleolacdu dernier interglaciaire pleistocenedans !'extreme-sudhyperaridede Ia Jordanie.Earth and Planetary Sciences330: 259-64. Abed A.M., and F.F. Helmdach(1981) Biostratigraphyand Mineralogy of the Lisan Series(Pleistocene)in the JordanValley. Berliner Geowissenschaft AbhandlungenA 23: 123-33. Abed, A.M., and R. Yaghan (2000) On the Paleoclimateof Jordan during the Last Glacial Maximum. Palaeogreography,Palaeoclimatology,Palaeoecology160: 23-33. Albert, R.M., and D.O. Henry (2005) Herding and Agricultural Activitites at the Early Neolithic Site of Ayn Abu Nukhayla (Wadi Rum, Jordan):The Resultsof Phytolith and SpheruliteAnalyses.Paleorient30: 8192. Albert, R.M., and S. Weiner (2001) Study of Phytolithsin PrehistoricAsh Layers:A QuantitativeApproach.In Phytoliths: Applications in Earth Sciencesand Human History, edited by J.D. Meunier and F. Coline (Rotterdam:A.A. Balkema):251-66. Al-Eisawi, D. (1985) Vegetationin Jordan.In Hadidi 1985: 45-48. Al-Eisawi, D., and B. Dajani (1987) A Study of Airborne Pollen Grainsin Amman,Jordan.Grana 26: 231-38. -(1988) Airborne Pollen of Jordan.Grana 27: 219-27. Barker, G.W., R. Adams, O.H. Creighton, D. Crook, D.D. Gilbertson, J.P. Grattan, C.O. Hunt, D.J. Mattingly, S.J. McLaren, H.A. Mohammed,P. Newson,C. Palmer,F.B. Pyatt, T.E.G. Reynolds,and R. Tomber (1999) Environmentand Land Use in the Wadi Faynan,SouthernJordan:The Third Seasonof Geoarchaeologyand LandscapeArchaeology(1998). Levant 31: 255-92. Besan' . . . . . . # , ,..... .....-......::::- , I; ... 15,000years).The archaeologicalfeaturesof Qasrat-Tilah wereconstructedin Qr2 deposits,andthe braidedstreamdepositsof Q0 accumulated after construction of the site. The agricultural fields, site architecture,and the postoccupationalnaturaldepositsare designatedQarbecausehumansettlementinfluencessedimentation processes(Fig. 1C).

New Survey and Excavations Four seasonsof archaeologicalexcavationand architecturalsurvey(1999, 2001, 2002, and 2003) were conductedat Qasrat-Tilah as part of the Wadi 'Arabah EarthquakeProject.Archaeological excavationsfocusedon exposingthe earthquake-damaged aqueductleadingfrom the settlingpool at the baseof the southwestcornerof the birkeh. In conjunctionwith the excavation,and in order to gain a comprehensiveview of the waterworksstructuresrelatedto the birkeh (Fig. 2), we mapped the aqueductsin Wadi at-Tilah to the eastof the site (Area B) and the walls and aqueductsin the agriculturalfields to the west of the site (Area C). The areaof the collapsedfort (Area D) was also surveyedsincethe relationshipof the multiple phasesof aqueductto the fort is not clear. A section of the southeastcornertower of the Area D fort exposedin the Wadi at-Tilah bankwasstudied.We madedetailedelevationand site plans of the reservoir(Area E) and other ruined structuresat the site using a total station (an electronictheodoliteequippedwith an electronicdistancemeasuring device).

Birkeh and Adjoining Structures The often-cited dimensionsof the standingmasonrywater reservoirof 34.2 x 33.6 mare taken from Glueck (1935: 12). From our plan map of the structure,it is clear that thesemeasurements were taken from the top of the westernwall and from the baseof the northernwall, respectively. Becausethe northwest corner of the birkeh has been displacedby earthquakefaulting, these measurements do not representthe original dimensions.It is not possibleto achievepreciseexterior measurements,due to the aforementionedcrack and telescoping,but our reconstructedinside dimensionsof the birkeh basedon the preservedsouthwestand southeastcorners(Fig. 3) indicate that the inside dimensionsof the reservoirwere 25.5 m2 • The north and southwalls of the birkeh are not uniform in thickness.They taperfrom a narrow width of about2.5 m wherethe structureabutsthe bedrockto the east,andthickenwestward,until the westwall is ca. 4 m thick at the top, and over 5 m thick at the base.The interior facing wall has largely collapsedtoward the centerof the birkeh exposingthe rubble core. Two aqueductsthat originate in the adjacentWadi at-Tilah feed the birkeh at the southeast cornerand in the middle of the eastwall. The 70-cm-wide channelflows alongthe top of the south wall and at the southwestcorner water drops down to a settling pool at the baseof the southwest corneror flows into the birkeh. The outflow from the settling pool collectsand water flows into a north-trendingaqueductconstructedbelow the surface.In excavationof this feature,we did not identify any pottery youngerthan Late Byzantineand Early Umayyadperiodwhich suggestthat the aqueductwas out of use and filled with sedimentafter the 7th century. Evidenceof seismicruptureis most obviousin the left-lateral offset of the northwesterncorner of the birkeh and the aqueduct(Figs. 2 and 3). The fault tracestrikesN8°Eand offsets the birkeh wall2.2 +/- 0.5 m (Klinger eta/. 2000). A repair to the northwestcornerof the birkeh wall suggeststhat an earthquakedamagedthe birkeh shortly after it was constructed.The birkeh wall was built directly over the fault, and an additionalcourseof stonewas addedto the top of the wall over the original plasterfollowing subsidenceof the northwestcorner.

414 CROSSING JORDAN We collectedcharcoalsamplesfrom within the mortarat the top of the northwestwall wherewe identified this repair.We also sampledcharcoal f rom the mortar at the baseof the exteriorof the westwall in the southwestcornerwhereit wasexposedin a looter'spit (Fig. 3C). The radiocarbon age from the baseof the exterior wall yielded a date of AD 607-680,similar to the resultsof the radiocarbondatingof the repair,indicatingthat this structure experiencedat leastone earthquake in the 7th century,which resultedin somesubsidenceand repairof the birkeh. However,it is clear from the dramaticdisplacementof the northwest cornerand the aqueduct that a subsequent and more severeearthquakeruptured the birkeh, probably resulting in its abandonmentin the 7th

century.Evidencefor three other ground-rupturing earthquakes that occurred in 9th, 11th, and possibly 16th centurywas documentedin the strati graphiclayers abovethe collapsedand offset aqueduct(Hayneseta/. 2006).

N2o"w35"NE I

Smndlngpluter-' COIII!fed wall

Rotated

Wall Sectlons

Ruins of the Oasr at-Tilah

Reservoir (Birkeh)

Area E

0

lOrn Sale

c "csample in mortar

Lower

(614-693 A·0 ·') plaster

11

/

-~-

Upper plaster

Qasr at-Tilah Birkeh West Wall Top course

I

missing

"

Fracture Top course

1

~-- -

missing j

~i

,.__N1~

w

Slop!

......,_rir-1~:~

Step4

Slep5

"c sample

in mortar

(607-680 A.D.) 012345

10m

Figure3. A) Photograph toward theNorth showingthe ruprureof the westwall of thebirkeh and north-trendingaqueduct. The width of the aqueductchannelis 0.74 m wide for scale.B) Plan map of the birkeh and Area A excavationsquares.C) Elevation drawingof the west wall of the birkeh. Note the gapingcrack in the northwestcornerwherethe Wadi ' Arabah fault crossesthe structure. Locations of radiocarbonanalyseson charcoalfrom the wall mortarareshown.The laboratory radiocarbonage resultswere calibratedto calendaragesusing the CALIB Radiocarboncalibration program(Rev 5.0.1); copyright 1986-2005 by M. Stuiverand P.J.Reimer (seeonline: ).

TORN ASUNDER: EARTHQUAKES AT QASR AT-TILAH

415

Aqueducts in Wadi at-Tilah and Field Walls There are two separatesystemsof aqueductsat Qasr at-Tilah, each containingmultiple phases. First, thereis the systemthat carrieswater from the birkeh to the fort and the field system(Area A). This project hasexcavatedthe partsof Area A that lie directly acrossthe fault line, and thoseparts that connectto the settling pool. Other parts of this systemhave beensurveyed,but a full understandingof the multiple phasesof the system,andevensomeof its basicplan still awaitsexcavation. Second,there is the systemof aqueductsthat carrieswater from the spring in the wadi to the birkeh (Area B). This systemconsistsof at leastthreephasesof aqueduct,cut into the north wall of the wadi. Thesevary in height,with the lowestbeingapparentlyNabataeanand the highestleading directly into the birkeh. Incidentally, modernplasticirrigation pipesfollow the sameroute from the springto the modernreservoir.This project hasbegunmappingand recordingthis system,and plan to finish this work in subsequentseasons A systemof field walls (Area C) are found over an areaof 1000 m X 400 m west-northwestof the Qasr at-Tilah birkeh and fort (Fig. 1). A network of walls separatesthe areainto rectangularshapedplots. The north and westernmostfields are only delineatedby the N-S terracewalls andare not subdivided into smaller plots suggestingthey were abandonedbefore the completion of an expansionphase.The terracedfields representan ancientirrigation and agriculturalsystemthat was fed by water from the aqueductsin Wadi at-Tilah. Someof theseterracedfields are still currently being farmed. The main ancient,east-westaqueductthat trendsfrom the Qasrat-Tilah site intersectsthe field 'walls' at right anglessuggestingthe 'walls' function as a water delivery conduit. No channelis evident. Nor is there good evidencefor either plaster or mortar. The walls are generally 1.2 m in width and are constructedof large boulderschinked with cobbles.In placesthe walls standthree courseshigh. There is a meter or more of elevationchangeacrossthe fields with lower elevations steppingalwaysto the west or north along the gradientof the Qr2 fan surface.The wall systemhas beenerodedby the incision and lateral migration of Wadi at-Tilah to the southandWadi as-Sidirto the north (Fig. lA and lB).

Fort This project has so far only engagedin preliminary surveyof the ruined structureof the fort from which the site getsthe name'Qasrat-Tilah'. The fort is locatedabout50 m west of the birkeh and appearsas low moundsof rubble where walls once stood. Severalmoredistinct sectionsof the curtainwall allowed us to surveyand determineat leastthe size of the fort, andto measurethe constructiondimensionsof the walls. Basedon our surveyeddata pointson exteriorwall alignments,the fort appearsto measure37m2 (Fig. 2A). It hasfour projecting cornertowersthat eachmeasureabout5 m2• However,elucidationof the exactrelationshipof the towers to the walls awaits excavation,as it is not clearly visible on the surface.The southeast cornertower hasbeenerodedby lateralmigrationof Wadi at-Tilah. This providesa completestratigraphic sectionthrough the foundation of the fort in the bank of the wadi, and revealsthe only direct evidenceof earlier structuresthat underlie the site. Much of the pottery collectedfrom this section is Nabataean,suggestingthat the structurevisible in the wadi bank beneaththe fort is Nabataean.This conclusion,that therewas earlier occupationof the site beforethe extantfort was built, is supportedby severallines of evidenceincluding the early surfacecollectionsand the Nabataean-stylePhaseI aqueduct.Excavationto revealearlierstructuresbeneathboth the fort andbirkeh will be a major goal of future seasonsof this project. Investigationsof damageto thesestructures may be usedto help constrainthe magnitudeand intensity of earlier earthquakesalong this fault.

416 CROSSING JORDAN Conclusion As it now stands,our working hypothesisis that there is clear evidenceat this site of a Nabataean farmsteadupon which was built a Romanquadriburgium,adjacentto which was built a Late Byzantine/Umayyadbirkeh. The birkeh rupturedin an earthquake,after which the site was abandoned until recently. All of thesephases,of course,dependon the perennialspring in Wadi at-Tilah. Each of these major phasescontainsmultiple sub-phases,of modification and repair, which reveal the history of seismicactivity along the DeadSeaTransformFault. The location of this site across the fault line itself provides an absolutely unique opportunity to observe these past earthquakes.

Acknowledgments This work could not have been accomplishedwithout the supportof Dr. Ghazi Bisheh and Dr. FawwazAl-Khraysheh, past and current Director-Generalof the Departmentof Antiquities of Jordan(DAJ), who grantedpermissionto surveyandexcavateat Qasrat-Tilah, andDAJ representatives, Mana! Basyouniand Aktham Oweidi, for assistingin the field. The generosityof the Jordan Valley Authority for useof their resthousefacilities in as-Safiis gratefully acknowledged.Numerous peoplehelpedin various phasesof the field work for which I would like to especiallythank Dr. MohammedAtallah (YarmoukUniversity), Dr. JanBrashier(GrandValley StateUniversity), Nasser Mansoor,John Rucker, Abdel-RahmanAbueladas,Alivia Allison, JaniceMcCabe,RachelSmith, and Ma'moonNasser.Fundingfor the field researchwas providedby two grantsfrom the Committee of Researchand Exploration of the National GeographicSociety (6842-00 and 7406-03) to T.M. Niemi, and a GeologicalSociety of America GraduateResearchgrant to J.M. Haynes.

References Frank, G. (1934) Aus der' Araba I: Reiseberich.Zeitschrift des DeuchenPalastina-Vereins57: 191-280. Glueck, N. (1935) Explorationsin EasternPalestine,II. The Annual of the American Schoolsof Oriental Research15: 11-14. Haynes,].,T.M. Niemi, andM. Atallah (2006) Evidencefor Ground-rupturingEarthquakeson the Northern Wadi Araba Fault at the ArchaeologicalSite of Qasr Tilah, Dead SeaTransform Fault System,Jordan. journal of Seismology10(4): 415-30. Kennedy, D. (2000) The RomanArmy in jordan (London: Council for British Researchin the Levant/The British Academy). Klinger, Y., J.P. Avouac, L. Dorbath, and N. Abou Karaki (2000) SeismicBehaviourof the Dead SeaFault along Araba Valley, Jordan.Geophysicaljournal lnternational142: 769-82. MacDonald,B. (1992) The SouthernGhors and Northeast'Arabah ArchaeologicalSurvey(Sheffield Archaeological Monographs5; Sheffield: J.R. Collis). Musil, A. (1908) Arabia Petrcea,v. III (Vienna). Niemi, T.M., H. Zhang,M. Atallah, andJ.B.J.Harrison(2001) Late Pleistoceneand HoloceneSlip Rateof the Northern Wadi Araba Fault, DeadSeaTransform,Jordan.journal of Seismology5(3): 449-74. Parker, S.T. (2000) The Defense of Palestine and Transjordan from Diocletian to Heraclius. In The Archaeologyofjordan and Beyond:Essaysin Honor ofjamesA. Sauer,edited by L.E. Stager,J.A. Green, and Micahel D. Coogan (Studiesin the Archaeologyand History of the Levant, 1; Winona Lake, IN: Eisenbrauns):367-88. Zhang, H. (1998) Late Pleistoceneand HoloceneSlip Rate of the Northern Wadi Araba Fault, Dead Sea Transform,Jordan(MS thesis,University of Missouri-KansasCity).

THEOCRATIC EMPIRETHE BYZANTINE PERIOD

This page intentionally left blank

48

Bioarchaeology of North Jordan A Decade of Cooperative American and Jordanian Student Research Jerome C. Rose, Mahmoud ei-Najjar, and Dolores L. Burke

Introduction At the initiation of the King Fahd Centerfor Middle Eastand Islamic Studiesat the University of Arkansas,a group of faculty were sentto tour universitiesin SaudiArabia andJordanin 1994.The Faculty of Archaeologyand Anthropology of Yarmouk University in Irbid, Jordan,respondedby expressingan interest in joint projects. Negotiation ensued between Yarmouk's Faculty of ArchaeologyandAnthropologyandthe Departmentof Anthropologyat Arkansas,representingand funded by the King Fahd Center.Thus, the Yarmouk-ArkansasBioarchaeologyField School was born, with its first joint project begunat Sa'adin 1995. The overall researchstrategyfor the field school'swork is to reconstructhealth and diet and level of well-being of the Byzantine inhabitantsof rural north Jordan. Recent scholarshiphas demonstratedthat the productionand export of wine and other agricultural productsbroughta reasonablelevel of prosperityto the easternMediterranean(Kingsley and Decker2001) andwe are testingthe plausibility that this prosperitywas sharedby inhabitantsat the lowest socio-economic levels, having a positive impact on health and diet within thesesmall communities. It was decidedby the co-directorsthat the field school excavationsand subsequentanalysis would be student-researchoriented, not only in excavation and data collection but in final productionof written analysis.In the following field seasonsat Sa'adin 1995-97,at al-Yasila in 1998,and at Ya'munin 1999-2005,the studentscarriedprojectsthroughfrom excavationto final analysis.This chapteris a synthesisof studentresearchin the bioarchaeologyand archaeologyof thesethree small sites in north Jordan: from the Arkansashonorsthesesby Matthew Anderson, Melinda Bond, Brian Dickinson,Kelly Gilbride, Katie Johnson,andJasonKizzia; from theYarmouk MA thesesby MohammedAl-Asagheirrin, FatimaAl-Awad, Nibal Khalil, Ali Khwaileh, HishamAlKoufahi, Hamza Morouj, and Ammar Al-Obeidat; from the ArkansasMA thesesby Catherine Alston, JadeBarnes,Kory Cooper,Melinda King, Kimberly Williams; and from the dissertationsby Abdulla Al-Shorman at the University of Arkansasand by Abdul Nasser Hindawi at Freiburg University (in progress).

Figure 1. The combinedstaff and studentsof the 2005 Arkansas/Yarmoukfield schoolat Ya'mun

Figure 2. Field Schoolstudentsbeginningthe excavationof a row of Byzantinetombsat Ya'mun in 2005

BIOARCHAEOLOGY OF NORTH JORDAN

421

Student Research Results Field Methods and Data Collection We beganrecord-keepingwith an assemblageof locus and level forms along with artifact and pail tagsfrom Yarmouk and skeletalforms from Arkansas,but soondiscoveredtheseto be inadequate. We attemptedto solvethe problemsby examiningexcavationmanualsand forms from archaeological teamsin Jordan,including the work at Petradirectedby MarthaJoukowskyand early versionsof the MadabaPlainsProjectmanuals(culminatingin Herr and Christopherson1998). Becausebioarchaeologistsseethe skeletonas the focus of research,we neededto modify the forms to place the skeletonin context.For us, ratherthan a tomb havinga skeleton,it is the skeletonthat hasa tomb. Consequently,Gilbride setaboutmodifying the MadabaPlainsforms to reflect our bioarchaeological orientation, and simplifying many to be filled out by checkingboxesand filling in blanks to reducethe possibility of errors madeby novice studentsor thosewhose English was elementary. Gilbride's honorsthesiswas testedin the field and ultimately becamethe first Handbookto Data Collection Forms, Bioarchaeologyof North Jordan, 2000 (modified in 2002). As we movedour excavationsfrom the small site of Sa'adto the largerand morecomplexsite of al-Yasila, we struggledto makesenseof the increasingvariation in tomb construction.So we were delighted to obtain a copy of The Necropolis of Hesban: A Typology of Tombs, by Waterhouse (1998) and quickly adaptedthe expandedtomb typology provided in the chapterby Krug. However, we still could not classify all of our tombs.So,Johnsontook on a project for her seniorthesis of taking all of the architecturalfeaturesfrom the 234 tombs excavatedat Sa'ad,al-Yasila, and Ya'mun and modifying Krug's systemto fit. Ultimately, sheproducedthe hierarchicalarrangement we use today: tombs for single burials with two types and five subtypesand tombs for multiple burials with nine types and zero to four subtypes. Population Continuity We establisheda Bronze to Late Antiquity presenceat Ya'mun, but the questionarose:Were we dealingwith the samepopulationover time or had migration alteredthe geneticsof the inhabitants? Geneticdifferencecould influencedisease,and so we had to answerthat questionbeforecontinuing with the skeletalresearch.Alston's master'sthesiscompared31 nonmetricmorphologicalgenetic variations of the teeth from Ya'mun-aminimum number (MNI) of 241 individuals from two Bronze Age tombs and 79 MNI from three Byzantine tombs. She found no genetic difference betweengroups;a similar peopleoccupiedYa'mun for 2,500 years. Comparableresults,using the samegenetictraits, tied the Late Roman/Byzantinepopulations together:Al-Awad reporteddentalmorphologyfrom Sa'ad;Khwaileh showedthat the ai-Yasilaand Sa'adteethwere drawn from similar populations;Al-Asagheirrin worked on teethfrom Waqqas,a Byzantinesite in the JordanValley, and showedthat the peopleof Waqqas,al-Yasila, and Sa'adall belongedtogether.We concludedthat the Late Roman/Byzantineinhabitantsin thesesmall sites were geneticallyhomogeneousand could be descendantfrom the BronzeAge inhabitants. Diet and Climate Diet is a critical factor in overall health; therefore,variation in prosperityshould influence nutritional quality and ultimately health. Wear on teeth is directly related to the kind of food, food preparationtechnology,and the subject'sage.So, if one controlsfor age, then differencesin wear indicate differencesin food or food preparationtechnology.Al-Awad reportedthe dental wear from Sa'ad,while Khwaileh'sdataindicatethat wearat al-Yasilawassimilar to Sa'ad.Al-Asagheirrin reportedthat the teeth of Waqqashad slightly more wear than al-Yasila, and thus Sa'ad.Morouj

422 CROSSING JORDAN reportedthat the wear at Ya'mun is slightly higher than Waqqasand al-Yasila. None of the differencesbetweenthese groups were statistically significant and so there appearsto be only small differencesin the foods eatenor how they were preparedat all three sites. Dental decayvariesby the amountand refinementof sugarsand carbohydrates.Al-Asagheirrin reportedthat Waqqashad less decay than al-Yasila (Khwaileh) and more than Sa'ad(Al-Awad), while Morouj reportedthat Ya'mun had fewer cariesthan Waqqasand al-Yasila. Wear and caries suggestsome minor diet differencesamong the three sites and especiallybetweenthose people buried in group and individual tombs. Seekingto understandthe reasonsfor thesedifferences,King conductedstablecarbonand nitrogen analysisof 127 teethfrom Sa'ad,al-Yasila,andYa'mun. Inasmuchasthe relative proportionsof carbonand nitrogen isotopescan suggestconsumptionof particular plantsand the proportion of animal proteins,the comparisonof the carbonand nitrogenisotopesof 26 BronzeAge teeth from Ya'mun showing no significant difference from the 101 Byzantineteeth from all three sites indicated that there is no major changein either the plants or proportionsof animal protein eaten during thesetwo time periods.All of thesepeopleseemto be eatingprimarily wheatandsome,but not much, meat. Further,therewere no significant differencesbetweenthe meanisotopevaluesof the threesites,althoughthe peopleat al-Yasilaate slightly moreanimal protein.Therewas,though, considerablevariationin the proportionof meateatenwithin eachof the sites.The valuesfrom the individual tombsat all threesiteswere clusteredtightly togetherindicatingvirtually identical diets. All the variation occurredin peopleburied in large group tombs, reflecting differential accessto animal proteinsresulting from variable levels of prosperity. Thesethreesiteswere includedamongothersin Al-Shorman'sdissertationdemonstratingthat oxygenisotopevaluesfrom humanteethtrackedthe samevariation in rainfall as indicatedby other sources.Of importancehereis that he showedrainfall at Middle-to-LateBronzeAge Ya'mun to be high and similar to the Roman/Byzantinerates,while different from the periodsbefore,between, and after. Given the similarity in rainfall, it is not surprisingthat we did not find dietary differences betweenthe Bronze and Byzantineperiods.

Tombs and Socio-Economic Ranking The tombsof the threesiteswere chronicledby studentsat varioustimes and in variousgroupings, andconsistgenerallyof the following: 81 excavatedtombsat Sa'adQohnson);48 excavatedtombs and 215 robbedtombsmeasuredand recordedat al-Yasila (AndersonandJohnson);and 161 tombs excavatedat Ya'mun (Barnesand Johnson). One assumptionof mortuaryarchaeologyis that the amountof humaneffort and resources(i.e. total energy expenditure)expendedon tombs reflects the socio-economicstatusof those buried (Pearson2000). Using the criteria of size, numberof rooms,architecturalcomplexitysuchaselaborate arcosoliawith stepsand multiple stonegraves,and supplementalfeaturessuchas lamp niches and othercarveddecorations,the threesiteswere given the following socio-economicrankingfrom lowestto highestby Johnson:Sa'ad,Ya'mun,and al-Yasila. Both BarnesandJohnsonsuggestedthat quality of workmanshipshouldbe an addedfactor for estimatingthe tomb status.They both documentthat sometombshaveobviouslybeencarvedby highly skilled artisanswhile othersof comparable size andcomplexity were carvedby lessskilled persons.Thus,the proportionof tombscarved by skilled artisanscan also be usedto rank the sites: Sa'adhaving none,Ya'mun having some,and al-Yasila having many finely carvedtombs. Cooper'smetallurgicalanalysisof personalornaments for his Master'sthesisalso confirms this ranking. Similarly, ranking can be found within the sites. At Sa'adthere is no apparentdifference in socio-economicstatus.At Ya'mun,architecturalcomplexity,tomb size,and workmanshipshowdifferencesin socio-economicstatuswithin the Roman/Byzantineperiod (Barnes).The socio-economic

BIOARCHAEOLOGY OF NORTH JORDAN

423

stratificationof al-Yasila hasbeenwell establishedby the site'sfirst excavator(Al-Muheisenand ElNajjar 1994) and confirmedby Anderson'sanalysisof the 9 tombsexcavatedby Al-Muheisen,the 3 9excavatedby the field school,and 215 recordedrobbedtombs.This differential rankingof group tombs supportsthe contentionof differential prosperityindicatedby the stableisotopeanalysis. Barnesdescribestombsfrom the Early BronzeAge throughthe Byzantineat Ya'mun,but identified no Iron Age tombs,suggestingthat there is a gap in occupation.Subsequentwork on the Tell by Hindawi for his dissertationfound numerousIron Age ceramics.Having establishedan Iron Age presence,we must searchfor the tombs. This has promptedus to begin a re-examinationof our architecturaland artifactual datato searchfor Iron Age tombsthat might havebeenmodified and re-usedby later peoples. Sources and Perceptions of Prosperity The presenceof large and elaboratewine pressesat Sa'ad(Roseand Burke 2004) and al-Yasila (AlMuheisenand El-Najjar 1994) establishthat surpluswine was beingproducedandcould havebeen sold to nearby cities, constitutingone of severalsourcesof local wealth. In order to verify that Ya'mun was in a similar situation, Cox undertookthe excavationand analysisof the wine presses for his honorsthesis. The four pressesrange in size and numberof pressingfloors-two with 7 floors, one with 4 floors, and one with 1 floor. The total areafor pressingis 81.2m2 with a minimum collection capacity of 2,745 liters. The size and capacity of the Ya'mun wine pressesare comparableto published pressesin Jordan, Palestine,and Israel, and Cox concludedthat they rankedat the 5Oth percentileor greater.Using a variety of methods,suchas total land for vineyards and populationestimates,Cox concludesthat Ya'mun was like the other two sitesand could have producedfar more than was needed,providing a local sourceof wealth. Cooperdemonstratedthat many of the copperitems of personaladornmentwere alloyed specifically to makethem appearlike silver and gold, thusgiving the impressionof wealth.This behavior suggeststhat the inhabitantsof thesesitesaspiredto more than the moderatewealth that they possessed.Kizzia's analysisof Ya'mun personalitems for his honorsthesisalso indicatesthe presence of moderatewealth. Further, he confirms the existenceof socio-economicdistinctions at Ya'mun wherelarger, more complex,and finely carvedtombsproducedmore personalornaments of greatervalue. The People The paleopathologicalanalysisof the bonesfrom thesethree sites was complicatedby extensive tomb robberies(approximately90%), both ancientand modern. Frequently,the robbersleft the bones,but jumbled,while in othercasesthe boneshad beentossedoutsidethe tomb and hencelost. Despitethesechallenges,the human remainshave much to tell us. Al-Koufahi focusedhis MA thesison the large group tombsnearthe churchat Sa'ad,while Williams analyzedall the othertombsand integratedthe two studies.The Late Romantombshad only 28 percentsubadultsand a significantunder-representation of infants (Al-Koufahi), while subadult proportionsin the Byzantine tombs rangedas high as 45 percent(Williams). The major skeletal pathology at Sa'adfor both periods is arthritis of the joints, while frequent damageto the neck vertebraeimplies that heavy loadswere being carried on the head.Al-Koufahi and Williams both showedthat arthritis of the shouldersand elbows ranged between29 to 33 percent,indicating heavy work. The healedfracture rate of 16.5 percentalso indicatesheavy and dangerouswork. Infections were few, at less than 1 percent. Khalil's MA thesison the skeletalremainsfrom 14 excavatedtombsat al-Yasila (MNI 153) again showedthat the Byzantinetombs producedmore subadultskeletonsthan earlier periods. Khalil

424

CROSSING JORDAN

statesthat thoseliving at al-Yasila were similar to thoseat Sa'ad,with few infectionsand common arthritis of the joints. The arthritic neckvertebraeagainindicatethe carrying of heavyloadson the head,while arthritis of the major joints is 47 percent.Healedfractures(7o/o) are lesscommonthan at Sa'ad,while infections remainedlow at 5 percent(Khalil). Barnesand Al-Obeidat wrote their theseswhen only 161 of the eventual208tombsat Ya'mun had been excavated,a 77 percentsample.Arthritis of the major joints and osteophytosisof the spineare 43 percentand40 percent,respectively.Theseare the mostcommonlesionsand suggesta workload similar to al-Yasila, but slightly lower than Sa'ad.Infectionsare rare, at 3 percent,as are healed fractures at 3.5 percent.Although there is not much difference betweenthe three sites, Ya'mun and al-Yasila are most similar with lower fracture ratesand osteoarthritisthan Sa'ad.An important note is that Barnesreports twelve casesof cribra orbitalia (pitting of the eye orbits), which indicatesthe presenceof iron deficiencyanemia,probablyresultingfrom a low animal-meat, high-grain diet. In summary,the demographyof all three sites showedthat infants and young children were more frequentin the Byzantineperiod, with ongoinganalysisof teeth pointing to more infants in the Byzantinegrouptombs.Sinceinfant mortality alwaysrangesbetween20 percentand40 percent increasesin proportionsof subadultswould be more a reflection of burial practicesthan mortality, and so it seemsthat Christianswere more likely to bury the infants in the sametombsas the adults. The most frequentpathologicallesion is spinal arthritis at all three sites.This suggestionof heavy workloadsis supportedby high ratesof major joint arthritis and healedfractures.Infection ratesare low andare not meaningfullydifferent betweenthe sites.We concludethat dietswereadequateand bacterialinfectionsinfrequent.

The Church In addition to tomb excavation,Arkansasfield schoolstudentsat times joined the YarmoukArchaeology Departmentfield schoolto conducttheir own researchprojectson the church.The church, situatedon the southeasterncornerof the Tell, was discoveredby the 1999 bioarchaeologyfield schooland subsequentlyexcavatedby Departmentof Antiquities archaeologists.The unexplored narthex and associatedecclesiasticalbuildings to the south of the church were excavatedby the YarmoukArchaeologyDepartmentfield school.Dickinson, for his honor'sthesis,excavatedin the unexplorednarthexwherehe found that at somepoint it had beendivided into threeroomsduring a major architecturalmodification. Underneatha baptismalfont (one of two found) and a pile of 44,900small Byzantinetesserae(matchingthe churchfloor) he found a floor madeof large white tesserae.Betweenit and the underlying original mosaicfloor (matchingthe nave) was a fill layer containingUmayyadsherdsand a fragmented6th/7th-centuryJarashlamp. This datesthe church modification and reuseat leastto the Umayyadperiod. In 2004 the churchwas thoughtto be of basilicastyle, with two aisles,a narthex,and an external apse;it was then coveredwith cleansandand plastic. As Bond beganher honorsthesison the church,shediscoveredthat the mosaicphoto recordwas not complete.Sheobtaineda grantto pay for uncoveringthe floor and photographingit in rows of 1 m squares.While clearingthe edgesof the mosaicadjacentto the apsefor photography,shefound them to continueeastward.Bond ultimately exposeda Diakonikon and Prothesis,demonstratingthat the churchwas rectangularwith an enclosedapse.In addition to detaileddocumentationof the floor and constructionphases,Bond solicitedwidespreadexpertopinionsto providea dateof 499-500CE for the dedicationinscription. It is clear that churchsize and mosaicquality also rank the threesitesin the sameorder of prosperity as the other measures.

BIOARCHAEOLOGY OF NORTH JORDAN

425

Figure 3. Detail of mosaicchurch floor at Ya'mun showingfirst-phasefigure removal by randomizedtesseraeon the right andsecond-phase figure removalwith large tesserae abovethat correspondsto Umayyadremodeling

Conclusions Large and complex wine pressesare just one sourceof wealth, while analysisof the quality and complexityof tombsandchurches,along with itemsof personaladornment,demonstratea prosperity ranking in the order of Sa'ad,Ya'mun, and al-Yasila. Further, tombs from the latter two sites exhibit variation in quality suggestingdifferential accessto wealth. Genetic analysis provides evidencefor geneticcontinuity over time and acrossthe three sitesduring the Byzantineperiod. Dental and isotopic researchsuggeststhe samecontinuity in diet over time and spacewith some variation in accessto proteinthat follows the prosperityranking. Workloadsappearto be arduous, but overall health is good. Thereis a trend for more infants to be buried with adults during the Christian period.

References Herr, L., and G. Christopherson(1998) ExcavationManual: Madaba Plains Pro;ect (Berrien Springs,MI: Institute of Archaeology,AndrewsUniversity Press). Kingsley, S., and M. Decker (eds.) (2001) Economyand Exchangein the East Mediterraneanduring Late Antiquity (Oxford: Oxbow Books).

426

CROSSING JORDAN

Al-Muheisen,Z., and M. EI-Najjar (1994) An AnthropologicalStudy of the HumanRemainsfrom Yasileh: A ClassicalSite in NorthernJordan.Mut'tah Journal for Researchand Studies9(2): 5-26. Pearson,M. (2000) The Archaeologyof Death and Burial (CollegeStation: TexasA&M University Press). Rose,]., and D. Burke (eds.) (2004) Sa'ad: A Late Roman/ByzantineSite in North Jordan (Irbid, Jordan: Yarmouk University Press). Waterhouse,S. (1998) The Necropolis of Hesban: A Typology of Tombs (Berrien Springs, Ml: Institute of Archaeologyand Andrews University Press).

49

Beyond the Rock Petra in the Sixth Century

CE

in the Light of

the Papyri Robert C. Caldwell and T raianos Gages

Over a dozen yearshave passedsince the discovery of more than 150 rolls of papyrusin a room attachedto a Byzantinechurch in Petrain Decemberof 1993. Thesetexts were poorly preserved becausethey were carbonizedin a fire that severelydamagedthe churchand adjoiningrooms.Since the discoveryscholarshaveconcentratedon the difficult tasksof conservingand reconstructingthe papyri. Despitetheir poor condition, many rolls have yielded enoughcontinuoustext to warrant publication,while somefragmentarypieceshaveprovideduseful snippetsof information. Recently the publicationproperof the papyri began:onevolume of documentshasalreadyappeared(P.Petra I), two more are nearly finished, and a fourth volume is plannedfor the next couple of years. Though the majority of the Petra papyri remain unpublished,enoughevidencehas emerged from the preliminarystudy of the texts to allow a generalassessment of their historical significance. Most importantly, the papyri attestthat for most of the 6th century,Petraand its environsdid not experiencethe economicandsocial declinethat had beenpostulatedprior to this discovery.Instead, the documentswitnesswidespreadcultivation of grainsand the vine, activitiesthat seeminglywere not impairedby the earthquakeof 551 (the earliestfirmly datedpapyruscomesfrom 537, the latest 593). This level of agriculturecould supportpoliteuomenoi(Lat. curiales), land-owningeliteswho servedas officers for the cities of the Byzantine state, and collected and paid taxes in a manner similar to thosein other provinces.The continuedprosperityof Petraalso meantthat it remained tied administrativelyto the Byzantinestateratherthan beingleft to the careof Ghassanidphylarchs. Along with the Byzantine administrationthe military too continued to be present. (For recent generaloverviewsand assessments of the Petrapapyri, seeP.Petra I: pp. 1-18 and Koenen,Daniel, and Gagos2003). Archaeologicalwork in the Petraareahasboth supportedtheseconclusionsandgonesomeway toward underminingthem. On the one hand,the large monasteryand pilgrim centeron the Mountain of Aaron a few kilometerssouthwestof Petraflourishedduring the time coveredby the papyri. Similarly, the church where the rolls were found underwentremodelingat the start of the 6th century. Amongthe improvementswere a colonnadedatrium and marble furnishings and wall mosaicsin the church itself. However, in contrastto this positive evidence,the inner wall of the city, which datesto the later Byzantineperiod,hadshrunkwhile residentialareaswithin its compass lay ruined and abandoned(Fiema et al. 2001: 53-77; Frosen and Fiema 2002: 69-70). The

428

CROSSING JORDAN

combinedevidencefrom archaeologyandthe papyri further suggeststhat Petrasharedin the trends of urbanismthat havebeenobservedin other cities of the sameperiod while it was able to support somemonumentalconstructionand a land-owningelite. All in all, the papyri supportthe general picture that Petrawas an 'average'provincial city during the 6th century. Theseinitial assessments demonstratethat the history of Byzantine Petra is being (re)written becauseof the discoveryof the carbonizedpapyri. As publicationsof the texts proceedand knowledge of the contentsof the documentsgrows, our understandingof the papyri will be inevitably refined and new inquiries will be undertaken.In this article, we would like to refrain from giving yet one more overview of the papyri, a task that has been undertakenseveraltimes in the past. Instead,we would like to proposea couple of new researchdirections.The first is an attemptto explore the natureof the city administrationin tandemwith the role of the local elites and their self-perceptionwithin the fabric of the local and broaderByzantineculture. The seconddirection we suggestexpandson the administrativeties Petrahadwith the wider Byzantineworld to consider the cultural onesas well.

Administration,

Elites, and Self-Perception

As a preliminary to thesenew directions,we would like to run through someof the evidencethat supportsour assertionsthat 6th-centuryPetrawas not an isolatedplace, that there was an active group of curiales who administeredthe area,and that Petramay not haveflourished to the degree suggestedby the papyri. AssumptionsaboutPetra'sisolation may be setasideby observingthe rapid receptionof imperial legislation, much faster in fact than Egypt or other part of the empire. Laws enactedby the emperorin Constantinopleappearto have been swiftly adaptedby the Petraean administration.Justinian'sNovella 47 issuedon August 31,537requiredall documentsto include his regnalyearalongwith consularand indiction yearsin the datingformula (which appearsright at the beginningof the Petrapapyri). A sworn settlementdatedas early as May 10, 538 (P.Petra I 2) conformsto this legislation. Two requeststo transfertax responsibility (P.Petra I 3 and 4), both from August538, also adhereto this legislation.Later, in a receiptof 578 (P.Petra I 10), a tax payer took advantageof a remissionof taxesin arrearsand a one-quarterremissionof taxesper year for four consecutiveindiction yearsstartingin 575; the receipt refers specifically to the reductionof one quartergrantedby 'our most pious and divine lords', the emperorsJustin II and Tiberius II (Novella 163 of 575). The numerousrequeststo changetax responsibilityandthe taxreceiptsfound amongthe Petra papyri reflect the detailed recordskept by the tax collectors. Indeed,property was registeredto taxpayers'personalaccountsthat were kept in public records.In anothertransferof tax responsibility (inv. 60.1 = P.Petra III 19) from 539-540,the propertyin questionis mentionedas 'registered in the "public codex"' of the city of Augustopolis,mostlikely the largesite of Udhruh about10 km eastof Petra.The recipientsof this request,however,are not just officials from Augustopolisbut also the 'currentcollectorsof gold from the metropolis[i.e. Petra]and of the areasfrom the "total holdings"' of AugustopolismanagedthroughPetra'.The term translatedas 'total holdings',homas in Greek, appearsextensivelyin the Petra papyri. Overall, this term appearsin two contexts: it representsthe total liability both of an entire city, as in P.Petra III 19, and of an individual. As we haveseen,landedpropertywas registeredin public records(demosioichartai, demosios codex).Eachperson'spropertywasassessed in units of tax liability basedon the type of land and its quality. The sum of this liability equaleda tax payer'shomas.In turn, the total of the liability of all tax payersamountedto the homasof the city. The unit of liability for accountingpurposesin the areaof Petrawas the iugum, as the taxreceiptsP.Petra I 7-10 makeclear.The iugum measuredthe shareof a tax assessment that a city and its registeredpayerswere liable to pay. For example,if a

BEYOND THE ROCK

429

city was assessedto produce 100 nomismata('golden coins') in tax and there were 100 payers registeredat a iugum each,then eachpayerwould be liable for one nomismaof tax. The Petrapapyri demonstratethat taxeswere organizedat the city level, as elsewherein the empire. Tax collectors were keeping and using city-based documentsto keep track of tax obligations.Thesecollectorswere kata kairon ('temporary',P.Petra I 3, 4, 7, etc.), and servedfor a specificterm. This designationsuggeststhat a rotatinggroup of citizenswere servingas hypodektai ('tax collector') or chrysypodektai(lit. 'tax collector of gold') in chargeof record keepingand tax collection. In at least one casethe tax collector is also called politeuomenos(P.Petra I 3 and 4), further emphasizingthe connectionbetweenfiscal administrationand the city. Politeuomenoi appearin a numberof other papyri, often in contextswhere it is clear that the bearerof the title is not performingan office. Accordingly, politeuomenosappearsto be more of a statusdesignation, indicating that an ancestorwas a curialis, ratherthan an office itself. Nevertheless,the survival of this designationreinforces that the city enduredas the center not only for tax collection and administrationbut also as the contextin which statuswas expressed. Not all the municipal officials chargedwith the keeping of records were called hypo- or chrysypodektai.In P.Petra III 23 (inv. 67), anothernotice of tax transfer,the addresseeis a single person, Alpheios son of Valens, 'the most honorable keeper of public records' (Gr. demosios chartophylax).Alpheios' honorific, eudokimotatos('most honorable'),differs from the hypodektai, who are alwaysaidesimotatoi('mostrespectable'),in additionto havinga different title. Despitethe different honorific and title, however,Alpheios musthaveworkedwith the taxcollectorson public, city-basedwork, and he was probablythe main official in chargeof maintainingthe public records. The receiptsmentionedabove demonstratethe complexity of the recordsthat Alpheios and otherswere chargedwith keeping.Someof them (e.g. P.Petra III 19) provide evidencethat Petra and Augustopolis were sharing responsibilitiesfor tax assessmentand collection. The receipts mentionin extremedetail the numberand fractionsof iuga the payerwas liable for in land classified asexaktorikafor PetraandAugustopolis(following Fournet2003: 403-4).Exaktorikarefersto one of the classificationsof land that the collectorshad to keeptrack of. Its precisenatureis not clear, but it is known that in the Petraregion otherland was classifiedas 'free' or 'patrimonial'.Different classificationswere probably subject to different kinds of taxes and perhapsdifferent rates of taxation or evendiffering assessments of their value in iuga. The details of how the city-collectors kept track of this land remain obscure,but we do know that since transfer noticesaddressedto them statedthesedifferent classifications,one of their dutieswas to know the various land categories. The collectorsmay evenhavebeencollectingtaxeson all classesof land regardlessof classification, as we have no evidenceat all for officials operatingoutsideof the city-basedadministration. That is to say, the samepool of wealthy families provided tax collectorsfor 'free' land collected throughthe cities and also for imperial estates,the patrimonialland. It was up to the public record keepersto track the different classificationsof the payer'sland. A man called Patrophilosson of Bassos,the father-in-law of Theodorosson of Obodianos,the centralfigure in the papyri (seethe family tree in P.Petra I 9), was payingback taxesover a number of years. The receiptsissuedto him cover all types of taxes, including extra taxesand a levy for oinokreon(a mixture of wine and meat).This tax may provide evidencefor a link betweenthe citybasedadministrationandthe provincial military administration,sincethis term, despiteuncertainty surroundingits precisemeaning,is associatedwith the annonamilitaris (annualmilitary tax). We know from other papyri that the Byzantine army maintaineda presencein the area, a fact not previouslyacceptedby all modernhistorians.In particular,the village of KastronZadakathon(modern Sadaqa,a few kilometerssouth of Petra)seemedto have remainedthe baseof a military unit called kathosiomenoidomesticikastrou Zadakathon(Imperial Honor Guardsof Kastron Zadakathon; inv. 44a). Another papyrusmentionsa prefectof anotherkastron called Ammatha.These

430

CROSSING JORDAN

soldiersmay well have benefitedfrom collection of the oinokreon.Additionally, it is conceivable that the Ghassanids,the tribal allies of the Byzantines,may have also receiveda portion of it. We know that Abu Karib, a Ghassanidphylarchappointedby Justinian,was involved in trying to end a disputein Sadaqa(inv. 83). It is perhapsno coincidencethat he appearsin a village with a military presenceand,althoughthe partiesin the disputeare not military but rathermenof cloth, one of the two mediatorsis in fact a military man, a prior. As Walter Kaegi has argued,Abu Karib and other phylarchsin return for their alliancewith Justinianwould havereceivedthe annonaas part of their wages(Kaegi 1985). Thus,the evidencefor tax collection links PetraandAugustopolisto the army andthe phylarchs, which were both concernsof the provincial and central administration.Taxation also links these two cities to the central administrationthrough the processof tax assessment. Inv. 8 recordsthe exchangeof four piecesof land, two for each party, that are classedas 'free' and 'patrimonial'. Relatedto the taxassessment of the land is a mentionof a FlaviusValerios, an epoptes('inspector'). Although the referenceis damaged,Valerios apparentlyperformedan epopsia('inspection')for the area.This usually happenedat the requestof cities or evenprovincesor diocesesto challengea tax assessment imposedon them. Perhapsmost well-known is the epopsiathat reducedan overly high assessment for the city of Cyrrhus in Syria but that was apparentlynever confirmed, as we learn from a number of letters by the city's bishop, Theodoret,to various people in Constantinople, including the empress,in which he pleadsthat the taxreductionthat resultedfrom the epopsiabe confirmed (Azema 1955: Ep. 42). In the caseof Cyrrhus,the official who led the epopsiawas not local but dispatchedto overseethe review of the tax assessment. We cannotknow whetherValerios was a local personchargedwith overseeingthe reassessment of the land or whether he was dispatchedby imperial officials. Thereare at leasttwo more referencesto epopsiaand to epoptoi in other Petrapapyri, and they may refer to the samereassessment of tax liability. At the least, they demonstratethat tax payersin the areawere concernedabouttheir assessments and were eagerto take advantageof meansto reducethem. We haveconcentratedon tax assessment and tax collectionbecausethroughit we haveseenhow importantcities remainedas units of administration.The addressof P.PetraIII 19, discussedabove, mentionsthat part of Augustopolis'holdingsweremanagedthroughPetra,andthe receiptsmention tax liability for land classifiedas exaktorikain both Petraand Augustopolis.Thus, for somepurposes,the administrationsof the two cities appearto havebeenmerged.But the mergeris between two cities, not betweencity and village. Nowherein the papyri is thereevidencethat tax collection functioned at the village level. Were this the case,it would suggestthat the city no longer was considereda viable unit of administration. Even though many of the officials mentionedin the papyri might not originate from the cities, they were neverthelessproviding the elite, who collected the taxes, and the officials for their churches.Theodoros,son of Obodianos,the central figure in the papyri, had special links with Sadaqa,where his father owned a houseand other property (e.g. inv. 83 and P.Petra III 23). It is possiblethat at somestagethe family movedto PetrawhereTheodorosservedas deaconand then archdeaconin the churchwherethe papyri were discovered.The family of the threebrotherswho divide the property recordedin inv. 10 owned housesin a village called Serila and in Petra.Vineyards and fields statedin the samedivision that have tenantsare also mentioned.But none of the housesare declaredas having tenants,so perhapsmembersof the family maintaineda dual residencein the village and in the metropolis.Indeed,the impressionone gets from thesepapyri is of an educatedgroup of property holdersthat may originatefrom villages like Serila and Sadaqabut provide the pool of city administrators.It is worth noting that in the surviving signaturesof the Petradocuments,with only two exceptions(a priestwho can barelywrite his nameandan illiterate woman from Sadaqa)everyonecan sign his own name.

BEYOND THE ROCK

431

Overall, the evidenceof the Petrapapyri eithersuggeststhat families of the elite suchas thoseof Theodoros,Patrophilos,and othersreally originatedfrom Petraor elsethat althoughfamily property may have beenheld in the city and village, it was more prestigiousfor them to statethat they camefrom Petraand not a village like Sadaqawheresoldiersof local origin were stationed.These families were the economicand social successes and they preferredto attachthemselvesto the place wherethey could exercisebettertheir position and enjoy their status:the city. The papyri also suggesta positiveassessment of Petraand Augustopolisasenduringlocal administrativecenters.Nevertheless,the evidencethat at leastsomemembersof the landowningelite might not originatefrom Petrabut rather from surroundingvillages recallsthe archaeologicalevidencementionedabove. Sucha contrastcompelsus to ask why Petra,as a centerof administration,was apparentlystill important even in the face of the city's obviously embarrassingcondition. Should perhapsthe mergerwith Augustopolisbe takenasa concessionto Petra'sreducedcircumstances? Tradition and habit may form one part of an answerto the questionjust posed.But there must be more to it than that and there might be a seriesof factors that are interlinked. Wealthy landownersare powerful, whetherin the contextof city or village. It is possiblethat by maintainingtheir ties to the central administrationthroughcity administrationthe elitesof this areawereableto distinguishthemselves from tribal leadersby more closely allying themselveswith the Byzantine state. But some texts, especiallythe receiptsthat recordpaymentof taxesin arrears(and thereare manyof them),may be signsof economicanxiety that forced the local elites to keep a closeeye to the imperial legislation from Constantinople,especiallywhen it provided tax relief for the local tax payers. As is clear, thesedocumentspresenttheir own problemsas they appearto reflect the practices, everydaylife, and anxietiesof the privileged local population. In many respects,thesepapyri, in addition to providing raw historical data for society at large, first and foremost witnessthe selfperceptionof those people that actually producedthem. It is clear that the Petraeanscribesand elites styled themselvesvery much as their counterpartselsewherein the empire,regardlessof the actualphysicalappearance and economyof Petra.They are all Flavii, they haveelaboratetitles, and they belong to similar administrativestructuresas other cities. This self-perceptioncannot be demonstratedbetterthan through the titulature of the city, which for the first time appearsin full form in someof our documents.Here we encounteran almost compulsivesensitivity to local history throughthe titles of the city arrangedin strict chronologicalorderas they werebestowedupon it in successionby the variousemperors:'Imperial colony, Antoniniana,distinguishedand noble [or local, dependingon interpretation]motherof colonies,HadrianicPetra,Metropolis of Third Palestine Salutaris'(seeP.Petra III 23.2-3 with note). As it hasbeenpointedout elsewhere,this seriesof epithets,on the one handsummarizes500 yearsof the city's history, but on the other,epitomizesits Romanizationand placesPetra in the political geographyof the Roman and Byzantine Empires. Thesetitles are signifiers of connectedpastsand commonvaluesfor a group and a communityand act as mnemonicdevicesto recall the sharedhistory of Petra. By this period, Petramay have lost its original grandeurand importanceas the archaeological evidencediscussedabovesuggests.The educatedlocal elites and bureaucratsrecountedPetra'spast as expressedin their institutions, titles, and the titulature of the city, but at least in their minds, Petraremainedpart of that glorious past if we understandthat thesetitles were not passedover without a secondthoughtby the writers and readersof the documentsthat usedthem.

Beyond the Rock Let us now turn to the secondsuggestionfor further researchthat we have mentioned,more specificallythe cultural ties of Petrato the broaderByzantineworld. That Petrano longerseemsto have been isolated from the army and administrationprompts us to ask whether the elite who

432

CROSSING JORDAN

servedin the administrationalso sharedin the broadercultural life of the ByzantineEmpire in the 6th century.And indeedthis seemsto havebeenthe case.As it hasbeenpointedout, the beautiful floor mosaicsin the church where the papyri were found bear some resemblanceto the mosaic schoolsof Gazaand Madaba(Fiemaet al. 2001: 219-70,esp.at 262-65).The handsof the scribes who producedthe documentscomparefavorably with contemporarypapyri from Egypt andtestify to a high level of skill and familiarity with the productionof suchtexts,while their contentattestsa level of legal competencethat Egyptiantexts do not alwaysmatch.Thus,artisanscapableof executing fine mosaic floors were active in Petra along with people capableof writing and producing preciselegal documents.Another way to explore Petra'scultural connectionswith neighboring provincesis to attemptto establishlinks betweenTheodoros'family and figures attestedoutsideof the papyri. One such personis a certain Gessios,whose father'snameis not known, a famous physician, teacherof medicine,and philosopherduring the reign of the emperorZeno (474-491).Damaskios, a main sourcefor Gessios'life, recordsthat he came from Petra, while Stephanosof Byzantium locateshis origin in el-Ji, an agricultural areaadjacentto Petranow known as Wadi Musa. (The fragmentof Damaskiosthat mentionsGessiosis found in the Suda; for text and commentary,see Athanassiadi1999, fr. 128; for Stephanos,see Meineke 1849: 200; for further citations, see Martindale 1980: s.v. Gessius3.) Leaving Petra and el-Ji, Gessiosis reportedto have taught in Alexandria.Time spentin Gazais implied by his friendshipwith Aeneasand Prokopios,two sophists from Gaza. Previousscholarshave understandablyremarkedlittle on Gessios'origins (e.g. Seitz 1892: 13). But the two sourcessuggestsomethingsimilar for Gessiosthat we have observedabove for membersofTheodoros'family: origin in an agriculturalvillage nearPetrabut alsoidentification with the metropolisitself. Given Gessios'origin, his connectionto Gaza,and his renownas an iatrosophistes('professorof medicine'),it is temptingto identify him with the deceasedphysicianwho appearsin P.Petra I 2, lines 150-53.In fact, this papyruswas drawn up in Gazaand datedMay 10,538.Its very fragmentary contentsare difficult to establish.The editorshavecalled it 'an agreementconcerninginherited property'. This is the only Petrapapyrusso far as we can tell to have beenwritten outsideof the Petraarea,and the provenancefrom Gazawould accordwith Gessios'known connectionsto that city. The date of Gessios'deathis not known from other sources,but if his acmewas during the reign of Zeno, it is plausiblethat he was deceasedby 538. The nameof the doctor in the Petrapapyrusunfortunatelydoesnot survive in full, but it begins Ge[. The gammaand the epsilonare printedwith dots to indicatethat the lettersare not complete. An examinationof the original papyrus,however,confirmedthat the readingis secure.After Ge[,ssiou would fill the line basedon the length of the following one. This doctor Ge[ is describedas being logias mnemes('of learnedmemory'),a descriptionusedfor learnedmen that is apt for the famousiatrosophistes.Moreover,this descriptionis rare in papyri and suggestssomeonewho was particularlyesteemed.In the papyri of this period deceasedpeoplearecommonlydescribedas 'most blessed'(makariotatoi). If this Ge[ were the same person as Gessiosthe iatrosophistes,then he would be the brotherof Dousariosson of Valens,a maternaluncle of Theodorosson of Obodianos. At this point the connectionbetweenTheodoros'family and Gessiosand the wider intellectual world that he took part in remainsno more than a plausibleinference.The nameGessiosappearsin a numberof unpublishedPetrapapyri, though we should keep in mind Stephanosof Byzantium's descriptionthat the name was common (Meineke 1849: 200)! We hope that further study will determinewhetherthe learneddoctor of P.Petra I 2 and Gessiosare the same.If this point can be established,then it is possiblethat membersof Theodoros'family could be linked to the vibrant intellectual centers of Alexandria and Gaza that are vividly attestedin the works of Aeneas, Procopios,ZachariasRhetor, and others.

BEYOND THE ROCK

433

We havedescribedthesetwo recentareasof researchon the Petrapapyri only as two examples of how the history of ByzantinePetracontinuesto be (re)written and how much work remainsto be done in terms of editing the texts and the historical interpretationof the documents.Our task of writing the 'bigger picture' history of Petrahasjust begun.

Abbreviations P.Petra I = J. Frosen,A. Arjava, and M. Lehtinen (eds.), The Petra Papyri I (Amman: American Center of Oriental Research,2002) P.Petra III =A. Arjava, M. Buchholz,and T. Gagos(eds.),The Petra Papyri III (Amman: American Centerof Oriental Research,in press)

References Athanassiadi,P. (1999) Damascius:The PhilosophicalHistory (Athens: ApameaCultural Association). Azema,Y. (1955) Theodoretde Cyr: Correspondance,I (Paris: Les Editions du Cerf). Fiema, Z.T., C. Kanellopoulos,T. Waliszewski,and R. Schick (2001) The Petra Church (Amman: American Centerof Oriental Research). Fournet,J.-L. (2003) Review of The Petra Papyri I in Antiquite tardive, 11 (Turnhout: Brephols). Frosen,J., and Z.T. Fiema(eds.) (2002) Petra: A City Forgottenand Rediscovered(Helsinki: Amos Anderson Art Museum). Kaegi, W. (1985) The annonamilitaris in the Early SeventhCentury. Byzantina 13: 591-96. Koenen,L., R.W. Daniel, and T. Gagos(2003) Petrain the Sixth Century: The Evidenceof the Carbonized Papyri. In Petra Rediscovered:Lost City of the Nabataeans,edited by G. Markoe (New York: Harry N. Abrams in Associationwith the Cincinnati Art Museum): 250-61 Martindale,J.R. (1980) The Prosopographyofthe Later RomanEmpire, II (Cambridge:CambridgeUniversity Press). Meineke,A. (ed.) (1849) StephanosByzantii Ethnicorumquaesupersunt(Berlin: G. Reimer). Seitz, K. (1892) Die Schulevon Gaza: eine LitterargeschichtlicheUntersuchung(Heidelberg: C. Winter).

This page intentionally left blank

50

Petra's Churches The Byzantines and Beyond Megan A. Perry and Patricia Maynor Bikai

The archaeologicalsite of Petra remains one of Jordan'smost recognizedtourist destinations; however,surprisingly little is known aboutthe city's ancientinhabitants,particularly thoseof the Byzantine period. Three American Center of Oriental Research-sponsored projects1 sought to rectify this gap by exploring Byzantine-periodecclesiasticalcomplexeson a large hillside to the north of the Petra'smain colonnadedstreet(the 'North Ridge') from 1992-2002.Situatedon the upperhalf of the North Ridge'ssouthernslope, the PetraChurch,the Blue Chapel,and the Ridge Church, with their associatedartifacts, attest to the presenceof a dynamic post-4th-century populationat the site and the Church'sprominentrole in southernJordan. Constructionof the three churcheswithin a relatively short period of time during the late 4th centuryand mid-5th centurymarkeda functional shift in useof the city's North Ridge. During the Nabataeanand Early Romanperiods,domesticstructurescoveredthe lower sectionof the slope from the sector of the Petra Church up to area of the Blue Chapel. The military may have had exclusiveuseof the upperslopeand the top of the ridge,2 chosenfor its 360°view of the city center as well as a view up the back entranceto Petra.The upper sectorof the North Ridge additionally was usedas a cemeteryduring the Nabataeanperiod, for rock-cut tombs honeycombthe hillside and top of the ridge. Two of thesetombsuncoverednearthe Ridge Churchwereexcavatedin 1998 (Bikai and Perry 2001). In the mid-to-late4th century,the Ridge Churchwas constructedin the 'military' areaat the top of the Ridge, possiblyby membersof the military themselves.The Christianpopulationof Petrahad grown substantiallyby the middle of the 5th century, necessitatingthe constructionof a major ecclesiasticalcomplex containingthe Cathedralof Petra,an elaboratebaptismalcomplex,and the presumedresidenceof the bishop(the Blue ChapelComplex).The Byzantine-periodpopulationat Petrawill be discussedwithin the contextof thesestructures'historiesof use and renovation.

1. All of ACOR'sPetraprojectsweresupportedthe Departmentof Antiquities and by grantsfrom the United States Agency for InternationalDevelopment;the PetraChurchProjectwasalsosupportedby manyotherdonors(seeFiemaeta/. 2001: vi); and the Blue Chapelexcavationwas also supportedby grantsfrom the Dick and Betsy DeVos Foundation. In 1994,a large Nabataeanbuilding west of the main study areawas also documentedby Pierre Bikai (Fig. l.P); 2. dubbedthe 'CourtyardBuilding', its function has not beendetermined;seeBikai and Bikai 1994: 634-36 and Pl. 57.1.

Figure I. The city centerof Petra by ChrysanthosKanellopoulos,using a map createdby HashemiteUniversity and the American Centerof Oriental Research. Commonnamesof the structures:A: Qasral-Bint; B: Small Temple; C: TemenosGate,Towers,andthe Baths; D: Great Temple; E: GardenandPool Complex; F: Middle Market; G: UpperMarket; H : SouthNymphaeon;J:North Nymphaeon;K: RidgeChurch;L: Blue ChapelComplex; M : PetraChurch; N: ByzantineTower; P: CourtyardBuilding; Q: Templeof the Winged Lions; R: Royal Palace;S: Shops.

PETRA'S CHURCHES

437

History of the Excavations In 1973, KennethRussellnoticeda largeburied structureapproximatelymid-way up the southern face of the North Ridge. Russell identified the structureas a large Byzantinechurch in 1990 and beganplansfor its excavation.Ken unfortunatelydied unexpectedlyin 1992,preventinghim from seeingthe implementationof his project. Excavationbeganunderthe sponsorshipof ACOR in May 1992 and continued almost continuously for the next four years. The project field directors, Zbigniew T. Fiema,RobertSchick,and Khairieh 'Arnr, almostcompletelyexcavatedthe structurein lessthan a year; Fiemasubsequentlyspentfrom 1993 until1996 preparingthe areafor placement of a protectiveshelter.During the secondexcavationphase,in December1993, the project made the remarkablediscoveryof the Petrapapyri (Chapter49, this volume). Zbigniew Fiema and his colleagues'final publication on the Petra Church details their architecturaland archaeological discoveries(Fiemaeta/. 2001), which will only be summarizedhere. ACOR continuedexcavationson the North Ridge in 1994 concomitantwith the final stagesof constructionof the PetraChurch shelter. In early 1994, ThomasDailey and Pierre Bikai noted a large structureat the top of the ridge, abuttingthe northernByzantinecity wall. The extensively eroded structure was identified as another Byzantine church, later named the 'Ridge Church'. Excavationsof this structureby project director Patricia M. Bikai and assistantdirector Virginia Egansoughtto exposeandconsolidateit to preventfurther erosionand provide additionalevidence of ByzantinePetra.The churchand relatedstructures,including a small 'shrine'likely devotedto an important military official, indicate the extentof the ByzantineChurch'spresenceat Petraand a possiblelink betweenthe Ridge Church and the military. As the PetraNorth Ridge Projectcompletedexcavationof the Ridge Churchin 1998,Virginia Egan left the project and MeganA. Perry joined as assistantdirector. From 1998 until 2002, the PetraNorth Ridge Projectfocusedon the excavationof two Nabataeantombsdiscoveredduring the final phasesof the Ridge Church excavation in addition to the exploration of another large structuresituatedbetweenthe PetraChurchandthe Ridge Church.This structure,the 'Blue Chapel Complex', turned out to be another ecclesiasticalcomplex consisting of a small chapel and associatedroomsattachedto a large tripartite building. The complex,with its location convenient to the PetraChurch, probablyservedas the residenceof the Bishop of Petra.

The Ridge Church The Ridge Church was the first ecclesiasticalstructureconstructedon the North Ridge, probably not long after the AD 363 earthquakethat devastatedmuch of Petra (Fig. l.K). A single apse, benches,and two rectangularside rooms, or pastophoria,were addedto an existing building near the top of the slope for conversionto its new function. The church properconsistsof a nave and two sideaislesseparatedby stylobatesthateachsupportedfive columns.The church'ssanctuarywas accessedthrough stepsleadingfrom an atrium to a 2.5 m deepportico. The southernentranceto the churchwas reachedvia a large openplaza,which in turn was accessedby a stairwayleadingup from what would later becomethe Blue ChapelComplex. A simple floor mosaic,wall mosaics,a marblealtar and altar screens,andan elevatedbemawere installed during a later renovationphasein the mid-5th century. Many of the original decorative elementswere lost during the collapseanderosionof the structure.Almost the entiresuperstructure of the Ridge Churcherodedor was clearedmanuallyfrom the top of the ridge, leaving the walls in someplacesbelow floor level.

438

CROSSING JORDAN

The residentsof Petratook every opportunityto conserverainwaterdue to the arid climate and lack of substantialnatural water sources,and the situation at the Ridge Church was no different. The church has a large cistern under the nave, apparently fed by a series of downspouts and channelsrunning rainwaterfrom the roof to the mouth of the cistern.

The Petra Church The PetraChurch (Fig. l.M; Fig. 2) was constructedduring the mid-to-late5th century.The original structurecontaineda singleapseand two pastophoriaat the easternendsof a broad nave and two aisles.The early chancelwas separated fromthe naveby a woodenchancelscreen,remnantsof which were discoveredwithin the chancelpost holes. The floor of the nave was apparentlypaved with sandstone.Mosaicswere also installed in the southernaisle and aroundthe altar areaduring this first phase;the existenceof mosaicsin the northernaisle during this period is unknown. The mosaicof the southernaisle hasa geometricbackgroundagainstwhich are portrayedanimals,birds, and fish. Personificationsof the Four Seasons,Wisdom, Earth, and the Oceanin addition to fishermen, a fowler, and a vasewith birds also run down the center(seeFiemaeta/. 2001: 218-305). A baptismalcomplexwas then built on the westernside of the atrium. The pristine condition of the baptisteryis remarkable,makingit one of the best-preserved in the Near East.The centralroom contains a large cruciform-type font under an elaboratelydecoratedcanopy supportedby four columns.A large stonejar, possiblyfor baptismof infants (Fiemaeta/. 2001: 47-48),was built into the platform surroundingthe font. This baptisterystyle is typical for 5th-centurysouthernPalestine and Jordan. The PetraChurch underwentextensiverenovationsduring the 6th century to increasechurch facilities and embellishits decorativeelements.A bell-shapedcisternwas constructedin the center of the atrium to conserverainwaterfrom the surroundingrooftops.The complexalsowasextended to the north of the atrium by the constructionof anotherroom, possibly a chapel (Fiema et al. 2001: 54). The baptisteryand other westernroomswere enlargedto accommodatelarger groups. The churchitself also saw significantrenovationduring this phase.The rectangularpastophoria flanking the central apse were converted into additional apsesto create the extant triapsidal structure.The altar areawas raised and new marble screenswere installed. Finally, an elaborate opussectile floor of marble and stonewas put down in both the raisedaltar areaand in the nave. This expensiveflooring can be found only in the more important churchesin the region. New mosaicsadditionallywere installedin the easternend of the southernaisleandin the northernaisle. The mosaicat easternend of the southernaisle depictsdeerand ostriches.The mosaicin the north aisle, installedfrom the door to the apse,includesgrapevines emergingfrom a vasejust inside the door to createrows of three roundelsextendingup the aisle. Within the roundelsare animalsand vesselsas well as a few humanfigures (seeFiemaet al. 2001: 218-305). Elaboratewall mosaicsalsowere installedin the PetraChurch,althoughonly fragmentsof these survived the fire and destructionof the church (Fiema et a/. 2001: 300-302). A few preserved sectionsshow humanfigures againsta backgroundof mosaiccubeswith gold leaf. The Petraand Ridge Church's wall mosaicsbear striking similarity to those of the mosaic masterpieceat the Churchof Mount St. Catherine'sin the Sinai, suggestingthat artists from the sameworkshopmay have createdthe Petramosaics. Finally, a synthronon,a step-likeseatingarea,probablywith a seatfor the bishopat the center, was installedin the centralapse.Most synthronafrom Byzantine-periodchurchesin Palestineand Jordanservedas seatingfor clergy during churchceremonies.The synthrononat Petra,however, was likely an aesthetic,rather than functional, addition to the church, for the width of the tiers makesthem impractical for sitting (Fiema eta/. 2001: 78).

PETRA'S CHURCHES

439

Churchofficials also beganto usea small 5.5 m x 5.2 m room to the north of the basilica,Room I, as a storageareaduring this period (Fiemaet al. 2001: 79). Room I had beenincorporatedinto the Petra Church complex upon its constructionduring the 5th century and used for storing liturgical and other objects. A 6th-centurychurch official then beganstoring a numberof papyri datingbetweenAD 537and593/594 thatbelongedto his extendedfamily. The eventualdestruction of the church by fire carbonizedmany of the objectsin the room, including the papyri.

Figure 2. Aerial view of the PetraChurch (photo by J. Wilson Myers and EleanorE. Myers)

440

CROSSING JORDAN

The Blue Chapel Complex The Blue Chapelcomplexwascreatedby convertingearlierstructureson theslopeat approximately the same time asthe constructionof the PetraChurch.The complexconsistsof two interconnected buildings (Fig. l.L; Fig. 3). The upperbuilding is a tripartite structurecontaininga numberof small roomsalong its northernside and entrancesto the eastandwest. The central sectionof this structure hasa room with a large door to the west anda courtyard to the east. A series of small rooms comprises the northernsectorof the upperbuilding, while the southernareais broken up into a vaulted room to the westand a portico to the east.The upper building may have servedas the actual residencefor the bishopand his retinue.

Figure 3. Reconstructionof the interior of the Blue Chapel by ChrysanthosKanellopoulosand Platon Konstadopoulos

PETRA'S CHURCHES

441

The Blue Chapel, an open atrium with a single-storeyportico, and anothersmall room are locatedin a lower building, accessedsolely via a staircaseleading down from the upper building; the narrownessof the staircasemakes it unsuitable for public access.An earlier building was convertedinto the chapelby addinga wall containingthreeapses.Later, an elevatedaltar areawas constructedto the west of the central apse.The altar area has a rather poorly placed pavement reusinga randomassortmentof marblepieces,a jarring featuregiven the careshownwhile building the restof the structure.Churchofficials may havehaphazardlyrepairedan originally elegantfloor at a later stage.Finally, a throne for a bishop was installed behind the altar and benchesfor the clergy placedon either side of the elevatedbema. Constructionof the Blue Chapelincluded the addition of many decorativeand functional elements,including a reliquary in the northernapsethat was empty upon discovery.An ornatelatticework pulpit and staircasewere installednearthe altar area,which includedchairsfor the clergy. A woodenbenchwas constructedalongthe northern,western,andsouthernperimeterof the chapel, as suggestedby rectangular-shaped notcheswithin the walls and the floor. One of these holes containedremnantsof a benchpostthat provideda 14C datebetween530 and 640 CAL AD (95.4o/o [2cr] calibratedageranges,relative areaunderdistribution 0.978),suggestingthe benches may have beenpart of later renovationsof the Blue Chapelduring the late 6th to early 7th centuries. The Blue Chapelcomplexutilized numerousfragmentsforagedfrom elsewherewithin the city, similar to the other Byzantine ecclesiasticalstructureson the North Ridge. These include four Anatolian blue granite columnstopped with well-preservedlimestone Nabataeancapitals. The columnsand capitalsof the Ridge Church and the PetraChurchsimilarly were scavengedfrom a variety of earlier monumentsin Petra (see Fiema et al. 2001: 172-73). Architects of the Blue Chapel,however,soughtto follow a cohesivedecorativeplan. The blue columns,with their white limestoneplinths, bases,and two-part capitals,were numberedand then movedto the site as a set. Additional effort was madeto carry the blue and white programthroughoutthe monument.The walls were plasteredwhite andthe floors werepavedwith blue-huedsandstone.The bluish marble of the pulpit, the chancelpostsand screens,and the reliquary continuedthe decorativeprogram createdby the blue columns. Excavation in the chapel and other parts of the lower building additionally uncoveredquantitiesof blue and white marble fragmentsthat may have coveredthe bishop'schair behind the altar and the seatsfor the clergy on either side of the altar. The chapelitself is very small; the 'proper'-theinterior areaexcludingthe apsearea-measures only 111 m2 as comparedto the Ridge Church at 158 m2 and the Petra Church at 358 m2 • The exactuseof this complexremainsunclear;however,the private natureof the chapel,the presence of a bishop'sthrone, and the proximity to the Petra church suggestsit may have servedas the residenceof the church'sbishop or anothermemberof the city's elite.

Byzantine Petra The excavationof thesethreeecclesiasticalstructuresand the discoveryand ongoingtranslationof the Petra papyri have contradictedearlier depictions of an abandonedByzantine Petra. Many structuresin Petra'scity centersufferedsubstantialphysicaldamageafter the A.D. 363 earthquake. Some,such as Qasr ai-Bint, the Great Temple, and the Temple of the Winged Lions, were never rebuilt, leading many scholars to surmise that the earthquakeprecipitated a decline in the populationand prominenceof Petra(Hammond1996: 6; Joukowsky1998: 139; Zayadine2003: 96). Many Byzantineperiod burialsalso were discoveredwithin thesestructures(Hammond198788: 82-83;Joukowsky2001: 336; Zayadine1982: 380; 2003: 96), suggestingthat the westernend of the former Nabataeancity centerwas usedasa cemeteryafter the 4th century.Indeed,in the city center,with the exceptionof the a few shopsrebuilt on the south side of the easternend of the

442 CROSSING JORDAN main road (Fig. l.S; Fiema 1998; Kanellopoulos2001), most of the areasouth of the main street was abandonedafter the earthquake. The westernend of the former city to the north side of the streetalso appearsto have been partially abandonedafter the 4th-centuryearthquake.The Byzantinecity centerat this time was centeredto the eastof the ecclesiasticalstructures.This was a city much smallerand lessimposing than in its Nabataeanform, but home to a vibrant community. A group of papyri documentinga variety of transactionsof an extendedfamily depositedin a room adjacentto the Petra Church provides one glimpse into Petra's6th-centurypolitical economy.The Petra papyri indicate that Petra was well integratedwithin the Byzantine imperial legal and taxation system (see Caldwell 2001: 67 and Chapter49, this volume). By the end of the 6th century,half a millennium had passed since the heyday of Petra'srole as a major trade entrep6t.Inhabitantsinsteadwere largely supportedby a localized,agricultural-basedeconomy(e.g. Caldwell2001:34-38).The papyri also note the PetraChurch was dedicatedto the 'Blessedand All-Holy Lady, the most Glorious Mother of God and Ever-Virgin Mary' (Fiemaet al. 2001: vii). The original namesof the other two churches on this slope are not known. At some point, likely during the 7th century, the three ecclesiasticalstructureson the North Ridge were abandoned,possibly due to a population decline in Petra. The PetraChurch burned either after cessationof its ecclesiasticalfunction or while it was in usebut undergoingrenovations (Fiemaet al. 2001: 80-91).The latestidentified datefrom the Petrapapyri, AD 593/94,implies that the PetraChurch remainedin use until the end of the 6th century. The churcheseventuallywere reoccupiedby someremainingresidentsof Petraduring the 6th-8thcenturies,usingthe abandoned structuresfor living and work space.These small groups also concentratedon looting the site, focusing primarily on the Nabataeanshaft tombsthat cover the upper sectorof the North Ridge. They continued these activities until at least the earthquakeof AD 748/749, which toppled the columnsof the Blue Chapel. The exactcauseof Petra's7th-centurypopulation decline remainsunclear. Little to no maintenanceof the city's water infrastructuresurely maderesidenceat Petra,with its paucity of springs or other natural water sources,difficult. Judgingby the amountof water availabletoday from the spring in the wadi below the Ridge Church,the post-7thcenturycity probablycontainedno more than a few hundredpersons.

References Bikai, Pa.M., and Pi.M. Bikai (1994) ACOR Excavationsat Petra 1994. In RicercaStorico-Archaologicain Giordania XIV-1994, pages634-36, edited by M. Piccirillo. Liber Annus44: 619-62. Bikai, Pa.M., and M.A. Perry (2001) PetraNorth Ridge ProjectTombs1 and 2: PreliminaryReport.Bulletin of the AmericanSchoolsof Oriental Research104: 59-78. Caldwell, R.C. (2001) Between State and Steppe: New Evidence for Society in Sixth-Century Southern Transjordan(PhD dissertation,University of Michigan). Fiema, Z.T. (1998) The Roman Street of the Petra Project, 1997: A Preliminary Report. Annual of the DepartmentofAntiquitiesofJordan 42: 399-424. Fiema, Z.T., C. Kanellopoulos,T. Waliszewski, and R. Schick (eds.) (2001) The Petra Church (Amman: American Centerof Oriental Research). Hammond, P.C. (1977-78) Excavationsat Petra 1975-1977.Annual of the Departmentof Antiquities of Jordan 22: 81-101. -(1996)The Templeofthe WingedLions, Petra, jordan, 1973-1990(FountainHills, AZ: PetraPublications). ]oukowsky, M.S. (1998) Petra Great Temple,I (Providence,RI: Brown University Press). -(2001) Brown University 2000 Excavationsat the Petra Great Temple. Annual of the Departmentof Antiquitiesofjordan 45: 325-42.

PETRAS CHURCHES

443

Kanellopoulos,C. (2001) The Architecture of the Shopsand ColonnadedStreet in Petra. Bulletin of the AmericanSchoolsof Oriental Research324: 9-22. Zayadine,F. (1982) RecentExcavationsat Petra (1979-1981).Annual of the Departmentof Antiquities of jordan 26: 365-93. -(2003) Evaluationchronologique.In Le Qasr al-Bint de Petra: /'architecture, le decor, Ia chronologie,et les dieux, edited by F. Zayadine, F. Larche, and J. Dentzer-Feydy (Paris: Editions Recherchesur les Civilisations): 81-97.

This page intentionally left blank

ISLAMIC CIVILIZATION IN JORDAN

This page intentionally left blank

51

From Nabataean King to Abbasid Caliph The Enduring Attraction of Hawara/ ai-Humayma, a Multi-cultural Site in Arabia Petraea John Peter Oleson

Scholarly Context and ResearchQuestions Ancient Hawara(in Nabataean,HWR; in Greek,Auara; in Latin, Havarra),now officially called aiHumayma,was a small deserttrading post andcaravanway-stationin Edom, the desertregion of southernJordan.It is located 80 km south of Petra,the capital of the Nabataeankingdom, and 80 km north of 'Aqaba (ancientAyla), on the Red Sea.According to Ouranios'sArabika, Hawara was foundedby a Nabataeanprince and later King Aretas-probablyAretasIII (84-62/59BC) rather thanAretasIV (9 BC-AD 39/40), sincearchaeologicalevidenceat the site commencesaround50 BC. Ouranioswrites (FGrH 675 frag. A.l.b): Auara: town in Arabia, namedby Aretasfrom an oracularresponsegiven to his father Obodas.Aretas set out in searchof the oracle'smeaning,for the oraclesaid 'to seekout a placeauara'-whichin the Arabian and Syrian languagesmeans'white'. And as he lay in wait, a vision appearedto him of a man clothed in white garmentsriding along on a white dromedary.But when the apparitionvanished,a craggy hill appeared,quite natural and rooted in the earth; and there he founded the town.

The name'White' and the white camel that led Aretasto the site may refer to the rounded,humpy hillocks of white Disi sandstonethat surroundHawaraon all sidesbut the west, or to a pre-existing camelmarketorganizedby local Bedouinbreeders.A shrine found at the site seemsto be oriented towardsthe red sandstonejabal (hill, mountain)that looms over Hawaraon the west, reinforcing the connectionwith a 'craggy hill' that also featuresin the foundation myth. The location and historical contextof the settlementsuggestit was intendedto serveasa centrefor sedentarizationof the nomadicNabataeanpastoralistswho occupiedthe region.Throughcareful managementof the meagerspringwaterand precipitation,the resultingcommunitywasableto enjoy a settledexistence basedon agriculture,stock-raising,and the servicingof caravans.A modestprosperitycontinued throughthe Roman,Byzantine,and Early Islamic periods,basedin part on the tradethat continued

dlt

076.

061

CHURCH

8126 BYZANTINE

B~H

I £125

.·.JJ

' E 122

E077

Figure L Map of structuresat al-Humayma/Hawara

CHURCH 066.

.!?.

010•

012e 01. . 078. 073.

one

~ ~~~~R

g~s~cH

060.

100

0~

S£At.lf'RASER

7

~

\1M]

W CH TOWER OR A MHOUSE COMPLEX

200m

HUMAYMA 2005

FROM NABATAEAN KING TO ABBASID CALIPH

449

to move along the north/southroad, the Via Nova Traiana, built by Trajan on the old Nabataean route that extendedfrom Damascus,pastPetra,to Ayla. The Abbasid family purchasedthe village site late in the 7th century,built a qasr (large residence)and family mosque(identified in 1992and 1993), and plotted the overthrow of the Umayyad caliphate there. After the successof their conspiracyin the mid-8th century,and the shift of the caliphateto Baghdad,the site wasessentially abandoned.The nameand historical associationsof the site survivedamongArab historiansandare still familiar to the local Bedouin. The site is beautiful to moderneyes:surroundedby white and red-brownsandstonejabals that rise steeplyfrom sandysoil; framed on the north by the steepash-Sharaescarpment,600 m high and 100 km long; enlivenedby a sparsebut beautiful desertflora and fauna. The stony, sandy region is now called the Hisma; in antiquity it was the southernportion of Arabia Petraea-'Stony Arabia'-which echoesthe nameof the capitalcity Petra,only 80 km to the northwest.Camelsstill serve the Bedouin inhabitantsas a picturesqueand practical meansof transportand subsistence. Nevertheless,life in this region poseschallenges.The annual precipitationaround Hawaratotals only 80-100mm, and this moistureis suckedaway by the strong,dry wind and the sun, resultingin an annual pan-evaporationof more than 3,000 mm. The region is classifiedas Hyper-Arid.

Summary of Field Work ( 1986-2005) To an outsider,the questionimmediatelyarisesof why so many rulers focusedtheir attentionon this isolated,seeminglyunremarkable,indefensible,waterlesssite. King Aretas,the emperorTrajan, Byzantinepriests,and the Abbasid family all focusedtheir attentionon Hawara/al-Humayma.In 1986, with a team of Canadianscholarsand students,and supportedby the Social Sciencesand Humanities ResearchCouncil of Canadaand the Taggart Foundation,I beganto researchthis question.Our work hascontinuedwith extraordinaryresultsuntil the present;therehasbeenlittle reoccupationof the site, so the archaeologicalremainsare consequentlywell preserved. My teamsurveyedand probedthe water-supplysystemin the settlementcentreandsurrounding region between1986 and 1989 as an exploration of cultural responsesto a desertenvironment (Oleson1986, 1988, 1990). I then undertookexcavationof the settlementcentrein ten field seasonsbetween1991 and 2005 (seeOlesonet al. 1993, 1995, 1999, 2003 for reportsand further bibliography).At varioustimes Khairieh 'Amr, RebeccaFoote,RobertSchick, M. BarbaraReeves, Erik de Bruijn, and Andrew Sherwoodhaveservedas co- or assistantdirectors.The fieldwork was carefully focusedon critical structuresand areasacrossthe large site in order to elucidateevery stageof its history, from first settlementto abandonment.The strategyof using both exploratory probesand focusedexcavationhas succeededadmirably (Fig. 1). Over this period the team excavateda Nabataeancampgroundand threeNabataeanand Late Romanhouses,the Romanfort and associatedbath, excavatedand probedfour Byzantinechurches,excavatedand consolidateda fifth church,and excavatedtwo Umayyad-Abbasidfarmhouses.We also excavatedmostof a large mudbrick complexforming part of the vicus,the civilian settlementassociatedwith the Romanfort. M. BarbaraReeveslater took on direction of this areaand identified a remarkableNabataeanshrine that continuedin useinto the Late Romanperiod. Partof the team,underthe direction of Rebecca Foote,also excavatedthe largeqasr (large residence)and mosquethat servedas the residenceof the famousAbbasidfamily (seeChapter52, this volume). Finally, we haveprobedmiscellaneousstructures within the settlementcentre and sampled many of the partly plundered rock-cut tombs surroundingit. We also consolidatedthe bath building, one church, the mosque,portions of the fort, and generallycleanedup the site for tourism. In 2006,with the supportof the CanadaFundin Jordanand the collaborationof the Friendsof Archaeology,RobbynGordon,RebeccaFoote,and I supervisedthe renovation of the galleries in the ArchaeologicalMuseum in 'Aqaba that are

450 CROSSING JORDAN dedicatedto the finds from al-Humayma,and arrangedthe installationof an exhibition of photographsand informative placardsin the Visitor's Centrerecently constructedat the site.

ResearchQuestions In addition to the overarchingquestion of precisely 'why' Hawara was so important, several subsidiaryclustersof researchquestionshave guided our excavations: • • • • •

What were the chronology,architecturalcharacter,and economicstructuresof the earliest settlement,and how can the exampleof Hawaracastany light on the processof sedentarization that Nabataeansociety as a whole was going through in the 1st century BC? What role did the water-supplysystemplay in the location and designof the settlement? When did the Romanoccupationof the site begin,and how did Romanadministrationand influence affect the water-supply system, settlementdesign, structures,and the local culture?What role did the fort play as part of the easternimperial frontier? What was the chronologyand what was the natureof the Christianizationof Hawara,and how deeply did Christianity and Byzantineculture affect the local people? What effect did the arrival of Islam have on the settlementand its inhabitants,and where did the Abbasid family build their historically attestedqasr and mosque?

This accountcan only summarizethe datawe have recoveredthat help to answereachof these researchfoci.

Nabataean Hawara In 1991-93I hopedto find evidencefor the earliest,presumablyearly 1stcenturyBC, occupationof the site in the centreof the settlement,nearsomereservoirsof typical Nabataeandesign(nos. 6768). Becauseof the presenceof deep silt depositslaid down during the Byzantine period and of unexpectedlyrich Islamic occupationstrata,only tracesof early 1st-centuryAD Nabataeanstructuresand-surprisingly-tombs were recoveredthere by 1993. Surveyand spot excavationin the plunderedrock-cuttombssurroundingthe town, however,yieldedcoinsandceramicsthat provided ample evidencefor the Nabataeanfoundation of Hawaraat somepoint in the 1st century BC. In 1996 we finally identified and excavatedtwo Nabataeanoccupationareas.Area C124, an open field adjacentto a Nabataeantype cisternon the south edgeof the settlement,yielded enormousamountsof 1st-centuryAD Nabataeanpottery and many contemporarycoins. The character of the depositssuggestsa campgroundfor caravansor seasonalinhabitantsof Hawara,locatedadjacent to a handy sourceof water. While this campgrounddid not date to the earliesthistorically attestedphaseof the site, it correspondswith the usual model for Nabataeanurbanism,in which nomadicbehaviorhelps createsettlementsand continuesalongsidepermanentarchitecture.This model has beendocumentedat Petraas well. In another open field, south of the Roman fort and on a rise above the settlementcentre, we uncoveredthe remainsof severalNabataeanstructuresin 1995-96 and 1998-2000beneath later buildings that belongedto the vicus, the civilian settlementserving the fort. The Nabataean houseat E122 was rebuilt on more or less its original plan in the early 2nd centuryAD. Another structure, beneaththe south end of E125, was a shrine or temple, probably destroyedduring the Roman takeover, but reused through the Late Roman period. All in all, the artifacts and architectureof NabataeanHawarafit the model of small Nabataeansettlementsof the 1st century BC and AD.

Q AR EA M

AREA C

ARfA tt .oRK9«lP 0' 8

AA EA P

E 1 1 6 - ROMAN F ORT

AR EA 8

HUMAYMA 2 0 0 5

Figure 2. Ma p of structuresat ai-Humaym

a!Hawara

452 CROSSING JORDAN

The NabataeanWater-Supply Systemof Hawara The water-supplysystemofNabataeanHawarawas regionalin scaleand integratedwith the settlementdesign,and the settlementlocation seemsto havebeenselectedwith this critical consideration in mind (Oleson1997).The settlementis locatedat the conflux of severalrun-off fields that provide a reliable supply and manageableamountof water to two public reservoirsand numerousprivate, domesticcisterns.Furthermore,it sits at the southernmostpoint that could convenientlybe reached by a gravity flow aqueductfed by springson the escarpment15 km to the north, and it is neargood agriculturalsoil and a route to Wadi 'Arabah.The scaleof the 1st-centuryBC or AD public reservoirs and aqueductindicatecentral,probablyroyal, planningandsponsorship.Drainagewasmorecasual, but it alsoinvolved somesite planningand construction(Oleson1996). Our surveyhasdocumented 61 cisternsin the settlementand within the 250 km2 catchmentarea,an aqueductsystem27 km long that brought spring water from the crest of the ash-Sharaescarpment,3 aqueduct-fed reservoirs,3 reservoirsfed by run-off, and a variety of terracesand wadi barriersto assistagricultural activity. The only post-Nabataean addition to the systemwas the reservoirbuilt as part of the Romanfort early in the 2nd centuryAD. The regionaltotal for the storagesystemwas 12,300 m3; for the urbancenter4,300 m3• Wheatand barley were cultivatedin the light, but fertile soil around the settlement,using water that ran off the surroundingsandstonehills during winter storms.

Roman and Late Roman Havarra Excavationof portions of the walls and interior structuresof the large (148 x 207 m) Romanfort (E116) between1993 and 2005 revealedthat the structurewas very well preservedand that it was built severalcenturiesearlier than expected(Fig. 2). It was traditionally assumedthat the fort was part of Diocletian'sreworkingof the frontier in the late 3rd century,but we recoveredceramicsand coins proving that the structurewas built very soon after Trajan's annexationof the Nabataean kingdom in 106. This is the earliestlarge Romanfort known in Jordan,and the useof a rectangular plan with projectingtowersat suchan early datehasderailedtheoriesconcerningthe development of Romanforts in the East. Geophysicalsurvey with ground-penetratingradar (GPR), resistivity, and magnetometrymethodsin 2003 and 2004 helpedto clarify detailsof the defensesand interior plan of the fort. We have documenteda decline in activity at the fort in the later 3rd century, rebuilding of someinterior structuresin the 4th-possiblyfor non-military occupation-andabandonmentearly in the 5th century (Olesoneta/. 1995, 1999, 2003). The original plan of the fort, which was laid out arounda grid of pavedstreets,included a principia (headquartersbuilding), praetorium (commander'sresidence)with a fountain and a dining and bathing areawith mosaic floors, horrea (granary),latrine, barracks,and workshops.All thesestructureshavebeenidentified. It is interestingto notethat the exchangeof cultural influencebetweenthe Nabataeaninhabitantsof Hawaraand the Romanoccupierswent strongly in both directions.The mosaicsin the bath areain the praetorium were executedby a Nabataeanworkshop that had been active for the previous decadesin Wadi Musa, outsidePetra.Furthermore,nearly all the fineware pottery usedin the fort during the 2nd and 3rd centurieswas manufacturedby Nabataeanworkshopsin Petra. Roman forts, occupiedby hundredsof soldiers, did not exist in isolation. They were either locatedat the edgeof an existing communityor a new communitysprangup outsidetheir walls to servicethe garrisonwithin. Soldierswent to thesevici for recreation(baths,taverns,and brothels), for worship, to visit their unofficial families, or to purchasestaplesor luxuries not availablein the fort itself. StructureE125, 80 m south of the fort, is a multi-unit stone and mudbrick complex constructedin the late 2nd centuryover theruins of severalNabataeanstructures.In its southeast

FROM NABATAEAN KING TO ABBASID CALIPH

453

corner,a long, pavedprocessionalway, orientedapproximately on the mountainbehindthe site, led to the Romanrebuild of a Nabataeanshine (Fig. 3). In its Romanphase,this shrinecontaineda column representinga Nabataeangod, Dushara,anddedicationsto JupiterAmmon and to Serapis. The Latin inscription on the altar to JupiterAmmon indicatesthat the soldiersstationedat the fort belongedto the Third Legion Cyrenaica, one of the two legions involved in the conquestof the Nabataeankingdom (Olesoneta/. 2002).The inscriptionalsomentionsthe Latinizedversionof the nameof the site: Havarra.

Figure 3. View of the processionalway and the Late Romanshrine;JabalQalkha is in the background

Byzantine Havarra Christianswere probably residentat Hawarafrom at leastthe 3rd centuryAD onward,given the early and enthusiasticinterestof Arab groups in the new religion, particularly in the Provincia Arabia. The first physical evidence,however,are the churches,which begin to appeararoundAD 400. We haveexcavatedand consolidatedthe largestand bestpreservedof these(ClOt), a threeapsebasili ca with chancel,fine marblechancel screen,five undisturbedburials, sacristy, and side entrancehall. A small, single-apsechapel(C119)-butwith paintedplasterwall s, marblefittings, and glasslamps- was partly excavatedon the westernslopeoverlookingthe site in 1993. Subsequently, we recognizedthat two sprawling Early Islamic habitationsunderexcavation(BlOO and F102) had been built into and on top of two more, large, single apsechurches,and that a late Ottomanhousenearthe centreof the site concealed the remainsof yet anotherlarge, single-apse church (B126)(for preliminary reports,seeOlesoneta/. 1993, 1995, 1999). The inhabitantsof Hawarabuilt at leastfive substantialchurches,the samenumberas at the larger mining centreof Faynan,and one more t han the numberso far identified at the regional capitalof Petra.This total seemsto be excessivefor a settlementwith a populationof no morethan

454 CROSSING JORDAN 650 persons,many of them probably residentat the site only part of the year and living in tents (Oleson1997: 181-82).It is possiblethat the doctrinal disputesendemicto this era fosteredseparate congregations,that Havarraservedas a religious centrefor Christiansthroughoutthe region, or that someof the churcheswere derelict when otherswere built. Although Hawaramust have beenthoroughly Christianizedby AD 400, the fort has not yet yielded any evidencefor Christian practices,raising interestingquestionsabout the characterof Romanizationat the site and the intensity of interactionbetweenthe soldiersand the civilians.

AI-Humayma and the Abbasid Family Christianitywitheredaway quickly after the arrival of Islam in the mid-7th century.At leasttwo of the churches(ClOt and BlOO) were abandonedfor a time, then destroyedby fire. Threeof them (BlOO, B126, and F102) were divided up by party walls and reoccupiedas habitationsin the Early Islamic period,prior to final abandonment in the first half of the 8th century.The initial occupation of the region by Islamic forceswas largely peaceful,but by the late 7th centurythe Christianpopulation of Hawara-nowcalledal-Humayma-hadconvertedor departed.Accordingto early Arabic historians,around 687/8, 'Ali ibn 'AbdAllah ibn al-'Abbas purchasedthe town and built a qasr (large residence)andmosque.Here, the Abbasidsplotted their revolt againstthe Umayyaddynasty, which was carriedout in 749. The family then left for Iraq (Oleson2001; Foote,Chapter52, this volume). We identified the Abbasidqasr in 1992.It is a rectangularstructure(ca. 61 x 50 m) consistingof a large trapezoidalcourtyardsurroundedby roomsfronting the court. Therewas a wide, recessed entranceon the east, and a small mosque (ca. 5.7 m2) outside the southeastcorner. Ceramics provide rich and precisedocumentationof the date,and the remainsof frescoesand carvedivory furniture indicatea tastefor luxury andwide commercialconnectionsappropriateto this politically active family (Olesoneta!. 1999, 2003).

Conclusions In summary,the original occupantsof Hawaracreateda water-supplystrategyappropriateto local conditionsand materialsand dependenton technologiesdevelopedor perfectedover the previous century at other Nabataeansettlementsites. The water-supplysystemwas linked with a mixed pastoral/agriculturaleconomicstrategyenlivenedby the presenceof a long-distancetrade route. The systemfunctionedwell enoughthat there was little expansionof facilities after the fall of the Nabataeankingdom, despite occupationby a substantialRoman military force and continued intenseoccupationof the site throughthe Early Islamic period. It is clearthat King Aretasplanneda sustainablewater-supplysystemthat functionedeffectively in a hyper-ariddesertfor 800 yearsand supportedthe only extensivesettlementin the Hisma.This watersystem,alongwith the arablesoil and the road, constitutedthe secretof Hawara'sappealto settlers.

References Oleson,J.P. (1986) The HumaymaHydraulic Survey: PreliminaryReportof the 1986 Season.Annualof the DepartmentofAntiquitiesin Jordan 30: 253-60. -(1988)The HumaymaHydraulic Survey: PreliminaryReportof the 1987Season.Annualofthe Department of Antiquities in Jordan 32: 157-69. -(1990)The HumeimaHydraulic Survey: PreliminaryReportof the 1989Season.Annualofthe Department of Antiquities in Jordan 34: 285-312.

FROM NABATAEAN KING TO ABBASID CALIPH

455

-(1995) The Origins and Design of NabataeanWater-Supply Systems. In Studies in the History and Archaeologyoffordan, V, edited by A. Hadidi (Amman: Departmentof Antiquities of Jordan):707-19. -(1996)SurfaceWaterManagementat an Ancient Site in Jordan'sSouthernDesert.In Acts, 16th Congressof the International Commissionon Irrigation and Drainage, Cairo (Cairo): 141-56. -(1997) Landscapeand Cityscapein the Hisma: The Resourcesof Ancient Al-Humayma. In Studiesin the History andArchaeologyofJordan, VI, editedby M. Zaghloul (Amman: Departmentof Antiquities): 17588. -(2001) King, Emperor,Priest,and Caliph: Cultural Changeat Hawara(Ancient ai-Humayma)in the First Millennium AD. In Studiesin the History and Archaeologyof jordan, VII, edited by K. Amr (Amman: Departmentof Antiquities): 569-80. Oleson,J.P.,K. 'Amr, R.M. Foote,]. Logan, M.B. Reeves,and R. Schick (1999) PreliminaryReportof the aiHumaymaExcavationProject, 1995, 1996, 1998. Annualof the DepartmentofAntiquitiesofJordan 43: 411-50. Oleson,J.P., K. 'Amr, R.M. Foote, and R. Schick (1995) Preliminary Report of the Humeima Excavation Project, 1993.Annual of the Departmentof AntiquitiesofJordan 39: 317-54. Oleson,J.P., K. 'Amr, R. Schick, R.M. Foote, and]. Somogyi-Csizmazia(1993) The Humeima Excavation Project,Jordan:PreliminaryReportof the 1991-1992Seasons.Annualofthe DepartmentofAntiquitiesof Jordan 37: 461-502. Oleson,J.P., G.S. Baker, E. de Bruijn, R.M. Foote,]. Logan, M.B. Reeves,and A.N. Sherwood(2003) Preliminary Reportof ai-HumaymaExcavationProject,2000, 2002.Annualofthe DepartmentofAntiquities in jordan 47: 37-64. Oleson, J.P., M.B. Reeves,and B. Fisher (2002) New Dedicatory Inscriptions from Humayma (Ancient Hawara),Jordan.ZPE 140: 103-21.

This page intentionally left blank

52

From Residence to Revolutionary Headquarters The Early Islamic Qasr and Mosque Complex at ai-Humayma and its 8th-century Context Rebecca M. Foote

When one thinks of who was crossingand controlling Jordan in the early 8th century CE one naturally thinks of the Umayyads,who ruled from 661-749.The Islamic caliphateexpandedits authority during the reign of the Umayyads,westwardinto present-dayPortugal, eastwardinto present-dayPakistan,and northwardinto present-dayUzbekistan,makingthe Umayyadcaliphatea world empire.The largessgainedthroughthe expansionfunded,amongotherthings, a considerable amountof monumentalbuilding. The official capital was at Damascus,where the Umayyadsbuilt the greatmosque,which still standslargely intact today. There was further building activity, both urban and rural, throughoutBilad al-Sham (literally the 'land of the north', expressedfrom an Arabian view, consistingof present-daySyria, Lebanon,Jordan,Palestine/Israel;it will be usedinterchangeablywith 'the Levant'in this chapter).Recentexcavatedfindings in urbancontextsin Jordan include the palacecity on the 'Amman citadel and the mosqueat the South Tetrapylonin Jarash. Significantbuilding activity carriedout in rural contextsduring Umayyadrule consistsmainly of the so-calleddesertcastles/palaces, numberingnow more than 20 and scatteredthroughoutBilad alSham.While the environmentsof mostof thesemonumentalbuildingsand building complexesare indeedarid; almostall are in fact locatedin steppeand semi-desert,and the word 'castle'overstates the fortified aspectof most; hence,a more apt term to use is simply the Arabic word qasr (pl. qusur),which is found in the historic written sourcesusually meaning'large residence'.Among the most notablesitesof the Umayyad-periodrural qusurin Jordanare 'Amra (especiallyits celebrated bath, the major surviving monumentof the multi-building complex there), al-Kharana,and alHallabat (fashionedwithin and around what was originally a Roman fort), all located between 'Amman and al-'Azraq. The qusurare usually attributedto Umayyadfamily patronage,yet at least one amongthem was of anotherelite family, that of the Abbasids,at al-Humayma,which is located in the south of Jordan(seemap on p. xxiii of the presentvolume). Sevenseasonsof archaeologicalexcavationin Field F103 at al-Humayma(ancientHawara)have unearthedan early 8th-centuryqasr andmasjid (mosque)complex,which is both the first definitive exampleof building patronageby a non-Umayyadelite during the Umayyadperiodandthe earliest extant material culture of the Abbasids (the dynasty that succeededthe Umayyadsand reigned

458

CROSSING JORDAN

750-1250CE).' The historic Arab writers relate that al-Humaymabecamehome of the Abbasids sometimeafter 687-8 CE and remainedso until the mid-8th century;thus it was their homeduring the decadesprior to their mid-8th-centurypolitical takeover. The findings of the excavationdisplay both similaritiesto anddifferencesfrom the otherregional qusur of the early Islamic period. For example,the qasr plan is similar to only a few othersin the Levant, while the small extramuralmasjid resemblesseveralin the region. Other aspectsdemonstrateintegral connectionsto variouspartsof the vastUmayyadempire,suchas finely carvedivory veneer panels from Khurasan (northeasternIran), coinage from Wasit (lower Iraq), decorative soapstonecookwarefrom Arabia, and agriculturalpracticefrom the Hijaz (northwesternArabia). Combined,the findings expandour understandingof the building milieu during Umayyadreign, in locale, form, and patronage,as well as reveal the tastes,affiliations, and activities of the Abbasids before and during their rise to caliphal power.

The Abbasids at ai-Humayma through Text2 The Abbasiddynasty'sprogenitor,'Abbas,was of the Meccanelite, a paternaluncle of the Prophet Muhammad.At somepoint after 687-8 but before705 CE, 'Abbas'sgrandson'Ali movedthe family to Bilad al-Sham, following the direction of his father, 'Abbas'sson 'Abd Allah, just before his death.A strugglefor control of the caliphateinspiredthis migration north: the Abbasidssidedwith 'Abd al-Malik, the Umayyad caliph reigning from Bilad al-Sham 685-705, rather than the rival caliph 'AbdAllah ibn Zubayr,who controlledthe religiousheartlandin northwestArabia calledthe Hijaz (where Meccaand Madina are located)until he was defeatedin 692. After consideringa numberof locations,'Ali ibn 'AbdAllah ibn 'Abbas'bought'the 'village' of al-Humayma,where he reportedly built a qasr and a masjid and had 500 olive trees. Family membersremainedbasedat al-Humaymaafter 'Ali's deathin 735, until the late 740swhen they fled eastwardto Iraq after the reigning UmayyadsrealizedAli's grandsonshad designson the caliphate. There is no clear reasongiven in the written historical sourcesas to why the Abbasidsselected al-Humayma.While the site today looks to be remote,evendesolate,al-Humaymahad sustaineda settledagrariancommunity(mixed with pastoralism)sincethe Nabataeanperiod. It is describedas near the haj (the annualMuslim pilgrimage to Mecca) route, and 'Ali is noted for his hospitality, which he offeredto passingtravelersbetweenSyria and the Hijaz. Furthermore,the Abbasidfamily still had ties with the Hijaz, so accessto points southwould havebeendesirable.Thus, the location could arguablybe seenas a perfectplaceto settle,providing both local sustenance and reasonable accessto the Umayyadcapital, yet offering a shorterjourney to their personaland religious homeland than sitesin the northernreachesof Bilad al-Sham(from Damascus,al-Humaymais aboutonethird the distanceto the Hijaz). In any case,at the time of choosingal-Humaymathe Abbasidsdid not have intentionsfor the caliphate.That was to developover subsequentdecadesby thosewho followed 'Ali.

1. The excavationsat al-Humaymahavebeendirectedby John PeterOleson,University of Victoria, andcarriedout with funding providedby the Social Sciencesand HumanitiesResearchCouncil of Canadaand the TaggartFoundation,with additional funds for the qasr-mosquecomplex from the Max van BerchemFoundation,BarakatTrust, British Academy, Aregon Corporation, and the Harris Endowmentof the American Schools of Oriental Research.The support of the Departmentof Antiquities of the HashemiteKingdom of Jordanhas been invaluable.The American Center of Oriental Researchand the work of many professionalandstudentvolunteershavebeenessentialto the successof the excavation.See Chapter51 in this volume, the contributionof J.P. Oleson,for discussionof the findings in other areasof the site dating to the earlier periods. 2. Akhbiir 1971: 107-8,131,142-45,149-51,154,195; al-Bakrl 1945-51:130; al-Baladhurl1978:72-73,75,121.

Humoymo 2002 F1 03 Qosr and Mosque(s)

, Key: .

N

j_

Original Qasr Architecture (excavated in 2002) Later architecture (excavated in 2002)

~

Conjecture

fa

Consolidated walls (in 2002)

20m

0 -

Figure 1. Planof qasr-mosquecomplexin Field F103

1::500

460

CROSSING JORDAN

Excavated Findings3 and Analysis Qasr

The Abbasidqasr-masjidcomplex is locatedat the southeastextremity of the al-Humaymasettlement (seeal-Humaymasite plan, Chapter51, Fig 1). The qasr is a large rectangularbuilding, measuring approximately61 x 50 m, and is centeredon an interior trapezoidalcourtyard,with evidence for a portico fronting the rooms on the westernside of the courtyard(Fig. 1 in this chapter).The main entranceis locatedalong the easternside of the building, inset from the line of the exterior. Interconnectedrooms of different sizessurroundthe courtyardin concentricrings on the eastern, northern,and westernsides,with an additional set of rooms on the southernside. Up to six coursesof the original masonrywalls survive in situ beneaththe surface,to a depthof nearly 2 m. The only room that clearly indicatesits original function is a bakery, located in the northwestcornerroom. Only one room (namedthe 'FrescoRoom', discussedbelow) wasoriginally ornamentedand containedhigh-statusmaterials(e.g. ivory). At leastpartsof the qasrwere lived in after the Abbasidsleft, but the natureof the occupationwas lesscoherentuntil the qasr receiveda major remodelingduring the Ottomanperiod (16th-18thcenturies). In the most general terms, the large qasr at al-Humaymawith a rectangularplan of rooms surroundinga courtyardcan be comparedwith the other Umayyad-periodqusurin Bilad al-Sham, yet it is not truly palatial. The masonryis cut blocks, someof it dressed,but much of it is spolia reusedfrom Nabataean,Roman, and Byzantine structuresat the site, and apart from the Fresco Room there is no ornamentation.The qasr is a big house,a large residence,but not a palace. Aspectsof the qasr-masjidcomplexthat display building practicesunlike mostother rural qusur erectedduring the Umayyadperiod include a recessedprincipal entry, wall lengthsthat are plain ratherthan punctuatedwith towers, and interconnectedroomsof varioussize and numberrather than a repeatingapartmentmodule. While it is often arguedthat the qusur locatedfurther north adoptedfortified architecturalmodelsof the Late RomanandByzantineperiods,the qasrplan at alHumayma,located in the very south of Jordan,may insteadreflect building traditions from the ArabianPeninsula,which the Abbasidscould havebroughtwith themwhenthey migratedfrom the Hijaz, or it may be an exampleof a surviving indigenousTrans-Jordanianbuilding tradition. Fresco Room and Ivory-veneered Objects Especiallyremarkablefindings datingto the early Islamic periodare in the room off the westsideof the courtyardthat is nearly on axis with the corridor foyer leadingfrom the main entry of the qasr into the east side of the courtyard. The deepeststratum in this room was createdby a fire and containedpaintedfresco fragments(henceits name,the FrescoRoom), finely carvedbut mostly charredivory fragments,large concentrationsof burnt wood, and much corrodediron. The pattern-paintedplasteroriginally decoratedthe upperwalls and possiblythe ceiling. There were a wide variety of expressivedecorativedesignsused, primarily floral and geometricwith a pearl-beadedborder, paintedin a rangeof colors in fresco secco(Fig. 2). The samepearl-beaded border is usedin more complex contemporaryfrescoes,for exampleat Gharbi (Syria) and 'Amra Uordan).The differencein quality may be due to caliphal patronagebeing able to employ higherskilled craft workersin urbancenters,suchas Damascusand 'Amman,which are nearerto the other qusur. Thereis somedebateover just how wealthy the Abbasidswerebeforecomingto power: the 3. Oleson eta/. 1993: 484-86; 1995: 343-49; 1999: 436-43; 2003: 55-62. Given the accuracywith which the excavatedqasr-masjidcomplexin Field F103 at ai-Humaymafits the historic descriptions,and that potteryand small finds from initial phasesof constructionand occupationdateto the 8th century,I identified the complexas the Abbasidresidence.

FROM RESIDENCE TO REVOLUTIONARY HEADQUARTERS

461

evidenceat ai-Humayma may indicate they hadsignifi cantly fewer resourcesthantheUmayyadsand wereonly able to aspireto conspicuousconsumption. It could also simply indicatethat they chose to expendtheir capital differently.

lOan

Figure 2. Frescofragment,from Room 02 (the 'FrescoRoom') in the qasr (C. Mundigler)

The ivory panel fragmentshad originall y beencarefully carved into very thin sheets(0.0030.005 m thick), measuring as large as 0.30 x 0.10 m. Six human fig ures and various bird, fish,

vegetal, trefoil , and geometricmotifs are depicted,and several small holes were drilled in each panel. The ivory is elephantine,of either Indian or African origin. Four of the humanfigures are dressedas soldiersand renderedin two sets;eachset of nearly identical and mirrored figures wears military headgear and a tunic over blousy pantaloons,and holdsa spear(Fig. 3). T he two other humanfiguresarealso renderedasa nearly identical set: fullcheekedwith thick wavy hair, dressedthe sameas the soldiers but without helmetsand spears; rather,eachis holding a staff or scabbard and is possiblyseatedon a throne. Someof the burnt wood discoveredin contextwith the coll apsedfrescowas in log-like forms and someflat. The flat wood is ash(Fraxinus), not common in Jordan,and is associatedwith the ivories. Corrodediron nails, hooks, eyes,hinges, and other fastenerswere also discovered.The panelsof wood, the drill holes in the ivory, and the iron fasteners suggestthatthe ivory panelswere veneerattachedto a woodenobject(s);however, the fire burnedthe object(s)to which the ivories wereaffixed beyondrecognition.A few curved ivory piecessuggestthe object(s) m_ay havebeena chestor chair.

462 CROSSING JORDAN

Figure 3. Reconstructedivory veneerpanel depictinga soldierset, from Room 02 (the 'FrescoRoom') in the qasr Q.P. Oleson)

No otherexcavatedroom in theqasr hasproducedsuchrich decorationandluxury objects.The highly ornamentedcharacter of the FrescoRoom, its placementon an almostdirect axis with the primary entranceto the building acrossthe courtyard,and the fine ivory-veneeredobjectsdiscoveredinside all suggestthe room served a ceremonial, reception, or otherspecial function. The facial featuresof all the figuresdepictedin the ivories areclearly not Levantine, pointing to a provenanceeastof the Euphrates - ancientSogdia(Uzbekistan),Persia, or the Indian Subcontinent. The costumingmay give the bestcluesfor placeof origin: the dress,with pantaloonsunder tunics,is clearly Persianat e; andthe differing patternin eachsoldier set'sheadgea r appearsto be the flexible fabric hood or capworn over the actual protectivemetal helmetknown in the Persianate world during the period (Nicolle 1997).

FROM RESIDENCE TO REVOLUTIONARY HEADQUARTERS

463

The military figures depictedin the ivory seemlikely to be evokingthe seditiousactivities of the Abbasidstowardsthe end of the Umayyadperiod (the middle of the 8th century).The ivories may testify to the historically reportedcloseaffiliation betweenthe popularpowerbasefor the Abbasids in Khurasan(northeasternIran) and al-Humayma,which had becometheir revolutionary headquartersafter the 730sCE, over 3000 km away. The 'Eastern'stylistic sensibility and predominant military iconographysuggestthat the object(s)to which the carvedivory panelswere affixed was producedin or aroundnortheasternIran, commissionedby a patronthereto expressmilitary support, and brought to al-Humayma.

Other Small Finds

Just outsidethe qasr doorway, the only Islamic coin yet found at the qasr was unearthed.It is an early Islamic silver dirham struck in the year 115 H (733-4 CE) in Wasit (in southernIraq), which wasthe largestmint at that time. Its presenceindicatesintegrationof the Abbasidoccupantsinto the larger caliphal economy. A numberof incisedsteatitevesselfragmentswere discoveredin the qasr, dating to the earliest phasesof occupation.Steatiteis a soapstone,which is easilycarvedand fashionedinto variousitems (e.g. cookingpots, lamps,incenseburners),most of which exploit the stone'sability to evenly distribute heat.Not native to Jordan,steatitewasquarriedat many sitesin the ArabianPeninsula,with importation to Jordanbeginningin the Late Romanperiod but only becomingwidespreadduring the early Islamic period (7th-10thcenturiesCE) (Grubisha2001). The steatiterecoveredfrom the Umayyad-periodlevels in the qasr demonstratesthat the inhabitantswere affiliated with Hijazi productionthrough gift, purchase,or other form of exchange. Mosque A few meterssoutheastof the qasr there is a small rhomboidalstructure,measuringjust over 5 m along each side, with modern walls built atop its three surviving original coursesof masonry. Establishedon a due north-southaxis, which differs from the qasr, discoveryof a bondedniche (mihrab) alongthe southwall indicatedthat the building was a mosque.Although correctorientation of the qibla (indicatingthe direction of Mecca,towardswhich Muslim prayeris focused)at alHumaymais south-southeast, due southqibla orientationsare not uncommonin the early Islamic period in Bilad al-Sham. Excavationrevealedlater rooms (strangelywithout doors) flanking the mosqueon the eastand west sides,and anothermosqueto the southwest.The secondmosqueis also small and was awkwardly abuttedto the first; its foundationpottery was scantand inconclusive,making an absolute dateunclear.Perhapsit was built shortly after the first mosqueto accommodategenderseparation, or perhapsit was built much later, during the Ottomanperiod, when a major reorganizationand remodelingwas carriedout in the qasr. Most of the early 8th-centuryqusurin the Levanthavemosques,somewithin the qasrandsome as a small separatestructure.Other siteswith extramuralmosquesinclude al-Hallabat,al-Qastal, al-Risha,Umm a!-Walid and al-Zabib in Jordan,and Saysin Syria; while all are of similar form to the first al-Humaymamosque,they are twice the size or more. It is curiouswhy the mosquesat someof thesealreadyremotequsur were situatedoutsidethe qasr when the small size of the mosqueimplies private use.At leastone reasonmay be found in an historic remarkaboutactivities in the mosqueat al-Humayma:in addition to prayer,'Ali is reported to have gatheredthere with othersto dine and discussnews (Akhbar 150-51). With the mosque functioning also as a majlis (receptionroom), its separationfrom the qasr suggeststhe qasr was a private familial domain,and that guestscomingfrom afar were accommodated elsewhere,perhaps in tents.

464

CROSSING JORDAN

Agriculture Locatedat an elevation 600 m lower than the mountainsto the north and the escarpmentto the west, the al-Humaymasettlementrestsin a natural catchmentareameasuring250 km2 at the head of the Hisma Desert.While the areais 'Hyper-Arid', receivinglessthan 100 mm of annualrainfall and with high evaporation,if seasonalwater supplies(i.e. run-off and wadi flows) are harvested cultivation is possible.In addition, the chemicalpropertiesof the local soil provide unexpectedly high agricultural viability for a desertarea(this is to say,the land is not too alkali, saline,or sodic, which are commonhindrances).The archaeologicalevidencefor historic hydraulic featuresconstructedto harvestwater for humanand animal consumptionand for agriculturalcultivation indicatesthat extensiveexploitationof the al-Humaymacatchmentbeganin the Nabataeanperiod and continuedthereafter,with crop irrigation increasingin the later periods.Constructionsto collect anddeflectprecipitationfor agriculturalpurposeincludedimpoundmentdamsor barrierwalls, sets of wadi barriers,and terracing.Botanicalremainsindicatethat domesticatedgrains(barley,wheat) and legumes(esp. lentil) were surely grown locally, and probably tree crops (fig, olive) and grape vines (Oleson1997; Ramsay2004).The '500 olive trees'reportedto exist during the Abbasidoccupation of the site could indeedhave beencultivated and not be hyperbole. Abbasid family memberswere longstandingagricultural property owners in the highlandsof Ta'if, just southwestof their native Mecca. 'Ali's father 'AbdAllah beganto developland in more arid environmentsof the Hijaz during the middle decadesof the 7th century. This was part of a regional phenomenonin which many elites soughtto extendcultivation beyondthe rainfed highlands and springfedoases(e.g. Madina, Khaybar) into the basalthinterlands.Much of the capital, labor, and expertiseto enableconstructionof the hydraulic systemsnecessaryto establishthese plantationscameinto the Hijaz as by-productsof the Islamic conquests(el-Ali 1959). The excavatedevidencefor farming at al-Humaymain the early Islamic period hasparallelsat severalother rural Umayyad-periodqusuralsolocatedin arid zones(50-250mm meanannualrainfall), for exampleat 'Amra, al-'Azraq al-Shishan,al-Hallabat,Ma'an,Mshash,al-Qastal,al-Sil, and Umm al-Walid/Qanatirin Jordan.At thesesites,early Islamic-periodhydraulicinstallationsto harvest and managewadi flows and run-off were discovered,and at some, processinginstallations (mills and presses)and archeobotanicalresults offered further proof; at least grains, grapes,and olives werecertainly cultivated.Pursuingagriculturein the steppeandsemi-deserteastandsouthof the traditional loci of agriculturein the Levant during the Umayyadperiod is likely relatedto the interestin arid farming in the Hijaz during the early Islamic period (Foote, in press).

Concluding Remarks The architectureand its contentsexcavatedat the early 8th-centuryCE Abbasidqasr-mosquecomplex at al-Humaymaoffer the first clear data on the building activity of an elite family other than the Umayyadsduring their reign, and provide the only extantmaterialculture of the Abbasidsprior to their reign. The evidencecontributesto a better understandingof the rural qusurof the Levant (the so-called desert castles/palaces),a centerpiecein early Islamic architectural studies, and expresseshow the hinterlandsite of al-Humaymawas integrated intoan inter-regional caliphate before and while its ownersaspiredto greaterpower.

FROM RESIDENCE TO REVOLUTIONARY HEADQUARTERS

465

References Akhbaral-Dawla a!-'Abbaszyah(1971) (ed. 'A. al-'A. ai-Duri and 'A. al-J. al-Mutallabi; Beirut: Dar al-Tali'ah li al-Tiba'ahwa al-Nashr). al-Bakri, Abu 'Ubayd (1945-51)Mu'jam ma ista'jam min asma'al-buldan wa mawac;li', editedby M. al-Saqa (Cairo: Marba'atLajnat ai-Ta'lif wa ai-Tarjamahwa ai-Nashr). ai-Baladhuri,Ahmad ibn Yai)ya ibn Jabir (1978) Ansabal-Ashraf, III, edited by 'A. ai-'A. ai-Duri (Beirut and Wiesbaden:Yurlabu min F. Shataynir). ei-Ali, S.A. (1959) Muslim Estatesin Hidjaz in the First Century A. H. Journal of the Economicand Social History of the Orient 2(3): 247-61. Foote, R.M. (in press) 500 Olive Trees? Abbasid Family Agriculture at Humayma in Context. Orient Archaologie21. Grubisha,D. (2001) An Analysis of the SteatiteArtifacts from the ArchaeologicalSite of Aila, Jordan (MS thesis.University of Wisconsin-Milwaukee). Nicolle, D. (1997) Arms of the UmayyadEra: Military Technologyin a Time of Change.In War andSocietyin the EasternMediterranean,7th-15th Centuries,edited by Y. Lev (Leiden: Brill): 9-100. Oleson,J.P. (1997) Landscapeand Cityscapein the Hisma: The Resourcesof AncientAl-Humayma.In Studies in the History andArchaeologyofJordan, VI, editedby M. Zaghloul (Amman: Departmentof Antiquities): 175-88. Oleson,J.P.,K. 'Amr, R.M. Foote,J. Logan, M.B. Reeves,and R. Schick (1999) Preliminary Reportof the aiHumaymaExcavationProject, 1995, 1996, 1998. Annualof the DepartmentofAntiquitiesofjordan 43: 411-50. Oleson,J.P., K. 'Amr, R.M. Foote, and R. Schick (1995) Preliminary Report of the Humeima Excavation Project, 1993. Annualof the DepartmentofAntiquitiesofjordan 39: 317-54. Oleson,J.P., K. 'Amr, R. Schick, R.M. Foote, and J. Somogyi-Csizmazia(1993) The HumeimaExcavation Project,Jordan:PreliminaryReportof the 1991-1992Seasons.AnnualoftheDepartmentofAntiquitiesof Jordan 37: 461-502. Oleson,J.P.,G.S. Baker, E. de Bruijn, R.M. Foote,}. Logan,M.B. Reeves,andA.N. Sherwood(2003) Preliminary Reportof ai-HumaymaExcavationProject,2000, 2002. Annual of the DepartmentofAntiquitiesof jordan 47: 37-64.

This page intentionally left blank

53

Paradox of Power Between Local and Imperial at Umm AI-Jimal Bert de Vries

Views from the Imperial Center The interpretiveframeworkfor Hellenistic-Roman-Byzantine archaeologyin the Levant haslived with the Westerntradition of historiographythat viewed the fate of the peoplesof the region from the classicalimperial centers.For outlying centerslike Umm al-Jimal this meanttheir placementon the frontier of empire,and the measureof successor failure, was relatedto the amountof security providedby the respectiveempireagainstvariousperceivedenemies-forexample,Parthian/Persian aggressors,raiding nomads.It was statedthus early in the Umm al-Jimal Project: 'This deliberate fortification provided the stability and safety that enabled the development of permanent settlementslike Umm el-Jimal' (de Vries 1985: 255). A decidedly'post-colonial-theory'anti-Imperial reaction to this model in the 1990s,recently very well summarizedby S.T. Parker (2006: 551-52), stressedthe greaterresponsibility of local populationsfor their own successesand failures, and de-emphasizedthe need for the extensive Romanand Byzantinedefensivestrategies,by arguing,for example,for greatersymbiosisbetween nomadsand sedentists. This revision triggereda shift in frameworkfor explainingthe apparentprosperityof late Umm al-Jimal in volume I of the field reports(de Vries 1998: 232-36).Here I arguedthat the prosperity of Umm al-Jimal in the 5th-8th centuries,apparentfrom its marveloushousesand churches, occurredat the very time that imperial frontier defenseswere breaking down. As a result, local populations-bothsettledand nomadicArab-Aramaicspeakers-were left to their own devicesfor securityand economy.This relatively prosperouscharacterof late antiqueUmm al-Jimal-seenin the high-quality domesticarchitecture-compared very favorably with the remnantsof unsophisticated village architectureat the extra-muralsite of al-Hirri (excavatedas Area R, Momani and Horstmanshof1995), contemporarywith the Roman imperial hegemonyof the 1st to early 4th centuriesAD. The poverty and poor health of this earlier communityis highlighted further by the poverty and poor healthof the good-sizedskeletalsampleexcavatedfrom the simple cist burials in the cemeteriesoutside the immediate perimeterof the built-up areas(Brashier 1995; Cheyney 1997). The debatebetweenthosewho stressthe essential,beneficialrole of empireand thosewho stress its limits has been unresolved.That the first view persistsis evident in two recent important publications.In his final reporton the Limes Arabicus Project,for which I servedas architect,S.T. Parkerconcludes:'The successof Diocletian'sprogramis evidencedby the considerablegrowth in

468 CROSSING JORDAN settlementduring the 4th and5th centuries,when the frontier appearsto havebeenwell protected' (Parker2006: 573-74).And in the latestvolume of his life-time magisterialstudy of the role of the Arabs in the Romanand ByzantineEmpiresbefore Islam, Irfan Shahidarguesthat the Ghassanid phylarchysucceededessentiallyby steppinginto the shoesof the Romandefendersof the frontier. Not only did they convertto Christianityand usemonumentalbuildingsto asserttheir sincerityand majesty,they also adaptedthe Romanlimes defensivestrategyby reusingthe surviving Tetrarchic and early Byzantinefortifications (Shahid2002).

Paradox of Power While I have somedifficulty with both thesepositions (seede Vries 2005 for a review of Shahid 2002), my own extensiveresearchpreoccupationwith Romanfortifications at Umm al-Jimalandin the Limes ArabicusProject(de Vries, Godwin, and Lain 2006) hasbroughtgreatawarenessof the critical massof imperial influenceson the local culturesof the Levant. That critical massis evident centrally in the prevalenceof Romanand Byzantinefortifications at sites like al-Lajjun, Umm alJimal (de Vries 1986),and everywherein the Levant. Whetherwilling or resisting,peoplewere certainly awareof and affectedby the productsof decisionsmadein Antioch, Rome,or Constantinople, and time and again felt the impact of power from such centers.Did they feel exploited, blessed, enriched?On the other hand, they were also centrally locatedin their own domainwithin which they had a measureof power, perhapsfluctuating with the extent of imperial controls, perhaps merely enhancedby sheerdistancefrom imperial centers,protectedby tribal and communalstructures, maintainedby traditionsof languageand craft. Did they feel free, proud, oppressed,stifled? Such questionscannot be answeredreadily from examining pots or buildings, but they do highlight the paradoxesof power in ancientlives. I proposeto adoptthis concept,'the paradoxof power',asthe model for understandingthe cultural history of a local settlementlike Umm al-Jimal. I envisionthis as a two- (multi-)directionalflow of influences,not so much betweencenter(Rome) and periphery (frontier), but betweentwo centers.In perspectiveof the project, in fact Umm alJimal is the center,while the imperial or provincial capital is distant, at the peripheryif you will. Thus our focus of attention,the people and culture of Umm al-Jimal should not be obscuredby interpretingit predominantlyin termsof Romanimperial presenceand phenomena.The operative question is Umm al-Jimal's cultural history, not the history of the influence of Rome. Yet that influenceis there and should be frankly recognized. 'Paradoxof power' makesone awarethat the flow is not just one way, from Rome to Umm alJimal. It is easy,of course,to seethat RomeinfluencedUmm al-Jimal, in its fortifications, in imperial inscriptions,and, contextually,in reamsof literary sourceswritten from the imperial point of view. In contrast,the reverseinfluence of Umm al-Jimal on Rome is virtually undetectable.It is difficult to imagine, let alone prove, that such a tiny and apparentlyremote rural place,without evidenceof a political statureof its own, would havethe powerto affect a distant,looming Rome. Think of one tiny hint, the slogan'Victory to the Blues!' inscribedtwice on stone,in Greek.Could you imaginethis as a shoutreachingall the way to the hippodromein Constantinople?(You could say it was in fact a shout from Constantinople!) Suchreverseflow can be most clearly seenin the larger contextof the history and archaeology of the Levant.An excellent,thoughstill uncommonexpressionof that is presentedby Warwick Ball in Romein the East (2001). Ball's presentationthroughoutthe book is of the local culture of the Levant (the 'East'). While not shying away from Roman influences,he succeedsin showing that often cultural elementsperceivedas Romando in fact havedeeproots in the pre-Romanhistory of the Levant.

PARADOX OF POWER

469

More importantly, he demonstratesconvincingly that the influ ence of the Easton Rome was probably greaterthan Rome'sinfluenceon the East,and he lamentsthe fail ure of most Western scholars' perception andexpressionof this. H e concludesthebookwith due recognition of the twoway flow of influences: 'The infl uenceof Rome on the East for those sevencenturies was profound... But more profoundwasthe influenceof the EastuponRome.The processwasinvariably a two-way one,with ultimately the easternelement predominating'(Ball2001: 450). ThusHorace's famousline beginninggraecia capta becomesroma capta!

Figure 1. The late Nabataeanhalf of the inscription may be found in the west wall of the courtyard of houseVI. 'This is the steleof Fihr, son of Shullai, teacherof Gadhimat,ki ng of the Tanilkh'.

The problem of seeingthe reverseflow is not only thatthe historians,beginning with the Roman ones,have tendedto represen t the winner'spoint of view. It is also that they confusepolitical and military domination with the notion of cultural successand vitality. (That, of course, was, and remains a significant rationale for imperialism.) Again, at a place like Umm ai-Jimal one has to imaginetheremight havebeen resistanceto Romanmilitary andcultural domination, but it cannot be documented. For this one has to look at the not infrequentrebelli ons for context,for example the two Jewish rebellions and the great rampageof Zenobia (indirectly all uded in the famous Greek-Nabataean Gadhimainscription at Umm ai-Jimal, Fig. 1). From such instancesone has to suppose that imperial domination, especiallywhen violently or incompetentlyasserted,is not acceptedwillingly as a benevolent act of cultural enlightenment.The unrecorded feelingsof the people of Umm al-Jimal had to include periodi c animosity and begrudgingsubmissionto acts of awesomeand violent force. For fathoming this, the work done by sociaVcultural anthropologists studyingthe impact of globalizationon modernsocietiesmay provide useful ethnographicparallels (cf. 'powerdistance'in Hofstede 2001). Thus the concept, 'paradoxof power', can bridge the tendencyto champio n eitherthe empire or the local village for a more balancedinterpretationof the archaeological evidenceof the Umm aiJimal Project. Inevitably, this balancewill be weighted towardsUmm ai-Jimal itself, both as the

470

CROSSING JORDAN

immediate objectof study andascorrectionfor the traditional empire-centeredperspective. Also, to seethe modelas a two-way flow is a simplification that mustbe adjusted in practice.Umm al-Jimal isat one poll of manyintersectingpowerrelationships,for example, the Hawran plane,the badiya, theJabalDruze, Nabataean Bosraand Petra,the Decapolis, Provincia Arabia, and so on.

Notes on the A r chaeology and History of Umm ai-J imal Recent work on the Umm ai-Jimal Project publicationsfits in well with the 'paradoxof power'

model.David Graf andSalah Said'spublicationof the new Nabataeaninscriptionsis shedding sig-

nificant light on Umm ai-Jimal underNabataeaninfluencein the 1st-2ndcenturies (Graf and Said 2006), and their in-processwork on a large body of new Greek funerary inscriptionswith its plethora of Arabic personalnamesis yielding further information on theArab-Aramaic community of Umm ai-Jimal.The larger religious context of the religious remainsof the period (the Dushara altar, etc.) was triggered by the PetraExhibit's visit to Calvin Collegein 2005.A joint projectwith one studentis producinga comprehensivecomparative study of the templesof GreaterSyria, in the 1st-2nd centurieswith focuson distinguishing the local, regional,and Romandesign elements (de Vries and Osinga2006).

Figure2. Doorway into a Byzantinerebuilt ' Barracks' room complex. The roomswere reoccupiedby Druze in the early 20th century,who paintedthe lintel and posts. The nameson the lintel, Talal er·Riim (?)andJamal es-Seriirappearto be post-Druze, and testify to the long-lived Arab presenceat Umm al-Jimal.

While 'Roman'Umm ai-Jimal,asa monumentalsite associatedwith the 2nd-century Commodus Gate, has mostly disappeared in later military and domestic construction phases,'local' Umm aiJimal has come to li ght in the lst-3rd-century adjacent settlement of al-Hirri (Momani and Horstmanshof 1995: 472-74). And while the monumenta l Nabataean/Roman chambertombswere

PARADOX OF POWER

471

alreadyrobbedof their elite burials,hundredsof excavatedcist burialsaregiving us insight into the physicalcultureof poorerlocal residents, perhapsmanyassociatedwith al-Hirri, in the 3rd and4th centuries(Brashier1995; Cheyney1997). Publicationof the LimesArabicusProjectreports(Parker2006),which includesmy own work, will enablea better understandingof the in-processinterpretationof the Tetrarchiccastellumat Umm al-Jimal. The puzzleof its early abandonmentasa fort, its replacementby the Barracks(Fig. 2) in 411 (Parker 1998),andits reuseasa site for a masonryquarry,a 5th/6th-centurychurch,and 5th-9th-centuryfarmsteadwill bear heavily on the questionof the waning of Roman imperial control at Umm al-Jimal.

Figure 3. Mangersof UmayyadHouse119, exposedtO ground level in a recentDepartmentof Antiquities project

The local characterof the Byzantinedomesticsite aslocal Arab town prosperingwhile Roman military control was diminishingwas discussedabove. Excavationat House 119 (Fig. 3) pointed dramaticallyto the fact that the domestic prosperitycontinued in the Umayyadperiod (de Vries 1995:422-30),and that the post-Romancommunityhad a continuoushistory from the Byzantine 5th centuryto the Islamic9th (de Vries 2000). The Praetorium,publishedby Robin Brown, is the only monumentalRomanbuilding surviving to the endof Umm al-Jimal. It's transformationsand reusesrepresentthe subsequenthistory of autonomyfollowed by incorporationinto the Islamic Empire. In the5th-6thcenturiesit wasincorporatedinto an elaboratedomesticcomplex,andin the Umayyadperiodit wascompletelyand carefully re-flooredandits walls weredecoratedwith newly frescoedplastersurfaces,akin to the DesertCastles(Brown 1998: 171-93). The role of Umm al-Jimal as an Arab town hasparticularsignificancein the archaeologicalhistory of the emergenceof pre-Islamic sedentarysociety.Irfan Shahid'slatestvolume on the material

472

CROSSING JORDAN

remainsassociatedwith the Ghassanidphylarchy (Shahid2002; de Vries 2005) hashighlightedthe needfor careful study of the patternof Arabic settlementin the Levant, not only to understandthe situationon the eve of the Islamic conquests,but now specificallyto get a much clearerunderstanding of the relationshipbetweenthe mobile and archaeologicallyephemeralGhassanidconfederation and local Arab Christiancommunities.A three-waystudy linking Byzantium,the Ghassanidphylarchy, and sedentarycommunitieswill put placeslike Umm al-Jimal at the center. Use of the 'paradoxof power' conceptpromisesto be a rewardingintegrativedevice for putting togetherall the variousstrandsof the Umm al-Jimal Projectin which imperial influencescan be given their due, but in which the people of Umm al-Jimal will emergeas the 'heroes'of their own history who influencedthe distant politically dominantforces at least as much as they were influencedby them.

References Ball, W. (2001) Romein the East: The Transformationof an Empire (London and New York: Routledge). Brashier,J. (1995)The 1994Umm el-Jimal CemeteryExcavations:AreasM and Z. Annualofthe Department of Antiquitiesofjordan 39: 457-68. Brown, R. (1998) The RomanPreatorium and its DomesticRe-use.In de Vries 1998: 161-94 Cheyney,M. (1997) Age, StatusandGender:Mortality Patternsand Mortuary Practiceat Umm el-Jimal (MA thesis,Kalamazoo,WesternMichigan University). de Vries, B. (1985) Urbanisationin the BasaltRegionof North Jordanin Late Antiquity: The Caseof UmmelJimal. Studiesin the History and ArchaeologyofJordan 2: 249-56. -(1986) Umm el-Jimal in the First Three CenturiesA.D. In The Defenseof the Romanand ByzantineEast, editedby P. Freemanand D. Kennedy(British ArchaeologicalReports,InternationalSeries,297; Oxford: B.A.R.): 227-41. -(1995)The 1993 and 1994Seasonsat Umm el-Jimal.Annualofthe DepartmentofAntiquitiesofjordan 39: 421-35. -(1998) Umm el-]imal: A Frontier Town and its Landscapein Northern jordan. I, Fieldwork 1972-1981 Qournal of Roman Archaeology SupplementarySeries, 26; Portsmouth, Rl: Journal of Roman Archaeology). -(2000) Continuity and Changein the Urban Characterof the SouthernHauran from the 5th to the 9th Century: The ArchaeologicalEvidenceat Umm el-Jimal. MediterraneanArchaeology13: 39-45. -(2005) Reviewof Irfan Shahid,Byzantiumand the Arabsin the Sixth Century.11/1,Toponymy,Monuments, Historical Geographyand Frontier Studies.SpaeculumUanuary): 321-24. de Vries, B., V. Godwin, and A. Lain (2006) The Fortificationsof Lejjun (Area C). In The RomanFrontier in Central jordan: Final Report on the LimesArabicus Project, edited by S.T. Parker(2 vols.; Washington, DC: DumbartonOaks): I, 187-212. de Vries, B., and E. Osinga(2006) The Changesin NabataeanTempleDesignin the Contextof the Hellenistic and Roman Empires. Minds in the Making, Il/4. Online: www.calvin.edu/minds/vol02/issue04/petra_ bdevries.php. Graf, D.F., and S. Said (2006) New NabataeanFuneraryInscriptionsfrom Umm el-Jimal.Journal of Semitic Studies51(2): 267-303. Hofstede, G. (2001) Culture's Consequences:International Differencesin Work-relatedValues (Thousand Oaks, CA: SagePublications). Momani, A., and M. Horstmanshof(1995) Umm el-Jimal Area R 1994 Field Season.Annual of the DepartmentofAntiquitiesofjordan 39: 469-76. Parker,S.T. (1998) The Later castellum('Barracks').In de Vries 1998: 131-42. -(2006) The RomanFrontier in Central jordan: Final Reporton the LimesArabicus Project (2 vols.; Washington, DC: DumbartonOaks). Shahid, I. (2002) Byzantiumand the Arabs in the Sixth Century, 11!1: Toponymy,Monuments,Historical Geographyand Frontier Studies(Washington,DC: DumbartonOaks).

54

Peasants,Pilgrims, and the Body Politic The Northern Jordan Project and the Landscapesof the Islamic Periods Bethany

J. Walker

An Ecological Approach to Islamic Archaeology The traditional image of the Middle Ages in Jordanis a far from attractiveone: Jordanwas colonized, first by the Crusadersin the 12th century, then by the Mamluks (a Cairo-basedmilitaristic stateof slaveorigin) in the following century,and finally by the Ottomans(an expansionistempire with firearms),who nominally held Jordanfrom 1516until the endof World War I. Popularassessment of the long Ottomanperiod has been particularly critical: this was a 'Dark Age', when the statetook from the peopleand their land, through military serviceand taxation,and gavenothing in return. Schoolsand medical clinics were built by local initiative, and the economywas revived throughthe vision and efforts of local entrepreneurs(Walker, underreview c). According to these scenarios,medievalJordanwas little more than a corridor that foreign armiespassedthroughand devastated,its villages regularly pillaged by Bedouin, and its land colonized and economically exploited. Not only are theseimagesdangerouslysimplistic from the perspectiveof the written record,as they deny the give-and-takethat always existedbetweenthe stateand local society, but they also onceprovidedthe languagefor archaeologicalunderstandingof the materialculture of the period, the natureof settlements,and, mostimportantly, the demographicshifts of the 15th century.Until recently,archaeologicalstudiesfrequentlyreferredto the collapseof the urbancentersof the classical period, replaced by a predominantly rural landscapein which even village life eventually declinedas a result of war, high taxes,and neglect.Least understoodwas the Mamluk-Ottoman transitionof the 15th and 16th centuries,which was supposedlycharacterizedby the abandonment of villages acrossthe country and general demographicdecline, a scenario based,in part, on archaeologicalsurveys.These surveys,however, relied on a dating of pottery that reflected the inability, at the time, to properly identify Ottoman wares. Moreover, they did not cover all of Jordanwith the sameintensity: plains and steppehave generallybeenmore systematicallyinvestigatedthan the difficult terrain of the northern hills. The rural decline positedfor the plains does not necessarilyreflect settlementpatternsin other parts of the country.

474

CROSSING JORDAN

In the last fifteen years,significant progresshasbeenmadein the seriationof late medievaland post-16th-centurywares,requiringrevisionsin the wayssiteswere previouslyidentified anddated, as well as a generalrethinking of settlementhistory. Historical sources, while they describethe damagedoneto rural societyby the warsand economichardshipof the 15th century,do not suggestan abandonment of villages acrossthe region. If anything, the opposite is true: contempora ry Arabic recordsconsistentlyattest to the continuedeconomicviability of northernJordan, for example,during the Mamluk andthroughthe Ottomanperiods,anddocumenta continuity in rural settlementthere.This disconnectbetweenthe archaeologicaland written recordseither reflectsa

real regionaldifferencein settlementhistory, or other factors that may have beenat work, which

affectedsettlementdensityanddistribution,suchasinternalmigration{from the plains to the hills) andchangesin subsistence strategies{a moveaway from the productionof cashcropsandintensive agriculture). Either way, the state, and its relations with local society, emergesas an important elementin settlementshifts of the period.

Figure 1. Hills surroundingMaika, in northernjordan

It was with theseconcernsin mind that we began the NorthernJordanProject(NJP) in 2003. Designedto explore patternsof expansionand reduction in village lif e during the Mamluk and Ottomanperiods,our units of study each seasonare individual villages,locatedbetweenIrbid and theYarmouk River, along with their hinterlands.The region is striking for its naturalbeauty,with its sweet-sme lling valleys and wind-sweptmountains,picturesquevi ll ages,extensiveolive groves, and freshair andwater (Fig. 1). The physicallandscapeis punctuatedby old farmhouses,historical mosques , and a wide variety of agriculturaland herding i nstallationsfrom the medievaland early

PEASANTS, PILGRIMS, AND THE BODY POLITIC

475

modern periods. The Sawad'ssuccessfulhistory of market agriculture in medieval and modern times, the extensivehistorical documentationaboutits crops and land in written sourcesfrom the Mamluk and Ottomanperiods,and the potential for recoveryof pollen remainshere for environmental history make it an ideal unit for investigation. The project'sresearchagendais, to a largedegree,generatedby questionsthat emergefrom contemporary Arabic sourcesand directly pertain to settlement: to clarify Mamluk and Ottoman agrarianpolicies in Jordanand assessstateinvestmentin the local agriculturalsectors;to evaluate the environmentalimpactof imperial land-management practicesin the region; and to determineto what degreeregionalmarketsinfluencedcroppingpatternsandsettlementin general.The historical record is a critical resourcefor the project, and fieldwork is precededeach seasonby intensive researchin archivesin Cairo and 'Amman on economicdocumentsrelatedto land use in the study area. By combining archaeologicaldata and the historical record, we hope to write a settlement history for the region that accountsfor economicand environmentalvariablesand is historically informed. Our projectdesign,therefore,requiresmaking useof all forms of datathat havethe potentialto shedlight on the relationshipsbetweenhumanaction and environmentalchangein the contextof settlementchange:archival,archaeological,environmental,and ethnographic.The physicalremains of the villages-theliving, working, andworshippingspaces-are also of interestto the project,for their relationshipto the naturalenvironmentand their contributionto cultural heritage.Fieldwork during the 2003 and 2006 seasonswas conductedin the villages of Maika, Sahm,and Hubras, whose history can be traced to the Mamluk or early Ottoman periodsand which experienceda social and economicrenaissance with the Ottomanreformsof the mid-19th century,known as the Tanzimat.1

The CaseStudies Maika The town of Maika is located23 km northwestof Irbid and 8 km to the eastof Umm Qays. Its earliestdocumentationin Arabic sourcesappearsin late 14th-centurybiographicaldictionaries,with entriesof scholarly'Malkawis',which is the nameof a prominentextendedfamily in Jordantoday.2 What makesMaika unique for this part of Jordanwas its statusas a sultanic estateand its early endowmentby a Mamluk sultan for an institution outsideJordan.The waqfiyyah (endowment document)of SultanBarquqof 796/1393,in scroll form in Cairo, describesin detail the contours, physical condition, and croppingof the village of 'Hay Maika', purchasedby the sultan from the imperial treasury and endowedin its entirety to financially support his madrasah (law school) complex in Cairo (Walker 2004, 2005, 2007). It describesa denselysettledvillage with thriving agriculturebasedon olive oil and wine, a few abandonedhousesand mills but the majority of its installationsfully functional, anda village mosqueendowedwith separatefunds. The identification in 2003 of a collection of olive oil pressesin a cavelocatedwithin the limits of the historical village, at leastpart of which could be datedto the 13th-15thcenturies,atteststo its olive oil productionin this period. Over the courseof the 16th century, its populationhad doubled,and its revenuesin summercrops(melons,beans,and vegetables)and trees(olive trees)were amongthe highestin the region.

1. For weekly reportsfrom the field, seethe project website: www.gvsu.edu/history/walker, NJP tab. 2. The medievalhistory of individual villages is basedon Mamluk-periodchroniclesand biographicaldictionaries and taken from my forthcoming monograph,Jordan in the Late Middle Ages:Transformationof the Mamluk Frontier.

476

CROSSING JORDAN

The village benefitedfrom the land reformsof the OttomanTanzimat,growing in population, prosperity, and influence following its early registration of land in the 1880s. Central to the religiousand social life of the late Ottomanvillage was the shrineof the wali Omar, a Sufi shaykh associatedwith the Qadariyyahorder. The shrine, a single-roomed,domed, and whitewashed building, still standstoday, in the centerof a family cemetery,and was a placeof ritual visitation until the 1950s(Fig. 2). Shaykhsassociatedwith this shrineprovideda variety of importantpublic services,including educationand health care,making Maika an important regional centerby the turn of the century(Walker, underreview c).

Figure 2. Omar shrine in Maika, 19th century

Sahm The modernvillage of Sahmis located22 km northwestof Irbid on a branchof the road leadingto Umm Qays. Situatedon a promontoryhigh above the Wadi Sahm,where the village's irrigated gardensand orchardsare located,the village hasvisual andphysicalproximity to the Syrianborder. The steepwadi dividesthe village into two parts,known traditionally by the residentsas Husnand Masarra.While apparentlynot attestedin medievalsources,the village of Sahmdoesbecomevisible historically in 16th century and again in Tanzimat-period(19th-century) tax registersof the OttomanEmpire.It is amongthe earliestin modernJordanto haveregisteredirs lands,and it grew quickly during the late 19th and early 20th centuries.The village was founded on the basisof marketagricultureand is blessedwith plentiful sourcesof water (at leastfour springs)and a broad planting baseof a variety of summerand winter crops(citrus, apples,olives, wheat). The physicalremainsof the village describedin the Ottomanregistersare locatedon the upper slopes of the hills above Wadi Sahm. At the heart of the original village was the mosque,a

PEASANTS, PILGRIMS, AND THE BODY POLITIC

477

single-room and vaultedstonestructure. This musthave oncebeenpart of a much largercomplex, asthe shaykh's home(long gone),which doubledasthe village'sfirst school house,wasbuilt beside it, and connectedto it was a cemetery, now neglected and separatedfrom the mosque by a road built thirty yearsago. The mosque was built underlocal initiative and appearsto have beenconstructed in two main phases:first in the late Ottomanperiod (19th century)and then again in the mid-20th centu ry. Contemporary with this was the constructionof a public springhouse(funded with local moniesand built by a masonfrom Safedin the 1920s) and the oldeststanding stone homesin the vicinity, the majority built over and incorporatingcaves(usedfor habitation, storage of cools andgrain surplus, animal pens, andlimited food processing-Fig. 3). The modern village is, then, very much a product of the Tanzimatperiod, as land registration and state investment in infrastructure opened opportunities in market agricultureand catalyzed rural growth (Walker, under review b).

Figure 3. Early 20th-centuryfarmhouse,Husn,in Sahm

Hubras

The modernvill age of Hubrasli es 16 km north of Irbid and approximately 6 km south of Sahm. The 'old vill age'consists of several stone farmhousesof the Mandateperiod acrossa modern road (built in the earl y 1970s)from two historical mosques,which formed the core of the original vil lage. Hubras is an old vill age widely attestedin historical sources since the 14th century CE. It appearsin M amluk-period chronicles and biographical dictionariesas a large village with its own market and a place from which many of the intellectual elite of Damascusoriginated (Walker

478

CROSSING JORDAN

2004). Like othervillagesin Jordan,it sufferedsomelevel of depopulationin the late Mamluk!early Ottoman period, only to be rejuvenatedby Ottoman investmentin the secondhalf of the 16th century,when it hadthreemosquesand was surroundedby shrines.It reappearsin written sources in the 1890s,when families in the village beganto register their land for tax purposeswith the Ottomangovernment. The oldeststandingstructurein the largely abandoned'Old Hubras'is a medievalmosque,one of the oldest in Jordan; its sanctuarywent through severalphasesof rebuilding and expansion, culminatingin the large courtyardthat remainstoday. The mosque,in its original form, was small, with a single mihrab and a black and white mosaicfloor. According to an inscription on a minaret (now gone),the sanctuarywas enlargedin the 13th century,to accommodatea village population that had grown significantly in size; a secondmihrab was addedat that point, the mosaic floor replacedwith a flagstonepavement,and piers addedto a systemof pre-existingcolumnsto support a series of cross vaults supporting a flat roof. Excavation revealeda large flagstone pavement outsideand to the eastof this mosque,built sometimein the 19th century,the purposeof which is unclear.The spaceto the eastof the older mosquewas useddomestically,as walls of the late 19th or early 20th centuries,interruptingthe exteriorpavement,indicate.As at Sahm,the mosquein the 19th centurywas part of a larger complex; travel accountsdescribethe shrine (maqam)of a holy man nearby,andelderly residentsof the village today rememberthe remainsof a small stoneshrine to the west of the mosque.The medievalmosque,largely in ruin, continuedto be usedasa placeof prayerand kuttab (religious school) until the Mandateperiod. In 1931 a new, and much smaller, sanctuarywas built in the courtyardof the ruins of this structure,and the older courtyardusedas the village Qur'anic school until the 1960s. By then the population was much reducedfrom its height in the 14th century, when it was one of the largest villages in the region (Walker and Kenney, in press).

On Landscapes and Environments The settlementshifts of the 15th and 16th centuriesalso coincidewith a largertrend of desertification, confirmed by pollen analysis, soil thin sections, Dead Sea salt developmentstudies, and dendrochronology.Referencesin contemporaryArabic sourcesto episodesof droughtwould, on the surface,provide further supportfor a dry cycle in the late medieval periods.Jordanis highly susceptibleto drought: land here hastraditionally beendry farmed,and rainfall levels are in many regionsof the country barely sufficient for rain-fed agriculture.As a result, eventoday, one out of every five years the wheat crop fails for lack of rain. A single bad harvestcan be disastrousto a village community,particularly in the caseof single-cropproductionand in the absenceof adequate droughtrelief by the state.In the most extremeof environmentalevents,villagers may leave their villages,an action describedtime andagainin contemporaryArabic chroniclesin timesof drought, flood, earthquakes,and locustinvasions.Whetheror not they return dependson a combinationof political, economic,andsocial factors. In consideringthe impactof climatechangeon settlement,it is importantto differentiatebetweenclimatic events(a year of drought) and long-termtrends(soil erosion,deforestation,global warming). Imperialland-management strategiesandfinancial reforms, alongwith the political action of peasants,who 'vote with their feet', can exacerbateor ameliorate the effects of climatic events,in the worst circumstancesrenderingcertain forms of agriculture unsustainable. In an effort to understandthe real impact of such environmentalfactors as drought on the history of settlement,andto describemoreclearly the relationshipsamongland use,climatechange, andsoil development,we haveincorporatedintensivesoil (soil genesis)and paleobotanical(pollen and phytolith) analysesinto our project design.Paleobotanicalanalysesdescribethe kinds of crops

PEASANTS, PILGRIMS, AND THE BODY POLITIC

479

grown and the compositionof the natural environmentand can be usedto more fully accountfor changesin the agrarianregime and climate. Soil thin sectionsand coresare used to describethe developmentof soils; takenfrom archaeologicalcontexts(in our case,wall mortar, cisternplaster, fields and field walls, springs), soils can be dated and their compositionaltransformationdocumentedin termsof humanland use.Soil erosionis a particularly useful indicator in this regard,asit is the direct combinedresult of land managementand settlementand climate change(Lucke et al., in press). While thesetechniqueshavematuredin the realmsof soil science,they are relatively new within archaeologicalcircles. Suchanalysis,however,has alreadyborn fruit for this project, as it is documenting the complex interactionsof environmentand land use that may have contributed to settlementshifts during the period of study. Higher temperaturesand reducedrainfall in the late Mamluk period was followed by abandonmentof fields in Maika and Hubras,as well as a shift in agriculturalproductionfrom 'cashcrops'(olives) to a broaderand more varied base(Walker 2005: 95-101). On the other hand,episodesof extremeprecipitationcontributedto architecturalcollapse in Hubrasin the Mandateperiod. In addition, some degreeof deforestation,followed by forest renewal,is indicatedfor the study area.

Conclusions The late Islamic periods in Jordanoffer a unique opportunity to explore the diversity of local responsesto the complex international eventsof the time. While the image of imperialism and colonization is accurate,to some degree,for the Mamluk and Ottoman periods, it does not do justice to the myriad ways in which Jordanianvillages stakedout their own destiny and survived politically troubled times. The settlementand land use histories of villages in northern Jordan documentthe ways in which the villages were incorporatedinto the imperial economy,underwhat circumstancesthey operatedindependently,and how their use of the land left a footprint on the physicallandscapeof the region.The problematic15th centuryis graduallyemerging,asa resultof theseefforts, as one of transformation,rather than overall decline, as local societiesreturnedto traditional agriculturalstrategiesduring the collapseof the Mamluk state.In addition,the Ottoman Tanzimatis an importantturning point in Jordanianvillage history, as villages were resettled(or permanentlysettledfor the first time), laying the demographicand economicfoundationsfor the modernstate. More than ever, the historical record is proving to be an indispensableresourcefor Islamic archaeology.For the Middle andLate Islamic periodsa wide rangeof written sourcesareavailable, mostly in Arabic, from edited and publishedchroniclesto less accessibleeconomicand administrative documentsfound in archivesthroughoutthe Levant. Currenttrendsin Islamic archaeology in Jordanare investigatingland use, the natureof rural settlement,and environmentalimpact, as well as improving chronologiesbasedon ceramicsandarchitecture.Boundariesbetweendisciplines are disappearing,as specialistsfrom archaeology,history, art and architecturalhistory, and soil scienceare closely collaboratingin the study of medievalJordan.

Project-related Publications Lucke, B., Z. ai-Saad,M. Schmidt, R. Baumler,S.O. Lorenz, P. Udluft, K.-U. Heussner,and B.]. Walker (in press)Soils and Land Use in the DecapolisRegion(NorthernJordan):Implicationsfor LandscapeDevelopment and the Impact of Climate Change.Zeitschrift der DeutschenPalaestina-Vereins123. Walker. B.]. (2004) Mamluk Investmentin the Transjordan:A 'Boom and Bust' Economy.Mamluk Studies Review8(2): 119-47.

480

CROSSING JORDAN

-(2005) The NorthernJordanSurvey 2003: Agriculture in Late Islamic Maika and HubrasVillages. A Preliminary Report on the First Season.Bulletin of the AmericanSchoolsof Oriental Research339: 67-111. -(2007)Sowingthe Seedsof Rural Decline?Agriculture as an EconomicBarometerfor Late MamlukJordan. Mamluk StudiesReview11(1): 173-99. -(in pressa) The Role of Agriculture in Mamluk-JordanianPowerRelations.In ProceedingsofRoundtableon theAgeofthe Sultanates,editedby BethanyJ. Walker andJean-Fran